Over the past few decades, obesity has emerged as a severe health problem in the world since it can increase the incidence of ailments such as cardiovascular disease, diabetes, and cancer, especially among middle-aged and elderly age populations (1). The World Health Organization (WHO) characterizes obesity as the excessive accumulation of adipose tissue beyond normal physiological requirements (2). The Centers for Disease Control and Prevention’s Prevalence of Obesity in Adults 2023 map (3)revealed that middle-aged adults were nearly twice as likely to be obese compared to young adults across all states. Obesity rates were lowest among 18- to 24-year-olds at 19.5% and highest among 45- to 54-year-olds at 39.2%. Slow metabolism, decreased exercise, unbalanced diet and energy intake, as well as life pressure, have led to a greater risk of obesity for middle-aged and elderly people (4). Therefore, it is essential to take suitable actions to prevent and manage obesity. Obesity is characterized by a prolonged imbalance between caloric intake and energy expenditure, typically indicated by a BMI of 30 or higher (5). Adopting habits that decrease energy intake while boosting calorie expenditure might help slow down the advancement of obesity, including following a proper diet and engaging in physical activity (6). In addition, sleep may be another important factor (7).

Sufficient sleep is favorable for regulating the immune system and human metabolism (8). On the contrary, insufficient sleep may result in metabolic disturbances (9). Generally, 7–9 h is deemed as the suitable sleep time for overall physical and mental (10). However, over one-third of American adults get less than 7 h of sleep (11). A similar trend of shorter sleep duration has been observed in some Western countries over the past decades (7). Thus, a comprehensive evaluation of the correlation between sleep duration and obesity is crucial for exploring weight management (12, 13).

Numerous previous researches have confirmed the correlation between the duration of sleep and obesity (14–30). Nevertheless, these articles mainly found a negative relationship between short sleep time and obesity, and the results of long sleep time were unclear. In addition, the amount of sleep required varies at different ages. This study sought to elucidate further the relationship between sleep duration and various anthropometric indices of obesity in middle-aged and elderly Americans, ≥45 years old, using the data from the National Health and Nutrition Examination Surveys (NHANES) 2005–2014. Currently, the indices typically employed for clinical evaluation of obesity include waist circumference (WC), waist-to-hip ratio (WHR), waist-to-height ratio (WHtR), and body mass index (BMI) (31–33). Due to some limitations of the National Health and Nutrition Examination Survey (NHANES) surveys analyzed in this study (missing data on the hip circumference in the cited surveys), we selected BMI, WC, and WHtR as measures of obesity in this paper.

The purpose of this research was to assess the correlation between sleep duration and three obesity indicators among middle-aged and elderly individuals in the United States. Our cross-sectional study included 14,383 middle-aged and elderly Americans from NHANES. After multiple data analyses, it was concluded that sleep duration was significantly negatively correlated with BMI, WC and WHtR in this population.

Examining three obesity indicators and sleep duration enhanced our conclusions and their applicability. BMI is commonly used for assessing body measurements due to its simplicity, but it cannot differentiate between fat and muscle. WC is proposed as a complementary metric that is particularly effective for detecting central obesity. However, excluding height in the WC measurement may result in inaccuracies when estimating abdominal obesity across individuals of varying statures, thereby underscoring its limitations (5). Thus, we included WHtR analysis. In the fully adjusted model, a one-hour increase in sleep time was associated with a 0.19 kg/m² decrease in BMI (β = −0.19, 95%CI: −0.26, −0.12). After full adjustment, each additional hour of sleep was associated with a 0.36 cm decrease in WC (β = −0.36, 95%CI: −0.53, −0.19). After full adjustment, a one-hour increase in sleep duration was associated with a 0.27% decrease in WHtR expressed as a percentage (β = −0.27, 95%CI: −0.38, −0.17). The subgroup analysis results suggested that statistically significant differences were observed regarding race, diabetes, or hypertension status. Thus, the above findings apply more to middle-aged and elderly Non-Hispanic White in the U.S. However, we found that the number of participants in specific races was much lower than that in other subgroups, which may decrease the power of significant effect detection.

Previous studies, including meta-analyses and systematic reviews, have consistently shown a correlation between sleep duration and obesity indicators (14–27, 39–41). According to the Korean Health and Nutrition Survey study conducted by Kim et al. (28), those who slept less than 5 h per day had an increased likelihood of obesity, as determined by their BMI. In a large number of wearable sensor participants, the higher the body mass index, the shorter the sleep duration (29). Another study by Yazdanpanah et al. (41) found that BMI, WC and WHtR were in a significant negative association with nighttime sleep in an Iranian population. Furthermore, a meta-analysis conducted by Bacaro et al. (30) discovered that adults with a shorter sleep duration were more prone to a heightened susceptibility to obesity, but results on long sleep were still inconclusive. These studies reveal a negative correlation between sleep duration and obesity indicators in the case of short sleep duration, although the effects of long sleep remain uncertain. Our results are consistent with theirs on short sleep conditions. At the same time, we found that in the case of appropriate and long sleep, sleep duration and BMI, WC still showed a negative correlation, which is a new finding. The shape of the curve suggested that the critical value of sleeping is 7 h (even though this cut-off value was not significant) and that sleeping less than 7 h can be more associated with worse anthropometric indicators. In a comparable cross-sectional study, the associations between sleep duration and obesity indicators were more pronounced in young adults than in older adults (40), so studies tend to target younger groups. Our article, which focuses on the middle-aged and elderly population in the United States, strengthens the research analysis on older adults.

Multiple mechanisms elucidate the associations between sleep duration and obesity indicators (42). Several researches have shown that during sleep the secretion of various hormones varies, contributing to the metabolism and energy balance of our body (43–47). Cortisol secretion has a pronounced circadian rhythm‌, reaching a trough at midnight (48). However, sleep deprivation disrupts the hypothalamic–pituitary–adrenal (HPA) axis and the circadian rhythm of cortisol, increasing cortisol levels and promoting increased fat storage (43–45). Thyroid hormones regulate basal metabolic rate and energy expenditure (49). Short sleep duration disrupts thyroid function, reducing its levels and slowing metabolism, leading to an energy surplus and weight gain (47). Additionally, short sleep duration decreases leptin levels and increases ghrelin, enhancing appetite (46). Additionally, people with short sleep duration may increase eating possibilities because of the longer period of wakefulness (50, 51). Some studies have found no significant relationship between sleep duration and energy consumption (47, 52, 53). However, other studies have shown that shorter sleep times increase their energy expenditure, but at the same time, their energy intake may be higher (54–57). Many other researchers have also investigated the link between long sleep and obesity, but there are limited mechanisms to explain this relationship (58).

Our study has several strengths. First, an extensive dataset obtained from NHANES produced a representative sample of middle-aged and elderly Americans, to which we can generalise our findings to the broader U.S. population. Second, the confounding covariates were adjusted to make the results more reliable. Third, more detailed statistical analyses were employed, including multiple logistic regression models with corresponding adjustment ranges, smooth curve fitting, and subgroup analysis, to fully assess the correlation between sleep duration and obesity indicators. However, our research still has some shortcomings. Due to the limitations of cross-sectional studies, there is a correlation between the duration of sleep and obesity indicators without causation. In addition, most of the published studies, including ours, relied on self-reported sleep measures (i.e., memory and subjective estimations, which can be imprecise and inaccurate). Subsequent studies could consider introducing other sleep indicators, such as the Pittsburgh Sleep Quality Index, to evaluate sleep quality better (59). Finally, the use of weighting in regression analysis made the conclusions more relevant to the US population but less relevant to the general population.

Sleep duration was significantly negatively associated with BMI and WC, suggesting that longer sleep duration may contribute to lower obesity indicators in middle-aged and elderly Americans. Subgroup analysis showed that their negative correlation differed between races or hypertension status. However, additional prospective studies are required to validate these findings and investigate potential causal relationships. These findings may facilitate exploring novel weight loss strategies tailored to more specific demographic groups.