Association of insomnia with diet quality and eating behavior

To the Editor: Insomnia affects 30%–50% of adults globally, leading to diminished quality of life and increased risks of developing physically and mentally demanding disorders.1 Previous research, primarily in Western populations, reported the association between insomnia and poor diet quality, characterized by higher energy intake, lower vegetable consumption, and a preference for fatty foods, risk factors contributing to major chronic conditions.2 This association, however, remains under-explored in Asian populations, where dietary patterns significantly differ. Additionally, research indicated that poor sleep quality was often associated with nocturnal behaviors like late-night eating and breakfast skipping.3,4 Consuming breakfast is generally associated with better diet quality, while night-eating tends to be related to poorer diet quality. Further, the specific impact of insomnia on such eating behaviors, and their role in potentially mediating the relationship between insomnia and diet quality, has not been thoroughly explored, particularly in community-based populations. We thus conducted a large-scale community-based study in China to evaluate whether participants with insomnia were more likely to have lower diet quality and unhealthy behaviors (i.e., frequently engaging in night-eating and breakfast-skipping). We further investigated whether these eating behaviors mediated the potential association between insomnia and diet quality. Our study used data from two ongoing cohorts in China, the Kailuan Study I and II in Tangshan City, China, which were described elsewhere.5 Both cohorts underwent biennial follow-ups with comprehensive health assessments. In 2014, 100,587 participants of the Kailuan I and II completed the survey that included assessments of dietary intake and sleep status. After excluding individuals with major diseases (i.e., diabetes, cardiovascular disease, and cancer) and those with implausible dietary data or incomplete sleep and diet information, our analysis included 50,218 participants Supplementary Table 1, Supplementary Figure 1, https://links.lww.com/CM9/B987. Insomnia was defined based on the National Institutes of Health (NIH) recommendations,6 incorporating non-restorative sleep plus one or more of the three nocturnal insomnia symptoms (difficulty initiating sleep, difficulty maintaining sleep, and early-morning awakenings). Dietary intake was assessed using a validated food frequency questionnaire (FFQ), and diet quality was evaluated based on a modified version of the Alternative Healthy Eating Index (mAHEI).5 The studied unhealthy eating behaviors, including habitual night-eating and breakfast-skipping. We investigated the association between insomnia, including specific insomnia symptoms, and diet quality, as well as eating behaviors. Our analysis adjusted for various demographic factors, lifestyle factors, health factors, energy intake, and sleep parameters (i.e., the use of hypnotics, frequent snoring, and sleep duration) using multivariable linear regression models and logistic regression models. We also conducted stratified and sensitivity analyses to account for factors like chronic disease and hypnotic usage. We further performed the mediation analysis using the R "mediation" package, a general algorithm for estimating causal mediation effects with a variety of statistical models, to test how the studied eating behaviors affected the relationship between insomnia and diet quality. All analyses were performed in R software (version 4.4.2, R Foundation, Vienna, Austria). The mean age of participants was 49.9 ± 14.6 years, and 82.70% (41,530/50,218) were men. In total, 3.54% (1780/50,218) of participants reported insomnia Supplementary Table 2, https://links.lww.com/CM9/B987. Participants with insomnia or individual insomnia symptoms had a lower mAHEI score than participants without insomnia or insomnia symptoms after adjusting for covariates (P-trend 3000 CNY), BMI (27.9 kg/m2), marital status (married, divorced/separate/single, or widowed), occupational type (blue collar or white collar), occupation status (retired or occupied), current smoker (yes or no), current drinker (yes or no), physical activity (IPAQ: low, moderate, or high), hypertension (yes or no), energy intake (kcal), use of hypnotics (yes or no), frequent snoring (yes or no), and sleep duration (9 h). BMI: Body mass index; CIs: Confidential intervals; IPAQ: International Physical Activity Questionnaire; mAHEI: Modified Alternative Healthy Eating Index; OR: Odds ratio.We observed significant associations between insomnia and the higher odds of having night-eating (adjusted odds ratio OR = 1.90; 95% confidence interval CI: 1.69–2.14) and breakfast-skipping (adjusted OR = 1.41; 95% CI: 1.24–1.62). A similar pattern was observed for each insomnia symptom Figure 1. Unhealthy eating behaviors (night-eating and breakfast-skipping) mediated the association between insomnia and diet quality, with a mean proportion of 6.60% for night-eating and 2.60% for breakfast-skipping Supplementary Figure 4, https://links.lww.com/CM9/B987. In this large-scale cross-sectional study, we observed that individuals with insomnia had lower diet quality and poorer eating behaviors. The unhealthy eating behaviors mediated a small but significant proportion of the association between insomnia and diet quality. Our results were broadly in line with previous studies on sleep disorders and overall diet quality, which were conducted in Western populations with relatively smaller sample sizes.7–9 In our previous study based on the Health Professional's follow-up study, we found a non-significant trend between insomnia and a low AHEI score.10 Contrary to some previous studies, we did not observe a significant association between insomnia and higher energy intake. This discrepancy may arise from differences in dietary assessment methods and population characteristics. Notably, certain demographics, particularly middle-aged people, men and blue-collar workers, appeared to possess more pronounced associations between insomnia and diet quality in comparison with the elderly, women, and white-collar workers. This could be attributed to lifestyle and health status variations among the different groups. In the current study, we found that participants with insomnia had a significantly lower score on five individual food items. This finding was consistent with previous studies that poor sleep quality may change the type of food intake, with lower fresh vegetables, fruits, and whole grain consumption, higher sodium and alcohol, and sugar-sweetened beverages consumption. Unexpectedly, we observed that individuals with insomnia tended to consume more nuts and seeds. This could be interpreted by the fact that insomnia may be associated with greater snack intake during nighttime, as nuts and sunflower seeds were the most common recreational snacks among the Chinese population. Consistent with previous studies, we observed a significant association between insomnia and the odds of night-eating and breakfast-skipping, suggesting that sleep disorders could lead to changes in eating behaviors. In the mediation analysis, we found that unhealthy eating behaviors (i.e., night-eating and breakfast-skipping) mediated a small but significant proportion of the association between insomnia and diet quality. It can be explained by the known adverse effect of eating snacks before bedtime, which may increase the consumption of unhealthy food items, such as trans-fat, sodium, and sugar-sweetened beverages. Furthermore, it reconfirms the hypothesis that insomnia may affect diet quality by increasing meal frequency. The underlying mechanisms for the observed associations might involve alterations in appetite-regulating hormones and taste sensitivity due to poor sleep quality, potentially leading to changes in food preferences and eating behaviors. However, these hypotheses need further exploration. Our study had several limitations. The cross-sectional design limited the causal inference, further prospective cohort studies are warranted to understand the temporal relationship. There was potential recall bias and misclassification bias due to the self-reported dietary assessments and sleep data. Generalizability is another concern because the study was conducted in a specific demographic in Northern China, predominantly involving men with lower-to-middle education levels and income. Furthermore, the clinical implication of our results needs careful consideration due to the cross-sectional study design and low mediated proportions of the eating behaviors on the association between insomnia and diet quality, indicating that the clinical value of insomnia and unhealthy eating behaviors needs to be explored with better dietary assessments. In conclusion, in a large Chinese cohort, we found that insomnia was associated with lower diet quality and unhealthy eating behaviors. Further prospective studies with an objective assessment of sleep and diet are warranted to clarify the temporal relationship between insomnia and diet quality or eating behaviors. Funding This study is supported by grants from the Startup Grant at Fudan University (Nos. JIF201036Y and JIF201040Y). Conflicts of interest None.