research paper on insomnia pdf

Affiliation

• 1 University of Washington, Seattle, Washington.
• PMID: 33683929
• DOI: 10.7326/AITC202103160

Insomnia-the unwelcome experience of difficulty sleeping-is common and can be acute, intermittent, or chronic. Insomnia can be the presenting symptom for several common sleep disorders, but it also often occurs comorbidly with mental and physical health conditions. Evaluating the symptom of insomnia requires assessing-largely by history-whether an underlying condition explains it. Insomnia disorder is the diagnostic term for the symptom of insomnia that merits specific attention. Cognitive behavioral therapy for insomnia is the preferred treatment approach because of its efficacy, safety, and durability of benefit, but pharmaceutical treatments are widely used for insomnia symptoms.

Publication types

• Research Support, Non-U.S. Gov't
• Hypnotics and Sedatives / therapeutic use
• Risk Factors
• Sleep Initiation and Maintenance Disorders / diagnosis*
• Sleep Initiation and Maintenance Disorders / drug therapy
• Sleep Initiation and Maintenance Disorders / etiology
• Sleep Initiation and Maintenance Disorders / therapy
• Hypnotics and Sedatives

• Open access
• Published: 29 March 2024

Insomnia and creativity in Chinese adolescents: mediation through need for cognition

• Xiaoyang Ren 1 ,
• Min Shi 1 &

BMC Psychology volume  12 , Article number:  180 ( 2024 ) Cite this article

69 Accesses

Metrics details

Creativity is an essential cognitive ability that plays a crucial role in advanced thinking. While previous research has demonstrated the impact of insomnia on cognitive function, its effects on creativity in Chinese adolescents remain unclear. This study explored the relationship between insomnia (specifically, daytime and nighttime disturbances) and creativity in adolescents. Additionally, it examined the potential mediating effect of the need for cognition on this relationship.

Questionnaires were administered to 302 adolescents to measure their creativity, need for cognition, and insomnia levels using the Williams Creative Tendencies Scale, Need for Cognition Scale, and Bergen Insomnia Scale, respectively. Regression analysis was conducted to examine the direct impact of insomnia on creativity. Furthermore, a mediation model was constructed to investigate the role of the need for cognition in mediating the relationship between insomnia and creativity.

The findings of the present study indicated that insomnia had a direct impact on the creativity of adolescents, demonstrating a time-of-day effect. Daytime disturbances were found to have a positive correlation with overall creativity and imagination, whereas no significant direct effect was found between nighttime disturbances and creativity. Further analysis revealed that insomnia, specifically daytime disturbances, might influence creativity by affecting the individual’s need for cognition. However, no similar indirect effects were observed for the relationship between nighttime disturbances and creativity.

Conclusions

Our findings indicate that adolescents might experience improved creativity as a result of daytime disruptions, and the level of need for cognition could play a crucial role in understanding the link between insomnia and creativity in adolescents.

Peer Review reports

Introduction

Insomnia is a condition characterized by an individual’s self-reported difficulties in sleeping [ 1 , 2 ]. It is characterized by symptoms such as taking a long time to fall asleep, waking up frequently during the night, experiencing prolonged periods of wakefulness during sleep, and frequent brief awakenings [ 3 ]. In recent years, issues like staying up late, not getting enough sleep, and struggling to fall asleep have become increasingly prevalent among adolescents [ 4 ]. The White Paper 2023 China Youth and Children’s Sleep Index, released by the China Sleep Research Association, reveals concerning statistics about the sleep patterns of junior high school students in China. According to the report, only 18.9% of these students manage to sleep for more than 8 h, while a staggering 59.4% sleep for less than 7 h. On average, these students only get 6.82 h of sleep, indicating that the majority of them do not meet the recommended amount of sleep. A study conducted in the Shandong Province of China explored the prevalence of sleep problems among adolescents. The findings revealed that 37.44% of adolescents suffered from insufficient sleep, while 26.89% reported experiencing poor sleep quality [ 5 ]. Another meta-analysis, which included 63 studies and a total of 430,422 Chinese adolescents, discovered that 104,802 adolescents experienced sleep disturbances. The overall prevalence of sleep problems was found to be 26%, with junior high school students having a detection rate of 20% [ 6 ]. As widely known, adolescents go through a crucial stage of psychological transformation. Issues such as sleep deprivation and sleep disorders appear to have a significant influence on their mental well-being, especially in terms of cognition and personality development [ 7 , 8 ].

Can tired minds generate creative ideas? Some researchers have found that the cognitive processes utilized before sleep by individuals with insomnia, such as rehearsing, planning, and problem-solving, are similar to the stages involved in creative thinking, such as preparation and incubation [ 9 , 10 ]. As a result, a hypothesis has emerged suggesting that individuals with disrupted sleep might exhibit greater creativity. In addition, it should be noted that disrupted sleep and the widely recognized consequences of sleep deprivation are symptoms of depression and anxiety [ 11 , 12 ], while depression and anxiety have also been associated with creativity [ 13 ]. This suggests that sleep issues could have been prevalent among individuals who are highly creative. However, it is important to consider that sleep problems have been shown to negatively affect cognitive function as well. For example, a study using fMRI have demonstrated that lack of sleep reduced the communication between various brain regions such as the amygdala, dorsolateral prefrontal cortex, dorsal anterior cingulate gyrus, and right inferior frontal gyrus. This weakened functional connectivity could result in a negative bias when it comes to encoding memories [ 14 ]. Additionally, research has found that sleep deprivation could also impact the activity of brain regions involved in fearful learning, namely the prefrontal cortex, hippocampus, and amygdala [ 15 ]. Since the activity of the aforementioned brain regions is crucial for individual creativity, some researchers have also suggested that problems such as sleep deprivation and sleep disorders may produce impairments in cognition, memory, etc., which in turn interfere with creativity [ 16 ].

It is noteworthy that only two studies have delved into the connection between insomnia and creativity until now. Firstly, researchers discovered a positive correlation between insomnia and creativity by comparing the prevalence of sleep disturbances in 30 creative children versus 30 control children. Notably, the highly creative children exhibited a higher incidence of sleep disturbances than the control group [ 17 ]. Subsequently, a recent study indicated a minor direct impact of a global insomnia factor on divergent thinking, implying time-of-day effects where nighttime sleep disturbances positively predicted divergent thinking more strongly than daytime disturbances [ 18 ]. These findings suggest that sleep disturbances may possess some beneficial predictive effects on creativity among children and adolescents. However, there could be disparities in the impact of sleep disturbances during the day and night. Despite this, the majority of existing studies have focused on the influence of insomnia on creative thinking, leaving a gap in research evidence regarding its effects on creative personality. It is established that insomnia is linked to personality traits [ 19 ]. Therefore, the primary objective of this study is to investigate the relationship between adolescents’ creativity (specifically creative personality) and insomnia. Building on the outcomes of previous studies, we hypothesized that insomnia would significantly and positively predict adolescents’ creativity (creative personality).

Although a tenuous link has been established between insomnia and creativity, it is postulated that additional variables might influence this relationship. Taking these observations into account, a crucial question arises: How does insomnia impact creativity? Since coming up with original and useful ideas requires several cognitively demanding processes [ 20 , 21 ], the need for cognition may also play an important role in creativity. The need for cognition refers to an individual’s tendency to engage in and derive pleasure from tasks that require cognitive effort [ 22 ]. Individuals with a strong need for cognition are more prone to innovate and have a deeper interest in addressing challenging problems. For instance, research suggested that those with a high cognitive need were more likely to generate ideas for ambiguous scenarios [ 23 ]. Furthermore, individuals with a strong need for cognition exhibited heightened creativity in problem-solving and possessed more pronounced creative personalities [ 24 , 25 ]. Therefore, the need for cognition might serve as a significant and positive predictor of creative personality [ 26 ]. In considering the role of insomnia in creativity, it is plausible that the need for cognition could act as a mediator, influencing the association between the two variables.

However, there was evidence that insomnia could impact individuals’ willingness to invest more time and effort when faced with complex tasks. The microanalytic model of insomnia highlighted hyperarousal as a key regulatory feature, which could distort perceptions of time and exacerbate the challenges associated with falling asleep and experiencing distress. As a result, the consequences of insomnia on the following day could include fatigue, mood disturbances such as irritability, cognitive impairments, and a reduced ability to engage in or enjoy mentally demanding tasks [ 27 ]. Furthermore, the maintained cognitive model of insomnia suggested that insomniacs tend to worry excessively about sleep and its consequences. This negative cognition leaded to emotional distress, and the resulting anxiety prompted individuals to hyperfocus on internal and external cues related to sleep-related threats. Consequently, this state of anxiety could lead to a lack of interest and motivation in solving complex problems, as well as crowding out the time needed for engaging in mentally challenging tasks [ 28 ]. Supported by neuroimaging and neurobiochemistry evidence, researchers have found that individuals with insomnia often exhibit impairments in various cognitive functions, including episodic memory, working memory, and certain aspects of executive functioning [ 7 ]. Given these findings, it is likely that insomnia can reduce an individual’s cognitive engagement and motivation to seek new knowledge, thereby suppressing the anticipated effect of insomnia on creativity. Therefore, the second objective of our study was to further investigate the psychological mechanisms that underlie the impact of insomnia on creativity. Drawing from the aforementioned theoretical and empirical evidence, we hypothesized that the need for cognition played a mediating role in the relationship between insomnia and creativity.

Taking into account that previous research primarily focused on young adults or children, who exhibited distinct sleep patterns compared to adolescents, the relationship between insomnia and the creativity of adolescents, particularly their creative personality, remained enigmatic. The objective of this study was to explore the impact of insomnia on adolescents’ creativity, specifically their creative personality, and to unravel the underlying mechanisms. Drawing from existing theoretical and empirical research, we postulated that: (1) insomnia, encompassing both daytime and nighttime disturbances, was associated with creativity in adolescents, and there might exist time-of-day effects (H1); and (2) the need for cognition might serve as a mediator between insomnia and creativity (H2).

Materials and methods

After a thorough literature review and consideration of previous research, the research questions and hypotheses were formulated in January 2023. Utilizing a cross-sectional research design, questionnaires were administered to a cohort of middle school students in Jinan, Shandong Province, in April of the same year. These questionnaires aimed to capture data on all the relevant research variables, including creativity, insomnia, and the need for cognition at the same time. Subsequently, the collected data was entered into a database and subjected to rigorous checking and analysis.

Participants and procedure

To ensure the validity and relevance of our study, we collaborated closely with a local school in the recruitment process. Initially, we liaised with the school’s head to disseminate recruitment details. Leveraging the assistance of class teachers, we carefully selected participants based on the following criteria: all participants were required to be native Chinese speakers with normal or corrected vision, exhibit no signs of mental or physical health issues, possess normal intellectual development, not encounter any reading difficulties, and not consume psychotropic drugs. Only students who expressed a willingness to participate and fulfilled the study’s criteria were ultimately chosen to participate in the testing process. This meticulous approach ensured that our sample population was representative and well-suited for the objectives of our research.

In this study, 318 junior high school students participated, of whom 302 were included in the primary analysis due to having complete datasets, yielding an effective participation rate of 94.97%. Participants’ ages ranged from 12 to 14, with an average of 12.97 years ( SD  = 0.49). Specifically, 41 were 12 years old, 229 were 13, and 32 were 14. 147 were females (48.7%) and 155 were males (51.3%). Regarding the parents’ educational backgrounds, the survey revealed that 31 fathers (10.3%) and 34 mothers (11.3%) held university degrees or higher qualifications. Notably, most parents had completed their education at the middle or high school level (70.2%). When it came to parental occupations, the survey found that the fathers’ top three professions were doctors (25.5%), self-employed individuals (11.6%), and drivers (8.3%). Meanwhile, for mothers, the most common occupations were self-employed (19.2%), salespeople (11.6%), and laborers (8.9%).

The Institutional Review Board of Shandong Normal University has granted approval for this study, ensuring that all measurements adhere strictly to the pertinent guidelines and regulations for psychological research. The group tests were conducted within the classroom setting, led by a psychology-major researcher as the primary tester. Initially, we secured the authorization and support of the school’s teaching department. Subsequently, we utilized the students’ self-study period to clarify the purpose of the research and underscore the principles of voluntariness, anonymity, and honesty. Ultimately, the participants were required to complete a psychological test within approximately 30 min, assessing various aspects such as creativity, need for cognition, insomnia, along with personal family information.

An adapted version of the Williams Creative Tendencies Scale (WCTS) was utilized to assess the creativity of the participants [ 29 ]. This scale was widely employed in numerous prior creativity studies and exhibited strong reliability [ 30 , 31 ]. The adapted version included 11 items to measure adventurousness, 14 items to measure curiosity, 13 items to measure imagination, and 12 items to measure challenge. Each item was rated on a Likert scale ranging from 1 (strongly disagree) to 3 (strongly agree). By compiling the total scores, we can effectively evaluate the creativity of the students. Notably, all the items demonstrated good reliability, with a Cronbach’s α value of 0.86.

• Need for cognition

The Need for Cognition Scale (NCS) [ 22 ] was employed in its shortened version to assess participants’ need for cognition. The 18-item Chinese version of the NCS was initially introduced [ 32 ] and subsequently validated as suitable for both adolescents and young adults in subsequent studies [ 33 ]. Participants were instructed to answer the questions based on their actual circumstances. Each item was rated using a Likert-type scale ranging from 1 (strongly opposed) to 5 (strongly agreed). The total score was calculated by summing up the responses to all 18 items, with higher scores indicating a stronger need for cognition. This measurement demonstrated good reliability in the current study, with a Cronbach’s α value of 0.76.

We utilized the Bergen Insomnia Scale (BIS) to assess insomnia among the participants [ 34 ]. The scale comprises six items, all aligned with the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) criteria for clinical insomnia. The validity of these items has been confirmed through subjective reports and polysomnographic data, encompassing sleep-stage progression, limb movement, and physiological measurements of respiration during controlled laboratory sleep. Three items focus on nighttime disturbances, such as “How many days a week did it take you over 30 minutes to fall asleep after switching off the lights in the past month?”; the other three items target daytime disturbances, like “How many days a week did you feel rested upon waking up in the past month?”. Participants were asked to rate their symptoms on a weekly basis using an eight-point scale ranging from 0 to 7. The total score for the first three items represents nighttime disturbances, while the last three items reflect daytime disturbances. This measurement demonstrated strong reliability for daytime disturbances (Cronbach’s α = 0.86), nighttime disturbances (Cronbach’s α = 0.62), and the overall insomnia score (Cronbach’s α = 0.80) in the Chinese population. According to previous literature, this scale demonstrated good reliability among the Chinese population [ 35 ].

Socioeconomic status (SES)

Recognizing the challenges in precisely measuring income, domestic researchers often turn to an alternative method: assessing a family’s socioeconomic status (SES) through a detailed analysis of their parents’ occupation and education level [ 36 ]. In the present study, we utilized the SES questionnaire to gather participants’ reports on their parents’ occupational and educational backgrounds [ 37 ]. These reports were then coded and graded, following established occupational classification standards, to ensure consistency and comparability across respondents. The occupational classification system employed in this study encompassed five distinct levels: (1) those engaged in temporary, unskilled, agricultural, or non-technical work; (2) self-employed individuals, manual laborers, and technicians; (3) general management and professional technical personnel, including clerks and employees in the commercial service industry; (4) middle-level professionals, managers, and technical personnel, as well as auxiliary professionals specializing in various fields of science, technology, and enterprise work; and (5) senior professional technicians, executives, and leading cadres exercising administrative functions in government, institutions, and social organizations, as well as high- and middle-level managers in large and medium-sized enterprises and private business owners. By utilizing this graded classification system, we aimed to capture a comprehensive representation of participants’ SES backgrounds, ensuring the validity and reliability of our findings.

Furthermore, the educational attainment of parents was categorized into distinct levels:“no schooling or primary education”, “junior middle school”, “high school or technical secondary school”, “junior college”, “university (undergraduate)”, and “graduate”. Participants were required to select the most appropriate category based on their parents’ educational qualifications, and each choice was assigned a numerical score ranging from 1 to 6 during the coding process.

Ultimately, the cumulative score served as an indicator of the family’s socioeconomic status, with a potential range spanning from 4 to 22. Notably, in this research endeavor, the SES scores for both mothers and fathers were computed separately, allowing for a nuanced understanding of each parent’s contribution to the overall socioeconomic profile of the family.

Data analysis

First, we employed the Pearson correlation to assess the relationships between the research variables in the present study. To explore the direct impact of insomnia (independent variables) on creativity (dependent variables), we resorted to multiple linear regression analysis. Specifically, gender, age, socioeconomic status of both parents, and insomnia total score (or daytime and nighttime disturbances) were simultaneously entered into the regression equation. Additionally, we utilized the mediation model to delve into the intricate relationships between insomnia, need for cognition, and creativity. To validate the mediation effects, we relied on the bootstrapping method. From the data, 5000 bootstrap samples were drawn, and 95% bootstrap confidence intervals (CI) were computed. For these statistical analyses, we employed SPSS 17.0 process SPSS macro PROCESS (model 4) ( http://www.afhayes.com ) [ 38 ]. This macro has been extensively used and developed for testing complex models incorporating mediating variables [ 39 ].

Common method deviation test

While the self-report method is a popular choice for data collection, it can potentially lead to common method variance (CMV) issues. To mitigate these concerns, we implemented various control measures to safeguard participants’ anonymity. Among these measures, we ensured that the collected data was strictly limited to scientific research purposes and employed reverse expressions for certain items [ 40 ]. Additionally, to enhance the study’s precision, we utilized the Harman single factor test to process the data. Specifically, we conducted a non-rotating principal component factor analysis on the aforementioned items. The results indicated that the first factor explained only 13.66% of the variation (falling below the 40% threshold). Consequently, this study did not exhibit significant common method variance issues in the collected data.

Descriptive statistics of study variables

Table  1 presents the means, standard deviations, bivariate correlations and gender differences among study variables. The independent samples t-test results revealed that females significantly scored higher than males on measures of insomnia, daytime disturbances, and imagination. Our findings further indicated a positive correlation between insomnia and daytime disturbances with imagination, whereas a negative correlation was observed with the need for cognition. Moreover, the need for cognition demonstrated positive associations with the creativity total score, adventure, curiosity, imagination, and challenge. Mother’s socioeconomic status (SES) exhibited a positive association with imagination. The data for all variables had no outliers and were within three standard deviations. The distributions of all variables approached normality, with skewness and kurtosis ranging from − 1 to 1.

Direct effect tests

The collinearity diagnosis revealed that the tolerance values for the variables of insomnia, daytime and nighttime disturbances, and need for cognition were greater than 0.2, ranging from 0.78 to 0.98, indicating the absence of significant collinearity issues.

The regression analysis results demonstrated that insomnia ( β  = 0.19, p  < 0.01) and daytime disturbances ( β  = 0.24, p  < 0.01) positively predicted imagination when controlling for gender, age, father’s SES, and mother’s SES. However, no significant direct effect of nighttime disturbances was observed on the creativity total score, adversity, curiosity, imagination, and challenge (Tables  2 and 3 ). Therefore, H1 was supported. Based on previous research, effect sizes of 0.10, 0.30, and 0.50 are considered small, medium, and large, respectively [ 18 , 41 ]. Consequently, insomnia ( β  = 0.19) and daytime disturbances ( β  = 0.24) exhibited small-to-medium positive effects on creativity, particularly in terms of imagination.

Indirect effect tests

The indirect effect of need for cognition between insomnia and creativity.

Firstly, the total effect of insomnia on creativity was tested, and it was demonstrated that the path coefficient was not significant. Subsequently, the mediating variable of cognition was added to the model to obtain the path type shown in Fig.  1 . The results showed that insomnia had a direct effect on creativity, and the need for cognition played an indirect role between insomnia and creativity (Table  4 ). The bootstrap test was utilized, and 5000 repeated samples were taken to test the mediating effect and estimate the confidence interval. The absence of 0 in the 95% confidence interval suggested that the indirect effect was significant (see Table  5 ). Therefore, H2 was supported. Similar results were found for imagination, while only indirect effects were found for adventure, curiosity, and challenge. According to the recently proposed mediation effect test method [ 42 ], the indirect effect of need for cognition on the relationship between insomnia and creativity was established, which manifested suppression effects. In other words, the inclusion of the need for cognition enhanced the relationship between insomnia and creativity.

Mediation analysis model testing relationships among insomnia, need for cognition (NC) and creativity

The indirect effect of need for cognition between daytime disturbances and creativity

Similar analysis processes were also conducted to investigate the relationship between daytime disturbances and creativity. Testing the total effect of daytime disturbances on creativity revealed that the path coefficient was not significant. The mediating variable, need for cognition, was then added to the model to obtain the path type shown in Fig.  2 . The results showed that daytime disturbances had a direct effect on creativity, and the need for cognition played an indirect role in the relationship between daytime disturbances and creativity (Table  6 ). Finally, the bootstrap test was employed, and 5000 replicated samples were taken to test the mediating effect and establish the confidence interval. The exclusion of 0 from the 95% confidence interval indicated a statistically significant indirect effect (see Table  7 ). Similar results were found for imagination, while for adventure, curiosity, and challenge, only indirect effects were found. Also, the inclusion of the need for cognition enhanced the relationship between daytime disturbances and creativity.

The indirect effect of need for cognition between nighttime disturbances and creativity

Although s imilar analysis processes were also conducted to examine the relationship between nighttime disturbances and creativity, no significant direct of nighttime disturbances or indirect effects of the need for cognition were found (see Supplementary Table S1 , Table S2 ).

Mediation analysis model testing relationships among daytime disturbances (DD), need for cognition (NC) and creativity

Insomnia and creativity in Chinese adolescents

The current study firstly examined the direct effect of insomnia (daytime disturbances, nighttime disturbances) on adolescents’ creativity. Based on our preliminary findings, insomnia was found to have a beneficial impact on the overall creativity score and imagination, aligning with prior research (H1 was supported). Further analysis showed that there indeed existed time-of-day effects: disturbances during the day had a significant effect on imagination, whereas the effect of disturbances during the night was not significant.

It’s worth nothing that the direct impact of insomnia on creativity was limited to imagination. Imagination is the ability to imagine things that have not yet happened and speculate intuitively, transcending the boundaries of the senses and reality [ 43 ]. It is the basis of all creative activities and a crucial part of culture life [ 44 ]. In a state of insomnia, individuals’ minds may be active, which may enhance individuals’ ability to visualize and increase their openness to new ideas and perspectives.

Moreover, our findings revealed a significant direct impact of daytime disturbances solely on imagination, with no comparable effect observed for nighttime disturbances. This seemed to contrast previous research conducted on young adults, indicating that ‘evening types’ - individuals who typically prefer staying up late and waking up late - tend to perform slightly better on certain measures of creativity [ 45 ]. However, it’s crucial to note that the sleep patterns of adolescents differ from those of young adults. Even if they stay up late, adolescents have less chance of waking up late. Consequently, nighttime disturbances may not be advantageous for them. Conversely, daytime disturbances resulting in fatigue and mood swings might lead to less stringent cognitive control, fostering opportunities for unconventional thinking. Hence, it becomes evident that the investigation of insomnia’s influence on creativity should take into account the time-of-day effects. Daytime disturbances appeared to positively predict creativity more strongly than nighttime disturbances in adolescents.

Mediation of need for cognition

Although the direct impact of insomnia on creativity was notable, the majority of the observed effects were of small-to-medium magnitude. Researchers postulated the existence of a third variable that could potentially mediate the relationship between sleep and creativity [ 46 ]. To delve deeper into the influence of insomnia on creativity, we investigated the intermediary role of the need for cognition. Our findings generally indicated that insomnia might exert its influence on creativity by modulating the need for cognition (H2 was supported). The introduction of need for cognition as a variable strengthened the predictive power of insomnia-related factors (such as daytime disturbances) on creative outcomes (like imagination). These observations suggests the emergence of a suppression effect, which refers to a scenario where a third variable attenuates the relationship between an independent variable (X) and a dependent variable (Y), even when the null hypothesis is true. In psychological research, the absence of a direct relationship between X and Y often poses a challenge. The suppression effect offered a valuable framework for addressing such scenarios and elucidating why significant relationships might not be immediately apparent [ 47 ]. Similarly, our results revealed that need for cognition acted as a suppressor, mitigating the effects of insomnia on creativity.

Despite the absence of a significant direct effect of insomnia, the need for cognition was supported as an indirect influence on adventure, curiosity, and challenge. Adventurousness, curiosity, and challenge-seeking all involve cognitive endeavors such as facing failure or criticism, inquiring into the root cause of problems, engaging in confusing situations, and making order out of chaos [ 43 ]. These creative personalities are strongly influenced by their need for cognition, and insomnia may influence them indirectly by altering their need for cognition.

Limitations

Although these findings offer valuable insights, it’s important to acknowledge several limitations. Firstly, the study employed a cross-sectional design, which assessed variables simultaneously, thus lacking evidence of a temporal link between insomnia and creativity. Longitudinal studies are needed to establish a definitive cause-and-effect relationship between these variables. Secondly, the current study primarily focused on creative personality, overlooking the impact of creative cognition. Given that insomnia is a small-to-medium predictor of divergent thinking [ 18 ], it’s crucial to investigate whether the need for cognition mediates this relationship, enhancing our understanding of the factors that truly influence insomnia’s predictive power over creativity. Finally, the findings of this study have not been replicated in other samples, limiting their generalizability. Future research should aim to replicate these results in diverse enrollment groups, particularly those experiencing severe insomnia, to gain a more comprehensive understanding of the phenomenon.

Despite the limitations described, the present study has two strengths. The primary strength is to first reveal the time-of-day effect associated with insomnia and adolescents’ creativity. These preliminary findings offer profound insights into the impact of sleep disturbances on adolescents’ creativity, thereby aiding in the development of accurate sleep concepts and promoting mental well-being. Secondly, insomnia was found to be more likely to influence creativity through affecting need for cognition. These revelations contribute to establishing scientific frameworks for understanding adolescents’ sleep patterns and suggest that the need for cognition is a crucial aspect in examining the link between insomnia and creativity. Notably, the suppression effect of the need for cognition offers an explanation for the tenuous association between insomnia and creativity, providing a theoretical foundation for fostering the emergence and development of creativity among adolescents with insomnia.

Data availability

Data is provided within the manuscript.

Abbreviations

Williams Creative Tendencies Scale

Need for Cognition Scale

Bergen Insomnia Scale

Socioeconomic Status

Common Method Variance

Father Socioeconomic Status

Mother Socioeconomic Status

Daytime Disturbances

Nighttime Disturbances

Need for Cognition

Standard Error

Savard MH, Savard J, Simard S, Ivers H. Empirical validation of the insomnia severity index in cancer patients. Psychol Oncol. 2005;14(6):429–41.

Article   Google Scholar  

Liu S, Zhang B. Interpretation of Chinese guidelines for the diagnosis and treatment of insomnia. Chin J Mod Neurol. 2017;17(9):633–8.

Google Scholar  

Sateia MJ, Doghramji K, Hauri PJ, Morin CM. Evaluation of chronic insomnia. An American Academy of Sleep Medicine review. Sleep. 2000;23:243–308.

Article   PubMed   Google Scholar  

Li L, Sun H. Review of research on bedtime procrastination. China J Health Psychol. 2020;28(2):316–20.

Pei Y, Wang X, Kong L, Ran X, Wei M, Gao J. The correlation between sleep and mental health of primary and secondary school students in Shandong Province. Chin School Health. 2023;44(11):1674–8.

Meng J, Ling G, Jing H, Gou l. Prevalence of sleep disturbances in Chinese adolescents: a systematic review and meta-analysis. PLoS ONE. 2021; 6(3), e0247333.

Fortier-Brochu É, Beaulieu-Bonneau S, Ivers H, Morin CM. Insomnia and daytime cognitive performance: a meta-analysis. Sleep Med Rev. 2012;16(1):83–94.

Kronholm E, Puusniekka R, Jokela J, et al. Trends in self-reported sleep problems, tiredness and related school performance among Finnish adolescents from 1984 to 2011. J Sleep Res. 2015;24:3–10.

Runco MA. Problem finding, problem solving, and creativity. Norwood, NJ: Ablex; 1994.

Wallas G. The art of thought. London: J. Cape; 1926.

Johnson EO, Roth T, Breslau N. The association of insomnia with anxiety disorders and depression: exploration of the direction of risk. J Psychiatr Res. 2006;40(8):700–8.

Neckelmann D, Mykletun A, Dahl AA. Chronic insomnia as a risk factor for developing anxiety and depression. Sleep. 2007;30(7):873–80.

Article   PubMed   PubMed Central   Google Scholar  

Carson SH. Creativity and psychopathology: a shared vulnerability model. Can J Psychiatry. 2011;56(3):144–53.

Shao Y, Lei Y, Wang L et al. Altered resting-state amygdala functional connectivity after 36 hours of total sleep deprivation. PLoS ONE 2014; 9(11), e112222.

Zhang J, Zhang H, Li H, Lei Y. The effect of sleep on fear learning and its cognitive neural mechanism. Adv Psychol Sci. 2023;31(4):631–40.

Dirk DC, Renato P. Sleepy but creative? How affective commitment, knowledge sharing and organizational forgiveness mitigate the dysfunctional effect of insomnia on creative behaviors. Personnel Rev. 2020;50(1):108–28.

Healey D, Runco MA. Could creativity be associated with insomnia? Creativity Res J. 2006;18(1):39–43.

Beaty RE, Silvia PJ, Nusbaum EC, Vartanian O. Tired minds, tired ideas? Exploring insomnia and creativity. Think Skills Creativity. 2013;9:69–75.

Ma G, Cui F. Research progress on the relationship between personality traits and insomnia. Chin J Health Med. 2017;19(1):87–9.

Amabile TM. A model of creativity and innovation in organizations. In Staw BM, Cummings LL, editors. Research in organizational behavior. Greenwich. 1988, 123–167.

Mumford MD, Mobley MI, Reiter-Palmon R, et al. Process analytic models of creative capacities. Creativity Res J. 1991;4:91–122.

Cacioppo JT, Petty RE. The need for cognition. J Personal Soc Psychol. 1982;42:116–31.

Dollinger SJ. Need for uniqueness, need for cognition, and creativity. J Creative Behav. 2003;37:99–116.

Du Z, Zhang L, Wang Y. Relationship between college students cognitive need, specialty commitment and innovative personality. Chin J School Health. 2011;32(4):454–6.

Hahn MH, Lee KC. Exploring the role of self-confidence, need-for-cognition, and the degree of IT support on individual creativity: multilevel analysis approach. Curr Psychol. 2017;36:565–76.

Du Z, Zhang L, Wang Y. The relationship between college students’ cognitive needs, professional commitment and innovative personality. Chin School Health. 2011;32(4):454–6.

Morin CM. Insomnia: psychological assessment and management. Guilford Press. 1993; 56–60.

Harvey AG. A cognitive model of insomnia. Behav Res Ther. 2002;40(8):869–93.

Lin X, Wang M. Williams creative thinking activity manual. Taiwan: Psychology publishing house; 1997.

Li WF, Li XT, Huang LJ, et al. Brain structure links trait creativity to openness to experience. Soc Cognit Affect Neurosci. 2015;10(2):191–8.

Tang G, Zou H, Hou K, et al. The characteristics of family innovation environment and their effects on everyday creative behavior: the mediating role of personality. J Psychol Sci. 2014;37(5):1125–31.

Gao Q. Conceptualization and measurement of need for cognition. Chin J Psychol. 1994;36(1):1–20.

Yang L, Zhang Q. An experimental study on Relationship between intrinsic-extrinsic motivation and Creativity of 1-Grade Junior High School Students. J Southwest China Normal Univ (Natural Science). 2004;29(1):123–6.

Pallesen S, Bjorvatn B, Nordhus IH, et al. A new scale for measuring insomnia: the Bergen Insomnia Scale. Percept Mot Skills. 2008;107(3):691–706.

Wei S, Chen B, Zhao Q, Bai Y. Reliability and validity of the Chinese version of the Bergen Shift Sleep Questionnaire. Chin J Mod Nurs. 2018;24(29):3508–12.

Qiao N, Zhang J, Liu G, Lin C. Effects of family socioeconomic status and parental involvement on junior students’ academic achievements: the moderating role of students’ perception of teachers’ support. Psychol Dev Educ. 2013;29:507–14.

Shi B, Shen J. The relationship between family socioeconomic status, intelligence, internal motivation and creativity. Psychol Dev Educ. 2007;23:30–4.

Hayes AF. Introduction to mediation, moderation, and conditional process analysis. New York: Guilford Press; 2013.

Carciofo R. Morningness–eveningness and affect: the mediating roles of sleep quality and metacognitive beliefs. Sleep Biol Rhythms. 2020;18(1):17–26.

Zhou H, Long L. Statistical remedies for common method biases. Adv Psychol Sci. 2004;12:942–50.

Cohen J. Statistical power for the behavioral sciences. Hillsdale, NJ: Erlbaum; 1988.

Wen Z, Ye B. Mediation effects analysis: methodology and model development. Adv Psychol Sci. 2014;22(5):731–45.

Liu C, Zhang H, Zhang L, et al. The influence of parental rearing style on students’ creativity tendency and self-efficacy. China J Health Psychol. 2013;21(4):589–91.

Vygotsky LS. Imagination and creativity in childhood. J Russian East Eur Psychol. 2004;42(1):7–97.

Giampietro M, Cavallera GM. Morning and evening types and creative thinking. Pers Indiv Differ. 2006;42:453–63.

Liu N, Yuan Y. Relationship between creativity and insomnia in children and adolescents. China J Health Psychol. 2022;30(5):790–4.

Liu Z, Liu T, Mu S. Statistical analysis framework of suppression effect and its application. Psychology: Techniques Appl. 2021;9(10):610–8.

Download references

Acknowledgements

Not applicable.

This research did not receive any specifc grant from funding agencies in the public, commercial, or not-for-proft sectors.

Author information

Authors and affiliations.

Department of Psychology, Shandong Normal University, No.88 East Wenhua Road, 250014, Jinan, China

Xiaoyang Ren, Min Shi & Si Si

You can also search for this author in PubMed   Google Scholar

Contributions

Xiaoyang Ren and Si Si wrote the main manuscript text and Min Shi prepared figures and tables. All authors reviewed the manuscript.Author Contributions Statement.

Corresponding author

Correspondence to Si Si .

Ethics declarations

Ethical approval and consent.

Ethical approval for the study was provided by the Institutional Review Board of Shandong Normal University, Jinan, China. Informed consent from their legal guardian for study participation was obtained. All measurements were performed in accordance with relevant guidelines and regulations of psychological study.

Competing interests

The authors declare no competing interests.

Conflict of interest

The author declared that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Consent for publication

Additional information, publisher’s note.

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Supplementary Material 1

Rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ . The Creative Commons Public Domain Dedication waiver ( http://creativecommons.org/publicdomain/zero/1.0/ ) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Cite this article.

Ren, X., Shi, M. & Si, S. Insomnia and creativity in Chinese adolescents: mediation through need for cognition. BMC Psychol 12 , 180 (2024). https://doi.org/10.1186/s40359-024-01663-3

Download citation

Received : 03 January 2024

Accepted : 15 March 2024

Published : 29 March 2024

DOI : https://doi.org/10.1186/s40359-024-01663-3

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

• Time-of-day effect
• Indirect effect

BMC Psychology

ISSN: 2050-7283

• General enquiries: [email protected]

Help | Advanced Search

Computer Science > Operating Systems

Title: aios: llm agent operating system.

Abstract: The integration and deployment of large language model (LLM)-based intelligent agents have been fraught with challenges that compromise their efficiency and efficacy. Among these issues are sub-optimal scheduling and resource allocation of agent requests over the LLM, the difficulties in maintaining context during interactions between agent and LLM, and the complexities inherent in integrating heterogeneous agents with different capabilities and specializations. The rapid increase of agent quantity and complexity further exacerbates these issues, often leading to bottlenecks and sub-optimal utilization of resources. Inspired by these challenges, this paper presents AIOS, an LLM agent operating system, which embeds large language model into operating systems (OS) as the brain of the OS, enabling an operating system "with soul" -- an important step towards AGI. Specifically, AIOS is designed to optimize resource allocation, facilitate context switch across agents, enable concurrent execution of agents, provide tool service for agents, and maintain access control for agents. We present the architecture of such an operating system, outline the core challenges it aims to resolve, and provide the basic design and implementation of the AIOS. Our experiments on concurrent execution of multiple agents demonstrate the reliability and efficiency of our AIOS modules. Through this, we aim to not only improve the performance and efficiency of LLM agents but also to pioneer for better development and deployment of the AIOS ecosystem in the future. The project is open-source at this https URL .

Submission history

Access paper:.

• HTML (experimental)
• Other Formats

References & Citations

• Google Scholar
• Semantic Scholar

BibTeX formatted citation

Bibliographic and Citation Tools

Code, data and media associated with this article, recommenders and search tools.

• Institution

arXivLabs: experimental projects with community collaborators

arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.

Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.

Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs .

An official website of the United States government

The .gov means it’s official. Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

The site is secure. The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

• Publications
• Account settings

Preview improvements coming to the PMC website in October 2024. Learn More or Try it out now .

• Advanced Search
• Journal List
• Front Psychiatry

The top 100 most cited papers in insomnia: A bibliometric analysis

The number of citations to a paper represents the weight of that work in a particular area of interest. Several highly cited papers are listed in the bibliometric analysis. This study aimed to identify and analyze the 100 most cited papers in insomnia research that might appeal to researchers and clinicians.

We reviewed the Web of Science (WOS) Core Collection database to identify articles from 1985 to 24 March 2022. The R bibliometric package was used to further analyze citation counts, authors, year of publication, source journal, geographical origin, subject, article type, and level of evidence. Word co-occurrence in 100 articles was visualized using VOS viewer software.

A total of 44,654 manuscripts were searched on the Web of Science. Between 2001 and 2021, the top 100 influential manuscripts were published, with a total citation frequency of 38,463. The top countries and institutions contributing to the field were the U.S. and Duke University. Morin C.M. was the most productive author, ranking first in citations. Sleep had the highest number of manuscripts published in the top 100 ( n = 31), followed by Sleep Medicine Reviews ( n = 9). The most cited manuscript (Bastien et al., Sleep Medicine, 2001; 3,384 citations) reported clinical validation of the Insomnia Severity Index (ISI) as a brief screening indicator for insomnia and as an outcome indicator for treatment studies. Co-occurrence analyses suggest that psychiatric disorders combined with insomnia and cognitive behavioral therapy remain future research trends.

This study provides a detailed list of the most cited articles on insomnia. The analysis provides researchers and clinicians with a detailed overview of the most cited papers on insomnia over the past two decades. Notably, COVID-19, anxiety, depression, CBT, and sleep microstructure are potential areas of focus for future research.

Introduction

Insomnia has been emerging with more public concerns over the past decades for affecting people’s health and well-being worldwide. The prevalence of insomnia disorder is approximately 10–20%, with approximately 50% having a chronic course ( 1 ). In America, 27.3% of adults reported insomnia 1 year, and the US annual loss of quality-adjusted life-years associated with insomnia (5.6 million) was significantly larger than that associated with any of the other 18 medical conditions assessed, including arthritis (4.94 million), depression (4.02 million), and hypertension (3.63 million) ( 2 ). The economic consequences of the disorder and the cost-effectiveness of insomnia treatments, in aggregate, exceeded $100 billion per year, with the majority being spent on indirect costs such as poorer workplace performance, increased health care utilization, and increased accident risk ( 3 ). Insomnia has been a public health issue and an extensive concern for medical practitioners. The number of insomnia-associated studies has gradually increased annually, of which 27,399 were published accumulatively in Web of Science (WoS) Core Collection from 1985 to 2021.

With a trend of research interest and explosive publication, it is worth identifying the most influential scientific achievements from an abundance of literature on insomnia related topics. So far, there is no perfect method for evaluating the scientific impact that a specific study has had on a scientific discipline, the number of citations of an article is a proxy to indicate the importance of the study ( 4 ). Bibliometric sciences offer both a statistical and quantitative analysis of published articles and provide a measure of their impact in a particular field of research. To date, no such analyses have been performed exploring the most influential works presented in the field of insomnia. In the present study, we aimed to analyze the top 100 most cited articles over the past decades in the field of insomnia with bibliometric citation analysis.

Identification of the top 100 cited articles

The Clarivate Analytics Web of Science Core Collection database was systematically searched on March 31, 2022. The search terms were “insomnia” and “disorders of initiating and maintaining Sleep,” with publication timespan (1985–2022). The publications were ranked by the number of citations, and these were reviewed to identify the top 100 papers with the most citations. Only original articles and reviews with full manuscripts that focused on insomnia as the main topic were included. Literature reviews that briefly summarized published studies were excluded; editorials and consensus statements were excluded. Two reviewers (SL and JJ) independently identified the top 100 papers according to the total citations of the papers, and any disagreement between the 2 reviewers was resolved by consensus involving a third reviewer (XM).

Analysis of the top 100 cited articles

Publications were stratified and systematically assessed according to publication year, country or institute, authors, and journal. Additionally, the frequencies of keywords extracted from the articles were assessed and then included in a network analysis of the development of insomnia.

All data were downloaded from the Web of Science and imported into the bibliometric package (Version 3.0.0) in R software (Version 4.1.3) ( 5 ), which converts and analyses automatically, including the distribution of countries/regions, years of publication, and authors. Publication quality by author was assessed based upon metrics that included the number of publications, citations in the research area, publication h-index value. The h-index is used to quantify an individual’s scientific research output and measure his citation impact ( 6 ).

Networks were constructed using VOS viewer v.1.6.18 ( 7 ) (Centre for Science and Technology Studies, Leiden University, Leiden, The Netherlands), which is commonly used to analyze and visualize relationships among authors, countries, co-citations, keywords, and the terms used in articles.

The Shapiro–Wilk test was applied to test the normality of the distribution of individual variables. We show the mean and standard deviation for data with a regularly distributed distribution and the median and range for data with a skewed distribution. The Tukey method was also employed for plotting the whiskers and outliers. The p -values from pairwise t -tests were adjusted according to either the Bonferroni post-hoc test or Mann–Whitney test to correct for the performance of multiple statistical analyses. All p -values were two-tailed, and a p -value of ≤0.05 was considered to indicate statistical significance. We used a one-way analysis of Kruskal–Wallis test for skewed data. The Mann–Kendall rank correlation was employed to test for correlations among non-parametric variables.

Global trends of annual publication

A total of 44,654 eligible publications were listed in peer-reviewed journals on the ISI Web of Knowledge WoS Core on 31 March 2022. Manuscripts were screened according to inclusion and exclusion criteria and ranked according to citation frequency. The top 100 influential manuscripts were obtained. General information is detailed in Table 1 .

The top 100 most-cited articles in insomnia.

Over the course of 20 years, the total number of citations for the top 100 works of literature varied, but reached a peak in 2021 ( Figure 1 ). The total citation frequency of the top 100 highly cited literature was 58,229 (ranging from 270 to 3,384), with a mean citation frequency of 582.29 and a median citation frequency of 427.5. To exclude the effect of year on citation volume, we analyzed the average annual citation rate of the 100 documents, the highest of which was “Factors Associated With Mental Health Outcomes Among Health Care Workers Exposed to Coronavirus Disease 2019” by Lai et al. ( 8 ) (average annual citation rate of 893.33; Table 1 ).

Global trends in publications on insomnia research.

Most articles on the list were published from 2005 to 2006 ( n = 24), followed by articles published from 2020 ( n = 9; Figure 2A ). The number of citations was high for articles published between 2001 and 2012 (mean total citations = 3762) and decreased for articles published after 2012, but reached a peak in 20 years for articles published in 2020 (citations = 7852; Figure 2B ). The total citation rate of an article was not related to the date of publication ( r = 0.07108, p > 0.05, Mann–Kendall test; Figure 3A ). However, the current citation rate of an article (as measured by the number of citations in 2021) suggests that articles published after 2011 are more likely to have been cited in recent years. This correlation was statistically significant ( r = 0.5394, p < 0.0001, Mann–Kendall test: Figure 3B ).

(A) The total number of publications for each type of article (clinical or review article) according to publication year. (B) The total number of citations publications for each type of article (clinical or review article) according to publication year. (C) Bar graph showing the number of citations (and standard deviation) for the 100 most-cited articles according to type of article (clinical research, review article). Box: lower linee= Box: lower linee number of= Box: lower linee number of citati= median value, white points = outliers. The Tukey method was used for plotting the whiskers and outliers. *Outlier.

(A) Overall citation rate since publication, and (B) current (2021 = last full year) citation rate for the 100 most-cited articles according to the publication date of the article.

Of the 100 articles, 50 were clinical research, 49 were review articles and 1 was basic research. Due to the small sample size of the basic research, we analyzed the number of citations for review articles and clinical research ( Table 2 ) and found that the review articles did not vary significantly with respect to total citations per article compared to the clinical research articles [Mann–Whitney test, p = 0.08; clinical research: median = 397.5 (range = 275–3384); review articles: median = 531(range = 270–2126): Figure 2C ].

Citations for review articles and clinical research.

* P ≤ 0.05.

Distribution of countries and institutes

The global contribution of insomnia research was analyzed and represented by a blue-coded world map in the R software ( Figure 4A ). Of the 35 countries and territories identified for this study, the USA had the highest number of articles ( n = 56), followed by Canada ( n = 22), Germany ( n = 11), Italy ( n = 10) and the UK ( n = 7) ( Figure 4B ). Studies from the USA were the most cited (24,423 citations), followed by Canada (11,832 citations), Germany (6,329 citations), China (5,587 citations) and the UK (3,097 citations) ( Figure 4C ).

Countries contributing to insomnia research. (A) World map showing the distribution of countries in this field. (B) Top 15 countries with the largest number of publications. (C) Total citations of related articles from different countries.

In the co-authorship analysis, a total of eight countries with more than five publications in the field were analyzed ( Figure 5A ). The five countries with the highest total connection intensity were the United States (total link strength = 19 times), Germany (17 times) and Canada (15 times). A total of 235 institutions are involved in this field. Laval University (38 articles) contributed the most publications, followed by Harvard University (11 articles), Stanford University (10 articles), University of Pittsburgh (10 articles), and Duke University (9 articles). We analyzed the co-authorship of 235 institutions with more than five publications. Eight institutional collaborations are shown ( Figure 5B ). The strongest institutions overall were Duke University (total link strength = 14 times).

Co-authorship analysis of countries and institutions. (A) Network map of co-authorship between countries with more than five publications. (B) Network map of co-authorship between institutions with more than five publications. The thickness of the lines indicates the strength of the relationship.

Analysis of author

Considering the number of publications, MORIN CM. is the most productive author, with 17 articles ( Figure 6A ) MORIN CM. was also the top-ranked author in terms of citations in this field (108 citations) ( Figure 6B ).

Analysis of authors. (A) Number of publications from different authors. (B) Total citations in the research filed from different authors. (C) Network map of co-authorship between authors with more than five publications. Size of the circles indicate the number of articles in the 100 most cited list, while the width of the curved line represents the link strength. The distance between two authors indicates approximate relatedness among the nodes.

We analyzed a total of 481 authors, 60 of whom were co-authors in more than two publications. Excluding 36 unrelated items, 24 authors were shown to have collaborated ( Figure 6C ). The author with the highest total linkage intensity was MORIN CM. (total link strength = 40 times).

Analysis of most cited journal

The 100 articles were published in 42 journals. Figure 7 shows the top ten h-index and cited journals that published related articles ( Figures 7A, B ). Of these 42 journals, the highest h-index was Sleep (h-index = 31), followed closely by Sleep Medicine Reviews (h-index = 9). Sleep was cited the most (928 times), followed by Sleep Medicine Reviews (193 times). In the co-citation analysis, we analyzed a total of 1,352 journals, and a total of 52 journals were cited more than 20 times ( Figure 7C ).

Analysis of journals. (A) Total citations in the research filed from different journals. (B) h-index of publications from different journals. (C) Network map of journals that were co-cited in more than 50 publications. The size of the circle represents the number of papers in the top 100 list.

Co-occurrence analysis of keywords

We analyzed a total of 33 keywords that were identified as appearing more than five times ( Figure 8A ). The colors in the overlay visualization shown in Figure 8B indicate the average year of publication of the identified keywords. The keywords which published after 2011 are colored more green or yellow. The density visualization shows the same identified keywords mapped by frequency of occurrence ( Figure 8C ).

Co-occurrence analysis of keywords. (A) Mapping of keywords of studies. (B) Distribution of keywords according to average publication year (blue: earlier, yellow: later). (C) Distribution of keywords according to the mean frequency of appearance. Keywords in yellow occurred with the highest frequency.

Citation and co-citation analyses

The citation analysis showed 94 pieces of literature with more than 50 citations ( Figure 9A ). As shown in Table 1 , “Validation of the Insomnia Severity Index as an outcome measure for insomnia research” [Bastien et al. ( 8 )] was cited 3,384 times, followed by “Factors Associated With Mental Health Outcomes Among Health Care Workers Exposed to Coronavirus Disease 2019” [Lai et al. ( 8 )] with 2,680 citations and the third most cited is “Epidemiology of insomnia: what we know and what we still need to learn” [Ohayon ( 10 )], with 2,126 citations.

(A) Network map of citation analysis of documents with more than 50 citations. (B) Network map of co-citation analysis of references with more than 10 citations. The size of the circle represents the number of papers in the top 100 list.

We analyzed 38 references that were co-cited more than 10 times in total ( Figure 9B ). The three most cited references were Ford de. ( 107 ) (1989, JAMA-J AM MED ASSOC; 33 citations), Ohayon ( 10 ) mm. (2002, sleep med rev; 30 citations), and Breslau ( 108 ) (1996, bio psychiat; 29 citations).

General trends in insomnia research

Bibliometrics allows for quantitative analysis of a researcher’s individual achievements, or even a country’s or institution’s contribution and international impact in the field, through a statistical analysis of the total number of academic papers published in a clinical field and the total frequency of citations ( 109 ). In this study, we combined bibliometric analysis with network visualization to identify the initial 100 most influential manuscripts in the field of insomnia based on global citation frequency, highlighting the contributions that have led to significant advances in insomnia research and pointing to current trends in the field.

With the largest number of publications and citations, and the highest co-authorship analysis ranking by country, the United States is currently the world leader in insomnia research. These results suggest that the US is likely to have a major impact on the direction of research in the field and has the strongest collaboration globally. The citations of articles from Canada, Germany, Italy and the UK have also increased significantly over the last three decades. China has a small total number of publications, but ranks fourth in total citations; it ranks sixth in collaborations with other countries, indicating that China has an influential publication in the field of insomnia and actively maintains close collaborations with other countries. The Laval University Institute is the most productive, with 38% of the publications, while Duke University ranks first in the co-authorship analysis, indicating its close cooperation with other institutes.

Influential authors and studies in insomnia

Morin, C.M. has the largest number of publications and citations and also ranks first in co-authorship analysis conducted by authors. Of these top 100 highly cited publications, Morin C.M. has published 17 articles, 8 of which he is the first author. Dr. Morin is interested in the validation of assessment scale for insomnia. Utilizing scales to evaluate the therapeutic effects of insomnia is the most convenient and widely used method, and the reliability and validity analysis results are critical for the scale to be used as an outcome indicator. Dr. Morin examined psychometric indices of the Insomnia Severity Index to evaluate treatment response in a clinical sample ( 9 ), and validated the Dysfunctional beliefs and attitudes about sleep, providing a variety of indicators for the assessment of insomnia ( 10 ). Nevertheless, the original version of the ISI and the PSQI are the most commonly used, and the original version of the ISI remains the only validated scale that is highly recommended for all insomnia research protocols. Dr. Morin has also conducted extensive research on cognitive behavioral therapy for insomnia and its comorbidities ( 110 , 111 ). His team’s current efforts continue to focus on evaluating the efficacy of cognitive behavioral therapy and optimizing the procedural approach ( 112 , 113 ).

According to the literature citation analysis and reference co-citation analysis, the most frequently cited Bastien et al. ( 8 ) reported clinical validation of the Insomnia Severity Index Scale (ISI) as a brief screening indicator for insomnia and an outcome indicator for treatment studies, indicating that the ISI is a reliable and valid tool for quantifying perceived insomnia severity. The ISI is a brief self-report instrument designed to assess subjective symptoms and daytime status of insomnia and the extent to which insomnia causes worry or distress. The ISI has been continually validated and the study by Bastien et al. was the first formal psychometric analysis of the reliability and validity of the ISI. Further validation of the ISI using item response theory (IRT) analysis was reported by Morin et al. ( 114 ), obtaining evidence on internal consistency, item response patterns and convergent validity, yielding new evidence on optimal sensitivity and specificity indices for case finding and assessment of minimal important changes following treatment ( 12 ). A recent meta-analysis ( 115 ) reported on the construct validity of the Insomnia Severity Index (ISI), which showed that studies reporting validated factor analyses (CFA) had more reliable results than those reporting only exploratory factor analyses (EFA), and that two-factor solution were strong expressions of dimensionality and higher reliability indicators for the ISI compared to three-factor solutions.

Future outlook

Our co-occurrence network diagram, categorized by subject area or date of publication, shows current hotspots and future directions in insomnia research ( Figures 8A–C ). The keywords indicate that insomnia research involves a wide range of populations (elderly, adolescents), causal factors (quality of life, coronavirus), disorders (anxiety disorders, depression) and therapies (cognitive behavioral therapy, pharmacotherapy). The most recent keywords indicating future trends in the field are as follows:

COVID-19 and insomnia

The total number of citations in the insomnia-related literature rose significantly in 2021, reaching a peak in the last 20 years. This may be related to the outbreak of the Corona Virus Disease 2019 (COVID-19). The impact of the new coronavirus pneumonia outbreak has led to an increase in psychological disorders in the population and a climb in the prevalence of insomnia, with 36.7% of adults from 13 countries having clinical symptoms of insomnia and 17.4% meeting diagnostic criteria for insomnia ( 24 , 114 ). Spielman identified negative life events and other stressors as triggers for the development of insomnia ( 116 ), with up to 37% of the population experiencing insomnia in the presence of stressful events ( 117 ). During the early stage of the COVID-19 pandemic, insomnia symptoms were mainly associated with acute psychological reactions due to the rapid spread of the disease and strict enforcement of restrictions, as well as poor sleep hygiene ( 118 ). During the late stage, insomnia symptoms are associated with economic stress associated with the COVID-19 pandemic ( 118 ) and the impairment of sleep patterns ( 119 ). Recent studies have shown that into the late stages, sleep is characterized by significant objective sleep fragmentation in the presence of adequate sleep duration ( 120 ), suggesting that the adverse effects of the initial pandemic outbreak on sleep will persist. Studies of insomnia during the COVID-19 pandemic highlight the importance of focusing not only on the primary diseases, but also on the psycho-psychological issues, particularly insomnia during global public health events.

Depression, anxiety, and insomnia

In our analysis of keywords, we found that “anxiety,” “depression,” and “mental disorders” were frequently mentioned in 100 documents as the second most frequently occurring keywords after “insomnia.” Studies have shown that there is a strong relationship between insomnia, depression and anxiety, with insomnia considered a risk factor for anxiety and depression ( 121 ), with those suffering from insomnia 9.82 times more likely to have clinically significant depression and 17.35 times more likely to have clinically significant anxiety compared to those without insomnia ( 38 ). Anxiety and depression are also considered risk factors for insomnia ( 122 ), suggesting that insomnia is bilaterally associated with psychiatric disorders such as anxiety and depression ( 41 ). In terms of biological mechanisms, polymorphisms and dysregulation of the serotonin, dopamine(DA), oxytocin (OXT) and genes may be associated with the development and maintenance of insomnia and mood disorders ( 123 ), while behavior and thoughts can in turn affect the activity of the serotonin, DA, OXT, and genes ( 124 ). In terms of brain function, sleep disturbances have been shown to disrupt the function of cortical neural circuits, including the amygdala, striatum, anterior cingulate cortex and prefrontal cortex (PFC) ( 125 ), which play a key role in the regulation of the affective system ( 126 ). In addition, there is growing evidence that insomnia disrupts brain functions associated with the reward system ( 127 , 128 ), and that dysfunction of the reward system is associated with a variety of neuropsychiatric disorders ( 129 ), including depression, bipolar disorder ( 127 , 128 , 130 ) and others.

Subtypes of insomnia

Insomnia is a heterogeneous disorder ( 131 ), and identifying clinically relevant subtypes of insomnia disorders can help reduce heterogeneity, identify etiology, and personalize treatment ( 132 ). In Ohayon ( 10 ) proposed that epidemiological studies should focus on distinguishing different subtypes of insomnia. Typing by sleep stage symptoms, such as difficulty falling asleep (DIS), difficulty maintaining sleep (DMS), early awakening (EMA), or a combination of four subtypes ( 133 ); typing by insomnia episodes and duration, such as chronic insomnia, short-term insomnia ( 134 ); and typing by primary and secondary clinical features of insomnia, such as primary insomnia, secondary insomnia ( 135 ). Although these subtypes can differ in terms of stable sleep-related characteristics, reliability and validity are lacking and heterogeneity still prevails ( 136 ). It remains difficult to find consistent insomnia features in terms of cognition, mood, personality, life history, polysomnography, and sleep microstructure, and this inconsistency suggests that different subtypes of insomnia disorder have not been fully identified ( 137 ). For a long time, researchers have been working on different aspects of the subtypes of insomnia disorders, such as natural history of insomnia ( 94 ), subjective and objective sleep duration ( 74 ), sleep microstructure ( 138 ), non-insomnia characteristics (life history, affective and personality traits) ( 137 ), and clustering subtypes of insomnia (subtyping based on subjective sleep variables as well as age at onset of insomnia, the severity of anxiety and depressive symptoms) ( 139 ). Vgontzas et al. ( 74 ) proposed that insomnia with short objective sleep duration is the most biologically severe phenotype of the disorder and is associated with a higher risk for hypertension, diabetes, and other diseases. Also, it appears that insomnia with objective short sleep duration is a biological marker of genetic predisposition to chronic insomnia. In the future, the underlying genetic, neurobiological, and neuropsychological mechanisms of insomnia with objective short sleep duration could be further explored. In terms of polysomnography ( 140 ), brain imaging ( 141 ), and genetics ( 142 ), we can also examine the association of other sleep variables with other phenotypes of insomnia.

Cognitive behavioral therapy

Cognitive behavioral therapy (CBT) is the most widely researched form of psychotherapy, which leads to changes in emotional distress and problem behavior by altering therapeutic strategies that are maladaptive to poor cognition ( 14 ). CBT for insomnia (CBT-I) has long been shown to be more effective than control therapy ( 143 ). Cognitive behavioral therapy for insomnia (CBT-I) is now commonly recommended as a first-line treatment for chronic insomnia because of the potential for sustained benefit from psychotherapy without the risk of tolerance or adverse effects associated with pharmacological approaches ( 65 ). Recent evidence suggests that CBTI can also be used to treat acute insomnia caused by stress ( 144 ). Many elements of this treatment approach can be applied to stressful events such as the current COVID-19 pandemic and can be adapted to treat and prevent sleep problems resulting from confinement, increased stress and changes in circadian and daily activities ( 78 ). The development of technologies such as the Internet, big data and artificial intelligence has brought about a boom in digital medicine in the healthcare industry, enabling the digitization of CBT-I ( 145 ), and the effectiveness of digital cognitive behavioral therapy (dBT-I) for insomnia has been validated ( 146 ). In recent years, dBT-I has been widely used during the COVID-19 pandemic, and Liu et al’s ( 147 ) study provides an entry point for building a dBTI platform and a theoretical basis for clinical application.

Sleep microstructure

In recent years, sleep microstructure has gradually gained widespread attention, and a number of the 100 articles we examined have begun to focus on slow-wave sleep. Slow wave sleep (SWS) is a component of non-REM sleep that is important in neurophysiological processes like memory and cognition ( 148 ). According to the “active system consolidation hypothesis,” slow oscillations, in conjunction with sleep spindle waves, drive the repetitive reactivation of newly encoded memories during slow-wave sleep, facilitating their integration into long-term memory storage sites ( 149 ). A growing body of research confirms that auditory stimulation ( 150 ), transcranial direct current stimulation ( 151 ), and medication ( 152 ) are all effective in improving memory function by enhancing slow waves of sleep ( 153 ). Slow-wave sleep is not only used for memory enhancement, but has also been widely used to improve cognitive function in patients with mental illness ( 154 , 155 ) and for sensory-motor recovery in stroke patients ( 156 ). Recent studies have shown that enhancing SWS in healthy individuals profoundly affects the connections between the endocrine and autonomic nervous systems ( 157 ), opening up a wide range of potential applications for enhancing SWS.

Strengths and limitations

To the best of our knowledge, this is the first bibliometric analysis of the Insomnia research trend. Using the R bibliometric package, we conducted a comprehensive survey of the literature to perform quantitative and qualitative analyses of the publication output and quality of studies from various authors. We also used a well-known scientometric software tool (VOSviewer) to build and visualize the bibliometric networks by analyzing co-authorship, co-citation, and co-occurrence. Nevertheless, our analyses have some limitations. Firstly, the search is primarily conducted in the WoS database. Although WoS is the most commonly used database in scientometrics, it is advisable to combine the results with those from other databases, such as PubMed and Scopus. Secondly, our search did not separate mechanistic studies from clinical studies, ignoring the research progress in mechanistic studies; however, this could also indicate that mechanistic studies in the field of sleep could be strengthened. Third, the keyword analysis results may have been influenced by incomplete keyword extraction. To better display the keywords, keywords that appeared more than five times in the network were shown. Fourthly, as this is a developing area of research, we may have overlooked the contribution of analyzing recently published studies because of their low citation frequency, despite some studies being published in high quality journals.

In conclusion, to our knowledge, this is the first bibliometric study to identify the 100 most cited papers in insomnia research. Our results suggest that the outbreak of the COVID-19 epidemic is strongly associated with the onset of insomnia and stimulates the researcher’s interest. The key words suggest “COVID-19;” “anxiety,” “depression,” “CBT,” and “sleep microstructure” are currently hot topics in the field of insomnia and will be future research trends in the field, indicating that the focus of research has shifted from insomnia epidemiology and scale validation to the study of co-morbidities and sleep microstructure of insomnia. Despite its limitations, citation analysis provides an important quantitative approach to research in the field of comparative science. The findings of this study may provide a valuable reference for researchers to guide and implement their scientific research interests in the field of insomnia.

Author contributions

QW, KL, and WW designed the study. QW and KL wrote and revised the draft manuscript and carried out data visualization and graphical interpretation. QW, KL, SL, JJ, and XW performed the literature search, retrieval, and data collection. WW provided the critical assistance or funding. All authors contributed and approved the final draft of the manuscript before submission.

Acknowledgments

We acknowledge the support of the Team of the Insomnia Research Team of Jiangsu Province Hospital of Chinese Medicine, Affiliated Hospital of Nanjing University of Chinese Medicine.

This work was supported by the National Natural Science Foundation of China (grant number: 82274631) and Jiangsu Provincial Department of Science and Technology (grant number: BE2021751).

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

• View all journals
• My Account Login
• Explore content
• About the journal
• Publish with us
• Sign up for alerts
• Open access
• Published: 26 March 2024

Predicting and improving complex beer flavor through machine learning

• Michiel Schreurs   ORCID: orcid.org/0000-0002-9449-5619 1 , 2 , 3   na1 ,
• Supinya Piampongsant 1 , 2 , 3   na1 ,
• Miguel Roncoroni   ORCID: orcid.org/0000-0001-7461-1427 1 , 2 , 3   na1 ,
• Lloyd Cool   ORCID: orcid.org/0000-0001-9936-3124 1 , 2 , 3 , 4 ,
• Beatriz Herrera-Malaver   ORCID: orcid.org/0000-0002-5096-9974 1 , 2 , 3 ,
• Christophe Vanderaa   ORCID: orcid.org/0000-0001-7443-5427 4 ,
• Florian A. Theßeling 1 , 2 , 3 ,
• Łukasz Kreft   ORCID: orcid.org/0000-0001-7620-4657 5 ,
• Alexander Botzki   ORCID: orcid.org/0000-0001-6691-4233 5 ,
• Philippe Malcorps 6 ,
• Luk Daenen 6 ,
• Tom Wenseleers   ORCID: orcid.org/0000-0002-1434-861X 4 &
• Kevin J. Verstrepen   ORCID: orcid.org/0000-0002-3077-6219 1 , 2 , 3  

Nature Communications volume  15 , Article number:  2368 ( 2024 ) Cite this article

42k Accesses

786 Altmetric

Metrics details

• Chemical engineering
• Gas chromatography
• Machine learning
• Metabolomics
• Taste receptors

The perception and appreciation of food flavor depends on many interacting chemical compounds and external factors, and therefore proves challenging to understand and predict. Here, we combine extensive chemical and sensory analyses of 250 different beers to train machine learning models that allow predicting flavor and consumer appreciation. For each beer, we measure over 200 chemical properties, perform quantitative descriptive sensory analysis with a trained tasting panel and map data from over 180,000 consumer reviews to train 10 different machine learning models. The best-performing algorithm, Gradient Boosting, yields models that significantly outperform predictions based on conventional statistics and accurately predict complex food features and consumer appreciation from chemical profiles. Model dissection allows identifying specific and unexpected compounds as drivers of beer flavor and appreciation. Adding these compounds results in variants of commercial alcoholic and non-alcoholic beers with improved consumer appreciation. Together, our study reveals how big data and machine learning uncover complex links between food chemistry, flavor and consumer perception, and lays the foundation to develop novel, tailored foods with superior flavors.

Similar content being viewed by others

BitterSweet: Building machine learning models for predicting the bitter and sweet taste of small molecules

Rudraksh Tuwani, Somin Wadhwa & Ganesh Bagler

Sensory lexicon and aroma volatiles analysis of brewing malt

Xiaoxia Su, Miao Yu, … Tianyi Du

Predicting odor from molecular structure: a multi-label classification approach

Kushagra Saini & Venkatnarayan Ramanathan

Introduction

Predicting and understanding food perception and appreciation is one of the major challenges in food science. Accurate modeling of food flavor and appreciation could yield important opportunities for both producers and consumers, including quality control, product fingerprinting, counterfeit detection, spoilage detection, and the development of new products and product combinations (food pairing) 1 , 2 , 3 , 4 , 5 , 6 . Accurate models for flavor and consumer appreciation would contribute greatly to our scientific understanding of how humans perceive and appreciate flavor. Moreover, accurate predictive models would also facilitate and standardize existing food assessment methods and could supplement or replace assessments by trained and consumer tasting panels, which are variable, expensive and time-consuming 7 , 8 , 9 . Lastly, apart from providing objective, quantitative, accurate and contextual information that can help producers, models can also guide consumers in understanding their personal preferences 10 .

Despite the myriad of applications, predicting food flavor and appreciation from its chemical properties remains a largely elusive goal in sensory science, especially for complex food and beverages 11 , 12 . A key obstacle is the immense number of flavor-active chemicals underlying food flavor. Flavor compounds can vary widely in chemical structure and concentration, making them technically challenging and labor-intensive to quantify, even in the face of innovations in metabolomics, such as non-targeted metabolic fingerprinting 13 , 14 . Moreover, sensory analysis is perhaps even more complicated. Flavor perception is highly complex, resulting from hundreds of different molecules interacting at the physiochemical and sensorial level. Sensory perception is often non-linear, characterized by complex and concentration-dependent synergistic and antagonistic effects 15 , 16 , 17 , 18 , 19 , 20 , 21 that are further convoluted by the genetics, environment, culture and psychology of consumers 22 , 23 , 24 . Perceived flavor is therefore difficult to measure, with problems of sensitivity, accuracy, and reproducibility that can only be resolved by gathering sufficiently large datasets 25 . Trained tasting panels are considered the prime source of quality sensory data, but require meticulous training, are low throughput and high cost. Public databases containing consumer reviews of food products could provide a valuable alternative, especially for studying appreciation scores, which do not require formal training 25 . Public databases offer the advantage of amassing large amounts of data, increasing the statistical power to identify potential drivers of appreciation. However, public datasets suffer from biases, including a bias in the volunteers that contribute to the database, as well as confounding factors such as price, cult status and psychological conformity towards previous ratings of the product.

Classical multivariate statistics and machine learning methods have been used to predict flavor of specific compounds by, for example, linking structural properties of a compound to its potential biological activities or linking concentrations of specific compounds to sensory profiles 1 , 26 . Importantly, most previous studies focused on predicting organoleptic properties of single compounds (often based on their chemical structure) 27 , 28 , 29 , 30 , 31 , 32 , 33 , thus ignoring the fact that these compounds are present in a complex matrix in food or beverages and excluding complex interactions between compounds. Moreover, the classical statistics commonly used in sensory science 34 , 35 , 36 , 37 , 38 , 39 require a large sample size and sufficient variance amongst predictors to create accurate models. They are not fit for studying an extensive set of hundreds of interacting flavor compounds, since they are sensitive to outliers, have a high tendency to overfit and are less suited for non-linear and discontinuous relationships 40 .

In this study, we combine extensive chemical analyses and sensory data of a set of different commercial beers with machine learning approaches to develop models that predict taste, smell, mouthfeel and appreciation from compound concentrations. Beer is particularly suited to model the relationship between chemistry, flavor and appreciation. First, beer is a complex product, consisting of thousands of flavor compounds that partake in complex sensory interactions 41 , 42 , 43 . This chemical diversity arises from the raw materials (malt, yeast, hops, water and spices) and biochemical conversions during the brewing process (kilning, mashing, boiling, fermentation, maturation and aging) 44 , 45 . Second, the advent of the internet saw beer consumers embrace online review platforms, such as RateBeer (ZX Ventures, Anheuser-Busch InBev SA/NV) and BeerAdvocate (Next Glass, inc.). In this way, the beer community provides massive data sets of beer flavor and appreciation scores, creating extraordinarily large sensory databases to complement the analyses of our professional sensory panel. Specifically, we characterize over 200 chemical properties of 250 commercial beers, spread across 22 beer styles, and link these to the descriptive sensory profiling data of a 16-person in-house trained tasting panel and data acquired from over 180,000 public consumer reviews. These unique and extensive datasets enable us to train a suite of machine learning models to predict flavor and appreciation from a beer’s chemical profile. Dissection of the best-performing models allows us to pinpoint specific compounds as potential drivers of beer flavor and appreciation. Follow-up experiments confirm the importance of these compounds and ultimately allow us to significantly improve the flavor and appreciation of selected commercial beers. Together, our study represents a significant step towards understanding complex flavors and reinforces the value of machine learning to develop and refine complex foods. In this way, it represents a stepping stone for further computer-aided food engineering applications 46 .

To generate a comprehensive dataset on beer flavor, we selected 250 commercial Belgian beers across 22 different beer styles (Supplementary Fig.  S1 ). Beers with ≤ 4.2% alcohol by volume (ABV) were classified as non-alcoholic and low-alcoholic. Blonds and Tripels constitute a significant portion of the dataset (12.4% and 11.2%, respectively) reflecting their presence on the Belgian beer market and the heterogeneity of beers within these styles. By contrast, lager beers are less diverse and dominated by a handful of brands. Rare styles such as Brut or Faro make up only a small fraction of the dataset (2% and 1%, respectively) because fewer of these beers are produced and because they are dominated by distinct characteristics in terms of flavor and chemical composition.

Extensive analysis identifies relationships between chemical compounds in beer

For each beer, we measured 226 different chemical properties, including common brewing parameters such as alcohol content, iso-alpha acids, pH, sugar concentration 47 , and over 200 flavor compounds (Methods, Supplementary Table  S1 ). A large portion (37.2%) are terpenoids arising from hopping, responsible for herbal and fruity flavors 16 , 48 . A second major category are yeast metabolites, such as esters and alcohols, that result in fruity and solvent notes 48 , 49 , 50 . Other measured compounds are primarily derived from malt, or other microbes such as non- Saccharomyces yeasts and bacteria (‘wild flora’). Compounds that arise from spices or staling are labeled under ‘Others’. Five attributes (caloric value, total acids and total ester, hop aroma and sulfur compounds) are calculated from multiple individually measured compounds.

As a first step in identifying relationships between chemical properties, we determined correlations between the concentrations of the compounds (Fig.  1 , upper panel, Supplementary Data  1 and 2 , and Supplementary Fig.  S2 . For the sake of clarity, only a subset of the measured compounds is shown in Fig.  1 ). Compounds of the same origin typically show a positive correlation, while absence of correlation hints at parameters varying independently. For example, the hop aroma compounds citronellol, and alpha-terpineol show moderate correlations with each other (Spearman’s rho=0.39 and 0.57), but not with the bittering hop component iso-alpha acids (Spearman’s rho=0.16 and −0.07). This illustrates how brewers can independently modify hop aroma and bitterness by selecting hop varieties and dosage time. If hops are added early in the boiling phase, chemical conversions increase bitterness while aromas evaporate, conversely, late addition of hops preserves aroma but limits bitterness 51 . Similarly, hop-derived iso-alpha acids show a strong anti-correlation with lactic acid and acetic acid, likely reflecting growth inhibition of lactic acid and acetic acid bacteria, or the consequent use of fewer hops in sour beer styles, such as West Flanders ales and Fruit beers, that rely on these bacteria for their distinct flavors 52 . Finally, yeast-derived esters (ethyl acetate, ethyl decanoate, ethyl hexanoate, ethyl octanoate) and alcohols (ethanol, isoamyl alcohol, isobutanol, and glycerol), correlate with Spearman coefficients above 0.5, suggesting that these secondary metabolites are correlated with the yeast genetic background and/or fermentation parameters and may be difficult to influence individually, although the choice of yeast strain may offer some control 53 .

Spearman rank correlations are shown. Descriptors are grouped according to their origin (malt (blue), hops (green), yeast (red), wild flora (yellow), Others (black)), and sensory aspect (aroma, taste, palate, and overall appreciation). Please note that for the chemical compounds, for the sake of clarity, only a subset of the total number of measured compounds is shown, with an emphasis on the key compounds for each source. For more details, see the main text and Methods section. Chemical data can be found in Supplementary Data  1 , correlations between all chemical compounds are depicted in Supplementary Fig.  S2 and correlation values can be found in Supplementary Data  2 . See Supplementary Data  4 for sensory panel assessments and Supplementary Data  5 for correlation values between all sensory descriptors.

Interestingly, different beer styles show distinct patterns for some flavor compounds (Supplementary Fig.  S3 ). These observations agree with expectations for key beer styles, and serve as a control for our measurements. For instance, Stouts generally show high values for color (darker), while hoppy beers contain elevated levels of iso-alpha acids, compounds associated with bitter hop taste. Acetic and lactic acid are not prevalent in most beers, with notable exceptions such as Kriek, Lambic, Faro, West Flanders ales and Flanders Old Brown, which use acid-producing bacteria ( Lactobacillus and Pediococcus ) or unconventional yeast ( Brettanomyces ) 54 , 55 . Glycerol, ethanol and esters show similar distributions across all beer styles, reflecting their common origin as products of yeast metabolism during fermentation 45 , 53 . Finally, low/no-alcohol beers contain low concentrations of glycerol and esters. This is in line with the production process for most of the low/no-alcohol beers in our dataset, which are produced through limiting fermentation or by stripping away alcohol via evaporation or dialysis, with both methods having the unintended side-effect of reducing the amount of flavor compounds in the final beer 56 , 57 .

Besides expected associations, our data also reveals less trivial associations between beer styles and specific parameters. For example, geraniol and citronellol, two monoterpenoids responsible for citrus, floral and rose flavors and characteristic of Citra hops, are found in relatively high amounts in Christmas, Saison, and Brett/co-fermented beers, where they may originate from terpenoid-rich spices such as coriander seeds instead of hops 58 .

Tasting panel assessments reveal sensorial relationships in beer

To assess the sensory profile of each beer, a trained tasting panel evaluated each of the 250 beers for 50 sensory attributes, including different hop, malt and yeast flavors, off-flavors and spices. Panelists used a tasting sheet (Supplementary Data  3 ) to score the different attributes. Panel consistency was evaluated by repeating 12 samples across different sessions and performing ANOVA. In 95% of cases no significant difference was found across sessions ( p  > 0.05), indicating good panel consistency (Supplementary Table  S2 ).

Aroma and taste perception reported by the trained panel are often linked (Fig.  1 , bottom left panel and Supplementary Data  4 and 5 ), with high correlations between hops aroma and taste (Spearman’s rho=0.83). Bitter taste was found to correlate with hop aroma and taste in general (Spearman’s rho=0.80 and 0.69), and particularly with “grassy” noble hops (Spearman’s rho=0.75). Barnyard flavor, most often associated with sour beers, is identified together with stale hops (Spearman’s rho=0.97) that are used in these beers. Lactic and acetic acid, which often co-occur, are correlated (Spearman’s rho=0.66). Interestingly, sweetness and bitterness are anti-correlated (Spearman’s rho = −0.48), confirming the hypothesis that they mask each other 59 , 60 . Beer body is highly correlated with alcohol (Spearman’s rho = 0.79), and overall appreciation is found to correlate with multiple aspects that describe beer mouthfeel (alcohol, carbonation; Spearman’s rho= 0.32, 0.39), as well as with hop and ester aroma intensity (Spearman’s rho=0.39 and 0.35).

Similar to the chemical analyses, sensorial analyses confirmed typical features of specific beer styles (Supplementary Fig.  S4 ). For example, sour beers (Faro, Flanders Old Brown, Fruit beer, Kriek, Lambic, West Flanders ale) were rated acidic, with flavors of both acetic and lactic acid. Hoppy beers were found to be bitter and showed hop-associated aromas like citrus and tropical fruit. Malt taste is most detected among scotch, stout/porters, and strong ales, while low/no-alcohol beers, which often have a reputation for being ‘worty’ (reminiscent of unfermented, sweet malt extract) appear in the middle. Unsurprisingly, hop aromas are most strongly detected among hoppy beers. Like its chemical counterpart (Supplementary Fig.  S3 ), acidity shows a right-skewed distribution, with the most acidic beers being Krieks, Lambics, and West Flanders ales.

Tasting panel assessments of specific flavors correlate with chemical composition

We find that the concentrations of several chemical compounds strongly correlate with specific aroma or taste, as evaluated by the tasting panel (Fig.  2 , Supplementary Fig.  S5 , Supplementary Data  6 ). In some cases, these correlations confirm expectations and serve as a useful control for data quality. For example, iso-alpha acids, the bittering compounds in hops, strongly correlate with bitterness (Spearman’s rho=0.68), while ethanol and glycerol correlate with tasters’ perceptions of alcohol and body, the mouthfeel sensation of fullness (Spearman’s rho=0.82/0.62 and 0.72/0.57 respectively) and darker color from roasted malts is a good indication of malt perception (Spearman’s rho=0.54).

Heatmap colors indicate Spearman’s Rho. Axes are organized according to sensory categories (aroma, taste, mouthfeel, overall), chemical categories and chemical sources in beer (malt (blue), hops (green), yeast (red), wild flora (yellow), Others (black)). See Supplementary Data  6 for all correlation values.

Interestingly, for some relationships between chemical compounds and perceived flavor, correlations are weaker than expected. For example, the rose-smelling phenethyl acetate only weakly correlates with floral aroma. This hints at more complex relationships and interactions between compounds and suggests a need for a more complex model than simple correlations. Lastly, we uncovered unexpected correlations. For instance, the esters ethyl decanoate and ethyl octanoate appear to correlate slightly with hop perception and bitterness, possibly due to their fruity flavor. Iron is anti-correlated with hop aromas and bitterness, most likely because it is also anti-correlated with iso-alpha acids. This could be a sign of metal chelation of hop acids 61 , given that our analyses measure unbound hop acids and total iron content, or could result from the higher iron content in dark and Fruit beers, which typically have less hoppy and bitter flavors 62 .

Public consumer reviews complement expert panel data

To complement and expand the sensory data of our trained tasting panel, we collected 180,000 reviews of our 250 beers from the online consumer review platform RateBeer. This provided numerical scores for beer appearance, aroma, taste, palate, overall quality as well as the average overall score.

Public datasets are known to suffer from biases, such as price, cult status and psychological conformity towards previous ratings of a product. For example, prices correlate with appreciation scores for these online consumer reviews (rho=0.49, Supplementary Fig.  S6 ), but not for our trained tasting panel (rho=0.19). This suggests that prices affect consumer appreciation, which has been reported in wine 63 , while blind tastings are unaffected. Moreover, we observe that some beer styles, like lagers and non-alcoholic beers, generally receive lower scores, reflecting that online reviewers are mostly beer aficionados with a preference for specialty beers over lager beers. In general, we find a modest correlation between our trained panel’s overall appreciation score and the online consumer appreciation scores (Fig.  3 , rho=0.29). Apart from the aforementioned biases in the online datasets, serving temperature, sample freshness and surroundings, which are all tightly controlled during the tasting panel sessions, can vary tremendously across online consumers and can further contribute to (among others, appreciation) differences between the two categories of tasters. Importantly, in contrast to the overall appreciation scores, for many sensory aspects the results from the professional panel correlated well with results obtained from RateBeer reviews. Correlations were highest for features that are relatively easy to recognize even for untrained tasters, like bitterness, sweetness, alcohol and malt aroma (Fig.  3 and below).

RateBeer text mining results can be found in Supplementary Data  7 . Rho values shown are Spearman correlation values, with asterisks indicating significant correlations ( p  < 0.05, two-sided). All p values were smaller than 0.001, except for Esters aroma (0.0553), Esters taste (0.3275), Esters aroma—banana (0.0019), Coriander (0.0508) and Diacetyl (0.0134).

Besides collecting consumer appreciation from these online reviews, we developed automated text analysis tools to gather additional data from review texts (Supplementary Data  7 ). Processing review texts on the RateBeer database yielded comparable results to the scores given by the trained panel for many common sensory aspects, including acidity, bitterness, sweetness, alcohol, malt, and hop tastes (Fig.  3 ). This is in line with what would be expected, since these attributes require less training for accurate assessment and are less influenced by environmental factors such as temperature, serving glass and odors in the environment. Consumer reviews also correlate well with our trained panel for 4-vinyl guaiacol, a compound associated with a very characteristic aroma. By contrast, correlations for more specific aromas like ester, coriander or diacetyl are underrepresented in the online reviews, underscoring the importance of using a trained tasting panel and standardized tasting sheets with explicit factors to be scored for evaluating specific aspects of a beer. Taken together, our results suggest that public reviews are trustworthy for some, but not all, flavor features and can complement or substitute taste panel data for these sensory aspects.

Models can predict beer sensory profiles from chemical data

The rich datasets of chemical analyses, tasting panel assessments and public reviews gathered in the first part of this study provided us with a unique opportunity to develop predictive models that link chemical data to sensorial features. Given the complexity of beer flavor, basic statistical tools such as correlations or linear regression may not always be the most suitable for making accurate predictions. Instead, we applied different machine learning models that can model both simple linear and complex interactive relationships. Specifically, we constructed a set of regression models to predict (a) trained panel scores for beer flavor and quality and (b) public reviews’ appreciation scores from beer chemical profiles. We trained and tested 10 different models (Methods), 3 linear regression-based models (simple linear regression with first-order interactions (LR), lasso regression with first-order interactions (Lasso), partial least squares regressor (PLSR)), 5 decision tree models (AdaBoost regressor (ABR), extra trees (ET), gradient boosting regressor (GBR), random forest (RF) and XGBoost regressor (XGBR)), 1 support vector regression (SVR), and 1 artificial neural network (ANN) model.

To compare the performance of our machine learning models, the dataset was randomly split into a training and test set, stratified by beer style. After a model was trained on data in the training set, its performance was evaluated on its ability to predict the test dataset obtained from multi-output models (based on the coefficient of determination, see Methods). Additionally, individual-attribute models were ranked per descriptor and the average rank was calculated, as proposed by Korneva et al. 64 . Importantly, both ways of evaluating the models’ performance agreed in general. Performance of the different models varied (Table  1 ). It should be noted that all models perform better at predicting RateBeer results than results from our trained tasting panel. One reason could be that sensory data is inherently variable, and this variability is averaged out with the large number of public reviews from RateBeer. Additionally, all tree-based models perform better at predicting taste than aroma. Linear models (LR) performed particularly poorly, with negative R 2 values, due to severe overfitting (training set R 2  = 1). Overfitting is a common issue in linear models with many parameters and limited samples, especially with interaction terms further amplifying the number of parameters. L1 regularization (Lasso) successfully overcomes this overfitting, out-competing multiple tree-based models on the RateBeer dataset. Similarly, the dimensionality reduction of PLSR avoids overfitting and improves performance, to some extent. Still, tree-based models (ABR, ET, GBR, RF and XGBR) show the best performance, out-competing the linear models (LR, Lasso, PLSR) commonly used in sensory science 65 .

GBR models showed the best overall performance in predicting sensory responses from chemical information, with R 2 values up to 0.75 depending on the predicted sensory feature (Supplementary Table  S4 ). The GBR models predict consumer appreciation (RateBeer) better than our trained panel’s appreciation (R 2 value of 0.67 compared to R 2 value of 0.09) (Supplementary Table  S3 and Supplementary Table  S4 ). ANN models showed intermediate performance, likely because neural networks typically perform best with larger datasets 66 . The SVR shows intermediate performance, mostly due to the weak predictions of specific attributes that lower the overall performance (Supplementary Table  S4 ).

Model dissection identifies specific, unexpected compounds as drivers of consumer appreciation

Next, we leveraged our models to infer important contributors to sensory perception and consumer appreciation. Consumer preference is a crucial sensory aspects, because a product that shows low consumer appreciation scores often does not succeed commercially 25 . Additionally, the requirement for a large number of representative evaluators makes consumer trials one of the more costly and time-consuming aspects of product development. Hence, a model for predicting chemical drivers of overall appreciation would be a welcome addition to the available toolbox for food development and optimization.

Since GBR models on our RateBeer dataset showed the best overall performance, we focused on these models. Specifically, we used two approaches to identify important contributors. First, rankings of the most important predictors for each sensorial trait in the GBR models were obtained based on impurity-based feature importance (mean decrease in impurity). High-ranked parameters were hypothesized to be either the true causal chemical properties underlying the trait, to correlate with the actual causal properties, or to take part in sensory interactions affecting the trait 67 (Fig.  4A ). In a second approach, we used SHAP 68 to determine which parameters contributed most to the model for making predictions of consumer appreciation (Fig.  4B ). SHAP calculates parameter contributions to model predictions on a per-sample basis, which can be aggregated into an importance score.

A The impurity-based feature importance (mean deviance in impurity, MDI) calculated from the Gradient Boosting Regression (GBR) model predicting RateBeer appreciation scores. The top 15 highest ranked chemical properties are shown. B SHAP summary plot for the top 15 parameters contributing to our GBR model. Each point on the graph represents a sample from our dataset. The color represents the concentration of that parameter, with bluer colors representing low values and redder colors representing higher values. Greater absolute values on the horizontal axis indicate a higher impact of the parameter on the prediction of the model. C Spearman correlations between the 15 most important chemical properties and consumer overall appreciation. Numbers indicate the Spearman Rho correlation coefficient, and the rank of this correlation compared to all other correlations. The top 15 important compounds were determined using SHAP (panel B).

Both approaches identified ethyl acetate as the most predictive parameter for beer appreciation (Fig.  4 ). Ethyl acetate is the most abundant ester in beer with a typical ‘fruity’, ‘solvent’ and ‘alcoholic’ flavor, but is often considered less important than other esters like isoamyl acetate. The second most important parameter identified by SHAP is ethanol, the most abundant beer compound after water. Apart from directly contributing to beer flavor and mouthfeel, ethanol drastically influences the physical properties of beer, dictating how easily volatile compounds escape the beer matrix to contribute to beer aroma 69 . Importantly, it should also be noted that the importance of ethanol for appreciation is likely inflated by the very low appreciation scores of non-alcoholic beers (Supplementary Fig.  S4 ). Despite not often being considered a driver of beer appreciation, protein level also ranks highly in both approaches, possibly due to its effect on mouthfeel and body 70 . Lactic acid, which contributes to the tart taste of sour beers, is the fourth most important parameter identified by SHAP, possibly due to the generally high appreciation of sour beers in our dataset.

Interestingly, some of the most important predictive parameters for our model are not well-established as beer flavors or are even commonly regarded as being negative for beer quality. For example, our models identify methanethiol and ethyl phenyl acetate, an ester commonly linked to beer staling 71 , as a key factor contributing to beer appreciation. Although there is no doubt that high concentrations of these compounds are considered unpleasant, the positive effects of modest concentrations are not yet known 72 , 73 .

To compare our approach to conventional statistics, we evaluated how well the 15 most important SHAP-derived parameters correlate with consumer appreciation (Fig.  4C ). Interestingly, only 6 of the properties derived by SHAP rank amongst the top 15 most correlated parameters. For some chemical compounds, the correlations are so low that they would have likely been considered unimportant. For example, lactic acid, the fourth most important parameter, shows a bimodal distribution for appreciation, with sour beers forming a separate cluster, that is missed entirely by the Spearman correlation. Additionally, the correlation plots reveal outliers, emphasizing the need for robust analysis tools. Together, this highlights the need for alternative models, like the Gradient Boosting model, that better grasp the complexity of (beer) flavor.

Finally, to observe the relationships between these chemical properties and their predicted targets, partial dependence plots were constructed for the six most important predictors of consumer appreciation 74 , 75 , 76 (Supplementary Fig.  S7 ). One-way partial dependence plots show how a change in concentration affects the predicted appreciation. These plots reveal an important limitation of our models: appreciation predictions remain constant at ever-increasing concentrations. This implies that once a threshold concentration is reached, further increasing the concentration does not affect appreciation. This is false, as it is well-documented that certain compounds become unpleasant at high concentrations, including ethyl acetate (‘nail polish’) 77 and methanethiol (‘sulfury’ and ‘rotten cabbage’) 78 . The inability of our models to grasp that flavor compounds have optimal levels, above which they become negative, is a consequence of working with commercial beer brands where (off-)flavors are rarely too high to negatively impact the product. The two-way partial dependence plots show how changing the concentration of two compounds influences predicted appreciation, visualizing their interactions (Supplementary Fig.  S7 ). In our case, the top 5 parameters are dominated by additive or synergistic interactions, with high concentrations for both compounds resulting in the highest predicted appreciation.

To assess the robustness of our best-performing models and model predictions, we performed 100 iterations of the GBR, RF and ET models. In general, all iterations of the models yielded similar performance (Supplementary Fig.  S8 ). Moreover, the main predictors (including the top predictors ethanol and ethyl acetate) remained virtually the same, especially for GBR and RF. For the iterations of the ET model, we did observe more variation in the top predictors, which is likely a consequence of the model’s inherent random architecture in combination with co-correlations between certain predictors. However, even in this case, several of the top predictors (ethanol and ethyl acetate) remain unchanged, although their rank in importance changes (Supplementary Fig.  S8 ).

Next, we investigated if a combination of RateBeer and trained panel data into one consolidated dataset would lead to stronger models, under the hypothesis that such a model would suffer less from bias in the datasets. A GBR model was trained to predict appreciation on the combined dataset. This model underperformed compared to the RateBeer model, both in the native case and when including a dataset identifier (R 2  = 0.67, 0.26 and 0.42 respectively). For the latter, the dataset identifier is the most important feature (Supplementary Fig.  S9 ), while most of the feature importance remains unchanged, with ethyl acetate and ethanol ranking highest, like in the original model trained only on RateBeer data. It seems that the large variation in the panel dataset introduces noise, weakening the models’ performances and reliability. In addition, it seems reasonable to assume that both datasets are fundamentally different, with the panel dataset obtained by blind tastings by a trained professional panel.

Lastly, we evaluated whether beer style identifiers would further enhance the model’s performance. A GBR model was trained with parameters that explicitly encoded the styles of the samples. This did not improve model performance (R2 = 0.66 with style information vs R2 = 0.67). The most important chemical features are consistent with the model trained without style information (eg. ethanol and ethyl acetate), and with the exception of the most preferred (strong ale) and least preferred (low/no-alcohol) styles, none of the styles were among the most important features (Supplementary Fig.  S9 , Supplementary Table  S5 and S6 ). This is likely due to a combination of style-specific chemical signatures, such as iso-alpha acids and lactic acid, that implicitly convey style information to the original models, as well as the low number of samples belonging to some styles, making it difficult for the model to learn style-specific patterns. Moreover, beer styles are not rigorously defined, with some styles overlapping in features and some beers being misattributed to a specific style, all of which leads to more noise in models that use style parameters.

Model validation

To test if our predictive models give insight into beer appreciation, we set up experiments aimed at improving existing commercial beers. We specifically selected overall appreciation as the trait to be examined because of its complexity and commercial relevance. Beer flavor comprises a complex bouquet rather than single aromas and tastes 53 . Hence, adding a single compound to the extent that a difference is noticeable may lead to an unbalanced, artificial flavor. Therefore, we evaluated the effect of combinations of compounds. Because Blond beers represent the most extensive style in our dataset, we selected a beer from this style as the starting material for these experiments (Beer 64 in Supplementary Data  1 ).

In the first set of experiments, we adjusted the concentrations of compounds that made up the most important predictors of overall appreciation (ethyl acetate, ethanol, lactic acid, ethyl phenyl acetate) together with correlated compounds (ethyl hexanoate, isoamyl acetate, glycerol), bringing them up to 95 th percentile ethanol-normalized concentrations (Methods) within the Blond group (‘Spiked’ concentration in Fig.  5A ). Compared to controls, the spiked beers were found to have significantly improved overall appreciation among trained panelists, with panelist noting increased intensity of ester flavors, sweetness, alcohol, and body fullness (Fig.  5B ). To disentangle the contribution of ethanol to these results, a second experiment was performed without the addition of ethanol. This resulted in a similar outcome, including increased perception of alcohol and overall appreciation.

Adding the top chemical compounds, identified as best predictors of appreciation by our model, into poorly appreciated beers results in increased appreciation from our trained panel. Results of sensory tests between base beers and those spiked with compounds identified as the best predictors by the model. A Blond and Non/Low-alcohol (0.0% ABV) base beers were brought up to 95th-percentile ethanol-normalized concentrations within each style. B For each sensory attribute, tasters indicated the more intense sample and selected the sample they preferred. The numbers above the bars correspond to the p values that indicate significant changes in perceived flavor (two-sided binomial test: alpha 0.05, n  = 20 or 13).

In a last experiment, we tested whether using the model’s predictions can boost the appreciation of a non-alcoholic beer (beer 223 in Supplementary Data  1 ). Again, the addition of a mixture of predicted compounds (omitting ethanol, in this case) resulted in a significant increase in appreciation, body, ester flavor and sweetness.

Predicting flavor and consumer appreciation from chemical composition is one of the ultimate goals of sensory science. A reliable, systematic and unbiased way to link chemical profiles to flavor and food appreciation would be a significant asset to the food and beverage industry. Such tools would substantially aid in quality control and recipe development, offer an efficient and cost-effective alternative to pilot studies and consumer trials and would ultimately allow food manufacturers to produce superior, tailor-made products that better meet the demands of specific consumer groups more efficiently.

A limited set of studies have previously tried, to varying degrees of success, to predict beer flavor and beer popularity based on (a limited set of) chemical compounds and flavors 79 , 80 . Current sensitive, high-throughput technologies allow measuring an unprecedented number of chemical compounds and properties in a large set of samples, yielding a dataset that can train models that help close the gaps between chemistry and flavor, even for a complex natural product like beer. To our knowledge, no previous research gathered data at this scale (250 samples, 226 chemical parameters, 50 sensory attributes and 5 consumer scores) to disentangle and validate the chemical aspects driving beer preference using various machine-learning techniques. We find that modern machine learning models outperform conventional statistical tools, such as correlations and linear models, and can successfully predict flavor appreciation from chemical composition. This could be attributed to the natural incorporation of interactions and non-linear or discontinuous effects in machine learning models, which are not easily grasped by the linear model architecture. While linear models and partial least squares regression represent the most widespread statistical approaches in sensory science, in part because they allow interpretation 65 , 81 , 82 , modern machine learning methods allow for building better predictive models while preserving the possibility to dissect and exploit the underlying patterns. Of the 10 different models we trained, tree-based models, such as our best performing GBR, showed the best overall performance in predicting sensory responses from chemical information, outcompeting artificial neural networks. This agrees with previous reports for models trained on tabular data 83 . Our results are in line with the findings of Colantonio et al. who also identified the gradient boosting architecture as performing best at predicting appreciation and flavor (of tomatoes and blueberries, in their specific study) 26 . Importantly, besides our larger experimental scale, we were able to directly confirm our models’ predictions in vivo.

Our study confirms that flavor compound concentration does not always correlate with perception, suggesting complex interactions that are often missed by more conventional statistics and simple models. Specifically, we find that tree-based algorithms may perform best in developing models that link complex food chemistry with aroma. Furthermore, we show that massive datasets of untrained consumer reviews provide a valuable source of data, that can complement or even replace trained tasting panels, especially for appreciation and basic flavors, such as sweetness and bitterness. This holds despite biases that are known to occur in such datasets, such as price or conformity bias. Moreover, GBR models predict taste better than aroma. This is likely because taste (e.g. bitterness) often directly relates to the corresponding chemical measurements (e.g., iso-alpha acids), whereas such a link is less clear for aromas, which often result from the interplay between multiple volatile compounds. We also find that our models are best at predicting acidity and alcohol, likely because there is a direct relation between the measured chemical compounds (acids and ethanol) and the corresponding perceived sensorial attribute (acidity and alcohol), and because even untrained consumers are generally able to recognize these flavors and aromas.

The predictions of our final models, trained on review data, hold even for blind tastings with small groups of trained tasters, as demonstrated by our ability to validate specific compounds as drivers of beer flavor and appreciation. Since adding a single compound to the extent of a noticeable difference may result in an unbalanced flavor profile, we specifically tested our identified key drivers as a combination of compounds. While this approach does not allow us to validate if a particular single compound would affect flavor and/or appreciation, our experiments do show that this combination of compounds increases consumer appreciation.

It is important to stress that, while it represents an important step forward, our approach still has several major limitations. A key weakness of the GBR model architecture is that amongst co-correlating variables, the largest main effect is consistently preferred for model building. As a result, co-correlating variables often have artificially low importance scores, both for impurity and SHAP-based methods, like we observed in the comparison to the more randomized Extra Trees models. This implies that chemicals identified as key drivers of a specific sensory feature by GBR might not be the true causative compounds, but rather co-correlate with the actual causative chemical. For example, the high importance of ethyl acetate could be (partially) attributed to the total ester content, ethanol or ethyl hexanoate (rho=0.77, rho=0.72 and rho=0.68), while ethyl phenylacetate could hide the importance of prenyl isobutyrate and ethyl benzoate (rho=0.77 and rho=0.76). Expanding our GBR model to include beer style as a parameter did not yield additional power or insight. This is likely due to style-specific chemical signatures, such as iso-alpha acids and lactic acid, that implicitly convey style information to the original model, as well as the smaller sample size per style, limiting the power to uncover style-specific patterns. This can be partly attributed to the curse of dimensionality, where the high number of parameters results in the models mainly incorporating single parameter effects, rather than complex interactions such as style-dependent effects 67 . A larger number of samples may overcome some of these limitations and offer more insight into style-specific effects. On the other hand, beer style is not a rigid scientific classification, and beers within one style often differ a lot, which further complicates the analysis of style as a model factor.

Our study is limited to beers from Belgian breweries. Although these beers cover a large portion of the beer styles available globally, some beer styles and consumer patterns may be missing, while other features might be overrepresented. For example, many Belgian ales exhibit yeast-driven flavor profiles, which is reflected in the chemical drivers of appreciation discovered by this study. In future work, expanding the scope to include diverse markets and beer styles could lead to the identification of even more drivers of appreciation and better models for special niche products that were not present in our beer set.

In addition to inherent limitations of GBR models, there are also some limitations associated with studying food aroma. Even if our chemical analyses measured most of the known aroma compounds, the total number of flavor compounds in complex foods like beer is still larger than the subset we were able to measure in this study. For example, hop-derived thiols, that influence flavor at very low concentrations, are notoriously difficult to measure in a high-throughput experiment. Moreover, consumer perception remains subjective and prone to biases that are difficult to avoid. It is also important to stress that the models are still immature and that more extensive datasets will be crucial for developing more complete models in the future. Besides more samples and parameters, our dataset does not include any demographic information about the tasters. Including such data could lead to better models that grasp external factors like age and culture. Another limitation is that our set of beers consists of high-quality end-products and lacks beers that are unfit for sale, which limits the current model in accurately predicting products that are appreciated very badly. Finally, while models could be readily applied in quality control, their use in sensory science and product development is restrained by their inability to discern causal relationships. Given that the models cannot distinguish compounds that genuinely drive consumer perception from those that merely correlate, validation experiments are essential to identify true causative compounds.

Despite the inherent limitations, dissection of our models enabled us to pinpoint specific molecules as potential drivers of beer aroma and consumer appreciation, including compounds that were unexpected and would not have been identified using standard approaches. Important drivers of beer appreciation uncovered by our models include protein levels, ethyl acetate, ethyl phenyl acetate and lactic acid. Currently, many brewers already use lactic acid to acidify their brewing water and ensure optimal pH for enzymatic activity during the mashing process. Our results suggest that adding lactic acid can also improve beer appreciation, although its individual effect remains to be tested. Interestingly, ethanol appears to be unnecessary to improve beer appreciation, both for blond beer and alcohol-free beer. Given the growing consumer interest in alcohol-free beer, with a predicted annual market growth of >7% 84 , it is relevant for brewers to know what compounds can further increase consumer appreciation of these beers. Hence, our model may readily provide avenues to further improve the flavor and consumer appreciation of both alcoholic and non-alcoholic beers, which is generally considered one of the key challenges for future beer production.

Whereas we see a direct implementation of our results for the development of superior alcohol-free beverages and other food products, our study can also serve as a stepping stone for the development of novel alcohol-containing beverages. We want to echo the growing body of scientific evidence for the negative effects of alcohol consumption, both on the individual level by the mutagenic, teratogenic and carcinogenic effects of ethanol 85 , 86 , as well as the burden on society caused by alcohol abuse and addiction. We encourage the use of our results for the production of healthier, tastier products, including novel and improved beverages with lower alcohol contents. Furthermore, we strongly discourage the use of these technologies to improve the appreciation or addictive properties of harmful substances.

The present work demonstrates that despite some important remaining hurdles, combining the latest developments in chemical analyses, sensory analysis and modern machine learning methods offers exciting avenues for food chemistry and engineering. Soon, these tools may provide solutions in quality control and recipe development, as well as new approaches to sensory science and flavor research.

Beer selection

250 commercial Belgian beers were selected to cover the broad diversity of beer styles and corresponding diversity in chemical composition and aroma. See Supplementary Fig.  S1 .

Chemical dataset

Sample preparation.

Beers within their expiration date were purchased from commercial retailers. Samples were prepared in biological duplicates at room temperature, unless explicitly stated otherwise. Bottle pressure was measured with a manual pressure device (Steinfurth Mess-Systeme GmbH) and used to calculate CO 2 concentration. The beer was poured through two filter papers (Macherey-Nagel, 500713032 MN 713 ¼) to remove carbon dioxide and prevent spontaneous foaming. Samples were then prepared for measurements by targeted Headspace-Gas Chromatography-Flame Ionization Detector/Flame Photometric Detector (HS-GC-FID/FPD), Headspace-Solid Phase Microextraction-Gas Chromatography-Mass Spectrometry (HS-SPME-GC-MS), colorimetric analysis, enzymatic analysis, Near-Infrared (NIR) analysis, as described in the sections below. The mean values of biological duplicates are reported for each compound.

HS-GC-FID/FPD

HS-GC-FID/FPD (Shimadzu GC 2010 Plus) was used to measure higher alcohols, acetaldehyde, esters, 4-vinyl guaicol, and sulfur compounds. Each measurement comprised 5 ml of sample pipetted into a 20 ml glass vial containing 1.75 g NaCl (VWR, 27810.295). 100 µl of 2-heptanol (Sigma-Aldrich, H3003) (internal standard) solution in ethanol (Fisher Chemical, E/0650DF/C17) was added for a final concentration of 2.44 mg/L. Samples were flushed with nitrogen for 10 s, sealed with a silicone septum, stored at −80 °C and analyzed in batches of 20.

The GC was equipped with a DB-WAXetr column (length, 30 m; internal diameter, 0.32 mm; layer thickness, 0.50 µm; Agilent Technologies, Santa Clara, CA, USA) to the FID and an HP-5 column (length, 30 m; internal diameter, 0.25 mm; layer thickness, 0.25 µm; Agilent Technologies, Santa Clara, CA, USA) to the FPD. N 2 was used as the carrier gas. Samples were incubated for 20 min at 70 °C in the headspace autosampler (Flow rate, 35 cm/s; Injection volume, 1000 µL; Injection mode, split; Combi PAL autosampler, CTC analytics, Switzerland). The injector, FID and FPD temperatures were kept at 250 °C. The GC oven temperature was first held at 50 °C for 5 min and then allowed to rise to 80 °C at a rate of 5 °C/min, followed by a second ramp of 4 °C/min until 200 °C kept for 3 min and a final ramp of (4 °C/min) until 230 °C for 1 min. Results were analyzed with the GCSolution software version 2.4 (Shimadzu, Kyoto, Japan). The GC was calibrated with a 5% EtOH solution (VWR International) containing the volatiles under study (Supplementary Table  S7 ).

HS-SPME-GC-MS

HS-SPME-GC-MS (Shimadzu GCMS-QP-2010 Ultra) was used to measure additional volatile compounds, mainly comprising terpenoids and esters. Samples were analyzed by HS-SPME using a triphase DVB/Carboxen/PDMS 50/30 μm SPME fiber (Supelco Co., Bellefonte, PA, USA) followed by gas chromatography (Thermo Fisher Scientific Trace 1300 series, USA) coupled to a mass spectrometer (Thermo Fisher Scientific ISQ series MS) equipped with a TriPlus RSH autosampler. 5 ml of degassed beer sample was placed in 20 ml vials containing 1.75 g NaCl (VWR, 27810.295). 5 µl internal standard mix was added, containing 2-heptanol (1 g/L) (Sigma-Aldrich, H3003), 4-fluorobenzaldehyde (1 g/L) (Sigma-Aldrich, 128376), 2,3-hexanedione (1 g/L) (Sigma-Aldrich, 144169) and guaiacol (1 g/L) (Sigma-Aldrich, W253200) in ethanol (Fisher Chemical, E/0650DF/C17). Each sample was incubated at 60 °C in the autosampler oven with constant agitation. After 5 min equilibration, the SPME fiber was exposed to the sample headspace for 30 min. The compounds trapped on the fiber were thermally desorbed in the injection port of the chromatograph by heating the fiber for 15 min at 270 °C.

The GC-MS was equipped with a low polarity RXi-5Sil MS column (length, 20 m; internal diameter, 0.18 mm; layer thickness, 0.18 µm; Restek, Bellefonte, PA, USA). Injection was performed in splitless mode at 320 °C, a split flow of 9 ml/min, a purge flow of 5 ml/min and an open valve time of 3 min. To obtain a pulsed injection, a programmed gas flow was used whereby the helium gas flow was set at 2.7 mL/min for 0.1 min, followed by a decrease in flow of 20 ml/min to the normal 0.9 mL/min. The temperature was first held at 30 °C for 3 min and then allowed to rise to 80 °C at a rate of 7 °C/min, followed by a second ramp of 2 °C/min till 125 °C and a final ramp of 8 °C/min with a final temperature of 270 °C.

Mass acquisition range was 33 to 550 amu at a scan rate of 5 scans/s. Electron impact ionization energy was 70 eV. The interface and ion source were kept at 275 °C and 250 °C, respectively. A mix of linear n-alkanes (from C7 to C40, Supelco Co.) was injected into the GC-MS under identical conditions to serve as external retention index markers. Identification and quantification of the compounds were performed using an in-house developed R script as described in Goelen et al. and Reher et al. 87 , 88 (for package information, see Supplementary Table  S8 ). Briefly, chromatograms were analyzed using AMDIS (v2.71) 89 to separate overlapping peaks and obtain pure compound spectra. The NIST MS Search software (v2.0 g) in combination with the NIST2017, FFNSC3 and Adams4 libraries were used to manually identify the empirical spectra, taking into account the expected retention time. After background subtraction and correcting for retention time shifts between samples run on different days based on alkane ladders, compound elution profiles were extracted and integrated using a file with 284 target compounds of interest, which were either recovered in our identified AMDIS list of spectra or were known to occur in beer. Compound elution profiles were estimated for every peak in every chromatogram over a time-restricted window using weighted non-negative least square analysis after which peak areas were integrated 87 , 88 . Batch effect correction was performed by normalizing against the most stable internal standard compound, 4-fluorobenzaldehyde. Out of all 284 target compounds that were analyzed, 167 were visually judged to have reliable elution profiles and were used for final analysis.

Discrete photometric and enzymatic analysis

Discrete photometric and enzymatic analysis (Thermo Scientific TM Gallery TM Plus Beermaster Discrete Analyzer) was used to measure acetic acid, ammonia, beta-glucan, iso-alpha acids, color, sugars, glycerol, iron, pH, protein, and sulfite. 2 ml of sample volume was used for the analyses. Information regarding the reagents and standard solutions used for analyses and calibrations is included in Supplementary Table  S7 and Supplementary Table  S9 .

NIR analyses

NIR analysis (Anton Paar Alcolyzer Beer ME System) was used to measure ethanol. Measurements comprised 50 ml of sample, and a 10% EtOH solution was used for calibration.

Correlation calculations

Pairwise Spearman Rank correlations were calculated between all chemical properties.

Sensory dataset

Trained panel.

Our trained tasting panel consisted of volunteers who gave prior verbal informed consent. All compounds used for the validation experiment were of food-grade quality. The tasting sessions were approved by the Social and Societal Ethics Committee of the KU Leuven (G-2022-5677-R2(MAR)). All online reviewers agreed to the Terms and Conditions of the RateBeer website.

Sensory analysis was performed according to the American Society of Brewing Chemists (ASBC) Sensory Analysis Methods 90 . 30 volunteers were screened through a series of triangle tests. The sixteen most sensitive and consistent tasters were retained as taste panel members. The resulting panel was diverse in age [22–42, mean: 29], sex [56% male] and nationality [7 different countries]. The panel developed a consensus vocabulary to describe beer aroma, taste and mouthfeel. Panelists were trained to identify and score 50 different attributes, using a 7-point scale to rate attributes’ intensity. The scoring sheet is included as Supplementary Data  3 . Sensory assessments took place between 10–12 a.m. The beers were served in black-colored glasses. Per session, between 5 and 12 beers of the same style were tasted at 12 °C to 16 °C. Two reference beers were added to each set and indicated as ‘Reference 1 & 2’, allowing panel members to calibrate their ratings. Not all panelists were present at every tasting. Scores were scaled by standard deviation and mean-centered per taster. Values are represented as z-scores and clustered by Euclidean distance. Pairwise Spearman correlations were calculated between taste and aroma sensory attributes. Panel consistency was evaluated by repeating samples on different sessions and performing ANOVA to identify differences, using the ‘stats’ package (v4.2.2) in R (for package information, see Supplementary Table  S8 ).

Online reviews from a public database

The ‘scrapy’ package in Python (v3.6) (for package information, see Supplementary Table  S8 ). was used to collect 232,288 online reviews (mean=922, min=6, max=5343) from RateBeer, an online beer review database. Each review entry comprised 5 numerical scores (appearance, aroma, taste, palate and overall quality) and an optional review text. The total number of reviews per reviewer was collected separately. Numerical scores were scaled and centered per rater, and mean scores were calculated per beer.

For the review texts, the language was estimated using the packages ‘langdetect’ and ‘langid’ in Python. Reviews that were classified as English by both packages were kept. Reviewers with fewer than 100 entries overall were discarded. 181,025 reviews from >6000 reviewers from >40 countries remained. Text processing was done using the ‘nltk’ package in Python. Texts were corrected for slang and misspellings; proper nouns and rare words that are relevant to the beer context were specified and kept as-is (‘Chimay’,’Lambic’, etc.). A dictionary of semantically similar sensorial terms, for example ‘floral’ and ‘flower’, was created and collapsed together into one term. Words were stemmed and lemmatized to avoid identifying words such as ‘acid’ and ‘acidity’ as separate terms. Numbers and punctuation were removed.

Sentences from up to 50 randomly chosen reviews per beer were manually categorized according to the aspect of beer they describe (appearance, aroma, taste, palate, overall quality—not to be confused with the 5 numerical scores described above) or flagged as irrelevant if they contained no useful information. If a beer contained fewer than 50 reviews, all reviews were manually classified. This labeled data set was used to train a model that classified the rest of the sentences for all beers 91 . Sentences describing taste and aroma were extracted, and term frequency–inverse document frequency (TFIDF) was implemented to calculate enrichment scores for sensorial words per beer.

The sex of the tasting subject was not considered when building our sensory database. Instead, results from different panelists were averaged, both for our trained panel (56% male, 44% female) and the RateBeer reviews (70% male, 30% female for RateBeer as a whole).

Beer price collection and processing

Beer prices were collected from the following stores: Colruyt, Delhaize, Total Wine, BeerHawk, The Belgian Beer Shop, The Belgian Shop, and Beer of Belgium. Where applicable, prices were converted to Euros and normalized per liter. Spearman correlations were calculated between these prices and mean overall appreciation scores from RateBeer and the taste panel, respectively.

Pairwise Spearman Rank correlations were calculated between all sensory properties.

Machine learning models

Predictive modeling of sensory profiles from chemical data.

Regression models were constructed to predict (a) trained panel scores for beer flavors and quality from beer chemical profiles and (b) public reviews’ appreciation scores from beer chemical profiles. Z-scores were used to represent sensory attributes in both data sets. Chemical properties with log-normal distributions (Shapiro-Wilk test, p  <  0.05 ) were log-transformed. Missing chemical measurements (0.1% of all data) were replaced with mean values per attribute. Observations from 250 beers were randomly separated into a training set (70%, 175 beers) and a test set (30%, 75 beers), stratified per beer style. Chemical measurements (p = 231) were normalized based on the training set average and standard deviation. In total, three linear regression-based models: linear regression with first-order interaction terms (LR), lasso regression with first-order interaction terms (Lasso) and partial least squares regression (PLSR); five decision tree models, Adaboost regressor (ABR), Extra Trees (ET), Gradient Boosting regressor (GBR), Random Forest (RF) and XGBoost regressor (XGBR); one support vector machine model (SVR) and one artificial neural network model (ANN) were trained. The models were implemented using the ‘scikit-learn’ package (v1.2.2) and ‘xgboost’ package (v1.7.3) in Python (v3.9.16). Models were trained, and hyperparameters optimized, using five-fold cross-validated grid search with the coefficient of determination (R 2 ) as the evaluation metric. The ANN (scikit-learn’s MLPRegressor) was optimized using Bayesian Tree-Structured Parzen Estimator optimization with the ‘Optuna’ Python package (v3.2.0). Individual models were trained per attribute, and a multi-output model was trained on all attributes simultaneously.

Model dissection

GBR was found to outperform other methods, resulting in models with the highest average R 2 values in both trained panel and public review data sets. Impurity-based rankings of the most important predictors for each predicted sensorial trait were obtained using the ‘scikit-learn’ package. To observe the relationships between these chemical properties and their predicted targets, partial dependence plots (PDP) were constructed for the six most important predictors of consumer appreciation 74 , 75 .

The ‘SHAP’ package in Python (v0.41.0) was implemented to provide an alternative ranking of predictor importance and to visualize the predictors’ effects as a function of their concentration 68 .

Validation of causal chemical properties

To validate the effects of the most important model features on predicted sensory attributes, beers were spiked with the chemical compounds identified by the models and descriptive sensory analyses were carried out according to the American Society of Brewing Chemists (ASBC) protocol 90 .

Compound spiking was done 30 min before tasting. Compounds were spiked into fresh beer bottles, that were immediately resealed and inverted three times. Fresh bottles of beer were opened for the same duration, resealed, and inverted thrice, to serve as controls. Pairs of spiked samples and controls were served simultaneously, chilled and in dark glasses as outlined in the Trained panel section above. Tasters were instructed to select the glass with the higher flavor intensity for each attribute (directional difference test 92 ) and to select the glass they prefer.

The final concentration after spiking was equal to the within-style average, after normalizing by ethanol concentration. This was done to ensure balanced flavor profiles in the final spiked beer. The same methods were applied to improve a non-alcoholic beer. Compounds were the following: ethyl acetate (Merck KGaA, W241415), ethyl hexanoate (Merck KGaA, W243906), isoamyl acetate (Merck KGaA, W205508), phenethyl acetate (Merck KGaA, W285706), ethanol (96%, Colruyt), glycerol (Merck KGaA, W252506), lactic acid (Merck KGaA, 261106).

Significant differences in preference or perceived intensity were determined by performing the two-sided binomial test on each attribute.

Reporting summary

Further information on research design is available in the  Nature Portfolio Reporting Summary linked to this article.

Data availability

The data that support the findings of this work are available in the Supplementary Data files and have been deposited to Zenodo under accession code 10653704 93 . The RateBeer scores data are under restricted access, they are not publicly available as they are property of RateBeer (ZX Ventures, USA). Access can be obtained from the authors upon reasonable request and with permission of RateBeer (ZX Ventures, USA).  Source data are provided with this paper.

Code availability

The code for training the machine learning models, analyzing the models, and generating the figures has been deposited to Zenodo under accession code 10653704 93 .

Tieman, D. et al. A chemical genetic roadmap to improved tomato flavor. Science 355 , 391–394 (2017).

Article   ADS   CAS   PubMed   Google Scholar  

Plutowska, B. & Wardencki, W. Application of gas chromatography–olfactometry (GC–O) in analysis and quality assessment of alcoholic beverages – A review. Food Chem. 107 , 449–463 (2008).

Article   CAS   Google Scholar  

Legin, A., Rudnitskaya, A., Seleznev, B. & Vlasov, Y. Electronic tongue for quality assessment of ethanol, vodka and eau-de-vie. Anal. Chim. Acta 534 , 129–135 (2005).

Loutfi, A., Coradeschi, S., Mani, G. K., Shankar, P. & Rayappan, J. B. B. Electronic noses for food quality: A review. J. Food Eng. 144 , 103–111 (2015).

Ahn, Y.-Y., Ahnert, S. E., Bagrow, J. P. & Barabási, A.-L. Flavor network and the principles of food pairing. Sci. Rep. 1 , 196 (2011).

Article   CAS   PubMed   PubMed Central   Google Scholar  

Bartoshuk, L. M. & Klee, H. J. Better fruits and vegetables through sensory analysis. Curr. Biol. 23 , R374–R378 (2013).

Article   CAS   PubMed   Google Scholar  

Piggott, J. R. Design questions in sensory and consumer science. Food Qual. Prefer. 3293 , 217–220 (1995).

Article   Google Scholar  

Kermit, M. & Lengard, V. Assessing the performance of a sensory panel-panellist monitoring and tracking. J. Chemom. 19 , 154–161 (2005).

Cook, D. J., Hollowood, T. A., Linforth, R. S. T. & Taylor, A. J. Correlating instrumental measurements of texture and flavour release with human perception. Int. J. Food Sci. Technol. 40 , 631–641 (2005).

Chinchanachokchai, S., Thontirawong, P. & Chinchanachokchai, P. A tale of two recommender systems: The moderating role of consumer expertise on artificial intelligence based product recommendations. J. Retail. Consum. Serv. 61 , 1–12 (2021).

Ross, C. F. Sensory science at the human-machine interface. Trends Food Sci. Technol. 20 , 63–72 (2009).

Chambers, E. IV & Koppel, K. Associations of volatile compounds with sensory aroma and flavor: The complex nature of flavor. Molecules 18 , 4887–4905 (2013).

Pinu, F. R. Metabolomics—The new frontier in food safety and quality research. Food Res. Int. 72 , 80–81 (2015).

Danezis, G. P., Tsagkaris, A. S., Brusic, V. & Georgiou, C. A. Food authentication: state of the art and prospects. Curr. Opin. Food Sci. 10 , 22–31 (2016).

Shepherd, G. M. Smell images and the flavour system in the human brain. Nature 444 , 316–321 (2006).

Meilgaard, M. C. Prediction of flavor differences between beers from their chemical composition. J. Agric. Food Chem. 30 , 1009–1017 (1982).

Xu, L. et al. Widespread receptor-driven modulation in peripheral olfactory coding. Science 368 , eaaz5390 (2020).

Kupferschmidt, K. Following the flavor. Science 340 , 808–809 (2013).

Billesbølle, C. B. et al. Structural basis of odorant recognition by a human odorant receptor. Nature 615 , 742–749 (2023).

Article   ADS   PubMed   PubMed Central   Google Scholar  

Smith, B. Perspective: Complexities of flavour. Nature 486 , S6–S6 (2012).

Pfister, P. et al. Odorant receptor inhibition is fundamental to odor encoding. Curr. Biol. 30 , 2574–2587 (2020).

Moskowitz, H. W., Kumaraiah, V., Sharma, K. N., Jacobs, H. L. & Sharma, S. D. Cross-cultural differences in simple taste preferences. Science 190 , 1217–1218 (1975).

Eriksson, N. et al. A genetic variant near olfactory receptor genes influences cilantro preference. Flavour 1 , 22 (2012).

Ferdenzi, C. et al. Variability of affective responses to odors: Culture, gender, and olfactory knowledge. Chem. Senses 38 , 175–186 (2013).

Article   PubMed   Google Scholar  

Lawless, H. T. & Heymann, H. Sensory evaluation of food: Principles and practices. (Springer, New York, NY). https://doi.org/10.1007/978-1-4419-6488-5 (2010).

Colantonio, V. et al. Metabolomic selection for enhanced fruit flavor. Proc. Natl. Acad. Sci. 119 , e2115865119 (2022).

Fritz, F., Preissner, R. & Banerjee, P. VirtualTaste: a web server for the prediction of organoleptic properties of chemical compounds. Nucleic Acids Res 49 , W679–W684 (2021).

Tuwani, R., Wadhwa, S. & Bagler, G. BitterSweet: Building machine learning models for predicting the bitter and sweet taste of small molecules. Sci. Rep. 9 , 1–13 (2019).

Dagan-Wiener, A. et al. Bitter or not? BitterPredict, a tool for predicting taste from chemical structure. Sci. Rep. 7 , 1–13 (2017).

Pallante, L. et al. Toward a general and interpretable umami taste predictor using a multi-objective machine learning approach. Sci. Rep. 12 , 1–11 (2022).

Malavolta, M. et al. A survey on computational taste predictors. Eur. Food Res. Technol. 248 , 2215–2235 (2022).

Lee, B. K. et al. A principal odor map unifies diverse tasks in olfactory perception. Science 381 , 999–1006 (2023).

Mayhew, E. J. et al. Transport features predict if a molecule is odorous. Proc. Natl. Acad. Sci. 119 , e2116576119 (2022).

Niu, Y. et al. Sensory evaluation of the synergism among ester odorants in light aroma-type liquor by odor threshold, aroma intensity and flash GC electronic nose. Food Res. Int. 113 , 102–114 (2018).

Yu, P., Low, M. Y. & Zhou, W. Design of experiments and regression modelling in food flavour and sensory analysis: A review. Trends Food Sci. Technol. 71 , 202–215 (2018).

Oladokun, O. et al. The impact of hop bitter acid and polyphenol profiles on the perceived bitterness of beer. Food Chem. 205 , 212–220 (2016).

Linforth, R., Cabannes, M., Hewson, L., Yang, N. & Taylor, A. Effect of fat content on flavor delivery during consumption: An in vivo model. J. Agric. Food Chem. 58 , 6905–6911 (2010).

Guo, S., Na Jom, K. & Ge, Y. Influence of roasting condition on flavor profile of sunflower seeds: A flavoromics approach. Sci. Rep. 9 , 11295 (2019).

Ren, Q. et al. The changes of microbial community and flavor compound in the fermentation process of Chinese rice wine using Fagopyrum tataricum grain as feedstock. Sci. Rep. 9 , 3365 (2019).

Hastie, T., Friedman, J. & Tibshirani, R. The Elements of Statistical Learning. (Springer, New York, NY). https://doi.org/10.1007/978-0-387-21606-5 (2001).

Dietz, C., Cook, D., Huismann, M., Wilson, C. & Ford, R. The multisensory perception of hop essential oil: a review. J. Inst. Brew. 126 , 320–342 (2020).

CAS   Google Scholar  

Roncoroni, Miguel & Verstrepen, Kevin Joan. Belgian Beer: Tested and Tasted. (Lannoo, 2018).

Meilgaard, M. Flavor chemistry of beer: Part II: Flavor and threshold of 239 aroma volatiles. in (1975).

Bokulich, N. A. & Bamforth, C. W. The microbiology of malting and brewing. Microbiol. Mol. Biol. Rev. MMBR 77 , 157–172 (2013).

Dzialo, M. C., Park, R., Steensels, J., Lievens, B. & Verstrepen, K. J. Physiology, ecology and industrial applications of aroma formation in yeast. FEMS Microbiol. Rev. 41 , S95–S128 (2017).

Article   PubMed   PubMed Central   Google Scholar  

Datta, A. et al. Computer-aided food engineering. Nat. Food 3 , 894–904 (2022).

American Society of Brewing Chemists. Beer Methods. (American Society of Brewing Chemists, St. Paul, MN, U.S.A.).

Olaniran, A. O., Hiralal, L., Mokoena, M. P. & Pillay, B. Flavour-active volatile compounds in beer: production, regulation and control. J. Inst. Brew. 123 , 13–23 (2017).

Verstrepen, K. J. et al. Flavor-active esters: Adding fruitiness to beer. J. Biosci. Bioeng. 96 , 110–118 (2003).

Meilgaard, M. C. Flavour chemistry of beer. part I: flavour interaction between principal volatiles. Master Brew. Assoc. Am. Tech. Q 12 , 107–117 (1975).

Briggs, D. E., Boulton, C. A., Brookes, P. A. & Stevens, R. Brewing 227–254. (Woodhead Publishing). https://doi.org/10.1533/9781855739062.227 (2004).

Bossaert, S., Crauwels, S., De Rouck, G. & Lievens, B. The power of sour - A review: Old traditions, new opportunities. BrewingScience 72 , 78–88 (2019).

Google Scholar  

Verstrepen, K. J. et al. Flavor active esters: Adding fruitiness to beer. J. Biosci. Bioeng. 96 , 110–118 (2003).

Snauwaert, I. et al. Microbial diversity and metabolite composition of Belgian red-brown acidic ales. Int. J. Food Microbiol. 221 , 1–11 (2016).

Spitaels, F. et al. The microbial diversity of traditional spontaneously fermented lambic beer. PLoS ONE 9 , e95384 (2014).

Blanco, C. A., Andrés-Iglesias, C. & Montero, O. Low-alcohol Beers: Flavor Compounds, Defects, and Improvement Strategies. Crit. Rev. Food Sci. Nutr. 56 , 1379–1388 (2016).

Jackowski, M. & Trusek, A. Non-Alcohol. beer Prod. – Overv. 20 , 32–38 (2018).

Takoi, K. et al. The contribution of geraniol metabolism to the citrus flavour of beer: Synergy of geraniol and β-citronellol under coexistence with excess linalool. J. Inst. Brew. 116 , 251–260 (2010).

Kroeze, J. H. & Bartoshuk, L. M. Bitterness suppression as revealed by split-tongue taste stimulation in humans. Physiol. Behav. 35 , 779–783 (1985).

Mennella, J. A. et al. A spoonful of sugar helps the medicine go down”: Bitter masking bysucrose among children and adults. Chem. Senses 40 , 17–25 (2015).

Wietstock, P., Kunz, T., Perreira, F. & Methner, F.-J. Metal chelation behavior of hop acids in buffered model systems. BrewingScience 69 , 56–63 (2016).

Sancho, D., Blanco, C. A., Caballero, I. & Pascual, A. Free iron in pale, dark and alcohol-free commercial lager beers. J. Sci. Food Agric. 91 , 1142–1147 (2011).

Rodrigues, H. & Parr, W. V. Contribution of cross-cultural studies to understanding wine appreciation: A review. Food Res. Int. 115 , 251–258 (2019).

Korneva, E. & Blockeel, H. Towards better evaluation of multi-target regression models. in ECML PKDD 2020 Workshops (eds. Koprinska, I. et al.) 353–362 (Springer International Publishing, Cham, 2020). https://doi.org/10.1007/978-3-030-65965-3_23 .

Gastón Ares. Mathematical and Statistical Methods in Food Science and Technology. (Wiley, 2013).

Grinsztajn, L., Oyallon, E. & Varoquaux, G. Why do tree-based models still outperform deep learning on tabular data? Preprint at http://arxiv.org/abs/2207.08815 (2022).

Gries, S. T. Statistics for Linguistics with R: A Practical Introduction. in Statistics for Linguistics with R (De Gruyter Mouton, 2021). https://doi.org/10.1515/9783110718256 .

Lundberg, S. M. et al. From local explanations to global understanding with explainable AI for trees. Nat. Mach. Intell. 2 , 56–67 (2020).

Ickes, C. M. & Cadwallader, K. R. Effects of ethanol on flavor perception in alcoholic beverages. Chemosens. Percept. 10 , 119–134 (2017).

Kato, M. et al. Influence of high molecular weight polypeptides on the mouthfeel of commercial beer. J. Inst. Brew. 127 , 27–40 (2021).

Wauters, R. et al. Novel Saccharomyces cerevisiae variants slow down the accumulation of staling aldehydes and improve beer shelf-life. Food Chem. 398 , 1–11 (2023).

Li, H., Jia, S. & Zhang, W. Rapid determination of low-level sulfur compounds in beer by headspace gas chromatography with a pulsed flame photometric detector. J. Am. Soc. Brew. Chem. 66 , 188–191 (2008).

Dercksen, A., Laurens, J., Torline, P., Axcell, B. C. & Rohwer, E. Quantitative analysis of volatile sulfur compounds in beer using a membrane extraction interface. J. Am. Soc. Brew. Chem. 54 , 228–233 (1996).

Molnar, C. Interpretable Machine Learning: A Guide for Making Black-Box Models Interpretable. (2020).

Zhao, Q. & Hastie, T. Causal interpretations of black-box models. J. Bus. Econ. Stat. Publ. Am. Stat. Assoc. 39 , 272–281 (2019).

Article   MathSciNet   Google Scholar  

Hastie, T., Tibshirani, R. & Friedman, J. The Elements of Statistical Learning. (Springer, 2019).

Labrado, D. et al. Identification by NMR of key compounds present in beer distillates and residual phases after dealcoholization by vacuum distillation. J. Sci. Food Agric. 100 , 3971–3978 (2020).

Lusk, L. T., Kay, S. B., Porubcan, A. & Ryder, D. S. Key olfactory cues for beer oxidation. J. Am. Soc. Brew. Chem. 70 , 257–261 (2012).

Gonzalez Viejo, C., Torrico, D. D., Dunshea, F. R. & Fuentes, S. Development of artificial neural network models to assess beer acceptability based on sensory properties using a robotic pourer: A comparative model approach to achieve an artificial intelligence system. Beverages 5 , 33 (2019).

Gonzalez Viejo, C., Fuentes, S., Torrico, D. D., Godbole, A. & Dunshea, F. R. Chemical characterization of aromas in beer and their effect on consumers liking. Food Chem. 293 , 479–485 (2019).

Gilbert, J. L. et al. Identifying breeding priorities for blueberry flavor using biochemical, sensory, and genotype by environment analyses. PLOS ONE 10 , 1–21 (2015).

Goulet, C. et al. Role of an esterase in flavor volatile variation within the tomato clade. Proc. Natl. Acad. Sci. 109 , 19009–19014 (2012).

Article   ADS   CAS   PubMed   PubMed Central   Google Scholar  

Borisov, V. et al. Deep Neural Networks and Tabular Data: A Survey. IEEE Trans. Neural Netw. Learn. Syst. 1–21 https://doi.org/10.1109/TNNLS.2022.3229161 (2022).

Statista. Statista Consumer Market Outlook: Beer - Worldwide.

Seitz, H. K. & Stickel, F. Molecular mechanisms of alcoholmediated carcinogenesis. Nat. Rev. Cancer 7 , 599–612 (2007).

Voordeckers, K. et al. Ethanol exposure increases mutation rate through error-prone polymerases. Nat. Commun. 11 , 3664 (2020).

Goelen, T. et al. Bacterial phylogeny predicts volatile organic compound composition and olfactory response of an aphid parasitoid. Oikos 129 , 1415–1428 (2020).

Article   ADS   Google Scholar  

Reher, T. et al. Evaluation of hop (Humulus lupulus) as a repellent for the management of Drosophila suzukii. Crop Prot. 124 , 104839 (2019).

Stein, S. E. An integrated method for spectrum extraction and compound identification from gas chromatography/mass spectrometry data. J. Am. Soc. Mass Spectrom. 10 , 770–781 (1999).

American Society of Brewing Chemists. Sensory Analysis Methods. (American Society of Brewing Chemists, St. Paul, MN, U.S.A., 1992).

McAuley, J., Leskovec, J. & Jurafsky, D. Learning Attitudes and Attributes from Multi-Aspect Reviews. Preprint at https://doi.org/10.48550/arXiv.1210.3926 (2012).

Meilgaard, M. C., Carr, B. T. & Carr, B. T. Sensory Evaluation Techniques. (CRC Press, Boca Raton). https://doi.org/10.1201/b16452 (2014).

Schreurs, M. et al. Data from: Predicting and improving complex beer flavor through machine learning. Zenodo https://doi.org/10.5281/zenodo.10653704 (2024).

Download references

Acknowledgements

We thank all lab members for their discussions and thank all tasting panel members for their contributions. Special thanks go out to Dr. Karin Voordeckers for her tremendous help in proofreading and improving the manuscript. M.S. was supported by a Baillet-Latour fellowship, L.C. acknowledges financial support from KU Leuven (C16/17/006), F.A.T. was supported by a PhD fellowship from FWO (1S08821N). Research in the lab of K.J.V. is supported by KU Leuven, FWO, VIB, VLAIO and the Brewing Science Serves Health Fund. Research in the lab of T.W. is supported by FWO (G.0A51.15) and KU Leuven (C16/17/006).

Author information

These authors contributed equally: Michiel Schreurs, Supinya Piampongsant, Miguel Roncoroni.

Authors and Affiliations

VIB—KU Leuven Center for Microbiology, Gaston Geenslaan 1, B-3001, Leuven, Belgium

Michiel Schreurs, Supinya Piampongsant, Miguel Roncoroni, Lloyd Cool, Beatriz Herrera-Malaver, Florian A. Theßeling & Kevin J. Verstrepen

CMPG Laboratory of Genetics and Genomics, KU Leuven, Gaston Geenslaan 1, B-3001, Leuven, Belgium

Leuven Institute for Beer Research (LIBR), Gaston Geenslaan 1, B-3001, Leuven, Belgium

Laboratory of Socioecology and Social Evolution, KU Leuven, Naamsestraat 59, B-3000, Leuven, Belgium

Lloyd Cool, Christophe Vanderaa & Tom Wenseleers

VIB Bioinformatics Core, VIB, Rijvisschestraat 120, B-9052, Ghent, Belgium

Łukasz Kreft & Alexander Botzki

AB InBev SA/NV, Brouwerijplein 1, B-3000, Leuven, Belgium

Philippe Malcorps & Luk Daenen

You can also search for this author in PubMed   Google Scholar

Contributions

S.P., M.S. and K.J.V. conceived the experiments. S.P., M.S. and K.J.V. designed the experiments. S.P., M.S., M.R., B.H. and F.A.T. performed the experiments. S.P., M.S., L.C., C.V., L.K., A.B., P.M., L.D., T.W. and K.J.V. contributed analysis ideas. S.P., M.S., L.C., C.V., T.W. and K.J.V. analyzed the data. All authors contributed to writing the manuscript.

Corresponding author

Correspondence to Kevin J. Verstrepen .

Ethics declarations

Competing interests.

K.J.V. is affiliated with bar.on. The other authors declare no competing interests.

Peer review

Peer review information.

Nature Communications thanks Florian Bauer, Andrew John Macintosh and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary information, peer review file, description of additional supplementary files, supplementary data 1, supplementary data 2, supplementary data 3, supplementary data 4, supplementary data 5, supplementary data 6, supplementary data 7, reporting summary, source data, source data, rights and permissions.

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/ .

Reprints and permissions

About this article

Cite this article.

Schreurs, M., Piampongsant, S., Roncoroni, M. et al. Predicting and improving complex beer flavor through machine learning. Nat Commun 15 , 2368 (2024). https://doi.org/10.1038/s41467-024-46346-0

Download citation

Received : 30 October 2023

Accepted : 21 February 2024

Published : 26 March 2024

DOI : https://doi.org/10.1038/s41467-024-46346-0

Share this article

Anyone you share the following link with will be able to read this content:

Sorry, a shareable link is not currently available for this article.

Provided by the Springer Nature SharedIt content-sharing initiative

By submitting a comment you agree to abide by our Terms and Community Guidelines . If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Quick links

• Explore articles by subject
• Guide to authors
• Editorial policies

Sign up for the Nature Briefing: Translational Research newsletter — top stories in biotechnology, drug discovery and pharma.