Circadian Alignment, Cardiometabolic Disease, and Sex-Specific Differences in Adults With Overweight/Obesity

Abstract

Rates of obesity have reached alarming levels, with over 70% of Americans currently classified as having an overweight or obese body composition, posing significant health concerns from resulting comorbidity and mortality (1, 2). Visceral adiposity specifically is a key contributor to obesity-mediated metabolic outcomes, increasing the risk of metabolic syndrome, diabetes, and cardiovascular disease (3). Conventional approaches to counter the rise in obesity have largely focused on lifestyle modifications such as dietary changes and increased physical activity. Yet, despite increased recognition of the adverse health consequences associated with overweight/obesity, the obesity epidemic continues worldwide (2).

Sleep and the misalignment of behaviors with the internal circadian clock (ie, circadian misalignment) have been identified as novel contributors to obesity risk and overall cardiometabolic health (4). For instance, overnight/rotating shiftwork, an extreme form of circadian misalignment, is associated with higher risk of obesity, diabetes, and metabolic syndrome (5-7). Moreover, in nonshift-working males with healthy weight, the association between circadian alignment with body fat percentage and abdominal fat distribution is inversely correlated (8), implying a potential mechanistic link between the circadian system and obesity. Interestingly, relationships between circadian alignment and body composition are not observed in females with healthy weight (8), suggesting that sexual dimorphism in circadian mechanisms may contribute to differences in health outcomes between the sexes.

Though considerable sex-specific differences in metabolic responses during circadian alignment/misalignment have been observed in tightly controlled laboratory studies (9), there is a lack of understanding into whether sex may independently play a role in the relationship between circadian alignment with cardiometabolic health, particularly in individuals already with overweight/obesity who may be more prone to circadian-mediated metabolic outcomes (10). While both females and males are susceptible to having obesity, widely recognized sex-specific differences in body fat distribution and metabolic regulation have significant implications for cardiometabolic disease (11, 12). Given the rise in precision medicine and the growing awareness of inequities in the prevention, diagnosis, and treatment of disease between females and males (13, 14), it is imperative to identify underlying circadian and biological mechanisms that contribute to sex-specific variations in health outcomes. Therefore, the purpose of this investigation was to identify sex-specific associations between circadian alignment and cardiometabolic health markers in females and males with overweight/obesity in real-world settings. It was hypothesized these relationships would differ in a sex-dependent manner.

Materials and Methods

Procedures were approved by the Oregon Health & Science University Institutional Review Board. This secondary, cross-sectional examination included a subset of participant data in newly hired nurses at Oregon Health & Science University and bus operators at a local transit organization from previous publications (15, 16); investigating sex-dependent effects of circadian alignment on cardiometabolic health is unique to the present study. Participants' ages ranged between 23 and 54 years and exclusion criteria included self-reported hypertension, diabetes, sleep disorders, pregnancy, and shiftwork within the previous 3 months. Written informed consent was obtained from all participants.

The circadian phase was determined in 30 participants (15 females) during an evening dim-light melatonin onset (DLMO) assessment at the Oregon Clinical and Translational Research Institute. Ambient light was dimmed to <5 lux and saliva samples were collected half-hourly from 17:00 to 23:00. Participants refrained from using personal light-emitting electronic devices and, in the 10 minutes prior to sample collection, did not eat or drink and maintained a seated posture. Dual-energy X-ray absorptiometry (Lunar iDXA, GE Healthcare) was used to assess body fat percentage and android/gynoid percent fat ratio. Waist circumference was measured around the abdomen at the level of the iliac crest. After the participant was seated for 5 minutes, 3 resting blood pressures measurements were taken 1-minute apart and averaged. A fasted (≥3 hours) blood draw was taken to assess total cholesterol, high-density lipoprotein and low-density lipoprotein cholesterol, triglycerides, and glucose. Self-reported sleep and wake times were collected from daily surveys for 1 week after the in-laboratory assessment.

DLMO was quantified as the linear interpolated timepoint at which melatonin levels crossed and remained above a 3 pg/mL threshold (17). The degree of circadian alignment was quantified as the time difference between DLMO and the average of diary-determined sleep onset (termed “phase angle of entrainment”), as described previously (8); DLMO occurred before sleep onset in all participants in the current study. Next, participants were categorized into narrow (shorter time between DLMO and sleep onset) or wide phase angle (more time between DLMO and sleep onset) groups based on whether their phase angle was below or above the group's median phase angle, respectively. Because of the known differences in metabolic markers between sexes (11, 12), participants were further subdivided based on sex into narrow (7 females/8 males) and wide (8 females/7 males) phase angle groups. Finally, sex-specific metabolic syndrome risk scores were calculated using 5 component variables of metabolic syndrome, as described previously (18).

All statistical analyses were conducted using R. Independent samples t-tests were used to assess differences in demographics between sex and sex differences in health outcomes between the narrow and wide phase angle groups. To determine sex-specific differences in health outcomes, the sample was stratified by sex and cardiometabolic variables within narrow and wide phase angle groups and compared using independent samples t-tests. Effect sizes were calculated using Hedge's g (small ≥0.2; moderate ≥0.5; large ≥0.8) between narrow and wide phase angle groups within each sex. Potential dose-dependent relationships between phase angle and cardiometabolic health metrics were assessed with Pearson correlations. Data are presented as mean ± SD. Statistical significance was set at an α-level of P < .05.

Results

There were no differences between females and males for age, body mass index (BMI), DLMO timing, or phase angle (all P > .20; Table 1). Female participants had significantly higher body fat percentage and high-density lipoprotein levels, while male participants had higher android/gynoid percent body fat, triglycerides levels, and metabolic syndrome risk scores (Table 1).

There were no sex differences in average phase angle within the narrow (P = .69) or wide (P = .11) groups (Table 1). Females in the narrow phase angle group had higher BMIs (P = .01; large effect size [g = 1.54], Fig. 1A) and body fat percentages (P < .01; large effect size [g = 4.01], Fig. 1B) than males in the narrow phase angle group. Males had greater android/gynoid body fat than females for both the narrow (P < .01; large effect size [g = 2.18], Fig. 1C) and wide phase angle groups (P < .01; large effect size [g = 1.49]), Fig. 1C). Finally, males in the narrow phase angle group had higher triglycerides (P < .01; large effect size [g = 1.73]), Fig. 1D) and had a trend of higher metabolic syndrome risk scores (P = .07; large effect size [g = 0.99]) than females in the narrow phase angle group.

Figure 1.

Cardiometabolic health metrics in females and males with narrow and wide phase angles. Box and whisker plots for BMI (A), body fat percentage (B), android/gynoid percentage (C), glucose (D), triglycerides (E), and metabolic syndrome risk score (F) in females and males with narrow vs wide phase angles, where the median (dark grey line), the first and the third quartile (lower and upper limits of the boxplots, respectively), and the minimum and the maximum value no more than 1.5× the interquartile range (lower and upper limits of the whiskers, respectively) are plotted. Data beyond the end of the whiskers are outlying points. n = 30 (15 F). Values are presented as mean ± SD. *Statistically significant difference between groups (P < .05) from independent t-test comparisons.

Abbreviation: BMI, body mass index.

Open in new tabDownload slide

Within sex, there was no effect of phase angle group on BMI for females or males (Fig. 1A); however, group differences showed higher body fat percentages within females in the narrow vs wide phase angle groups (P = .04; Fig. 1B), whereas males in the narrow phase angle group had higher android/gynoid body fat percentage than males in the wide phase angle group (P = .03; Fig. 1C). Hedge's g indicated a large effect size for android/gynoid body fat within males between the phase angle groups (g = 1.02) and a negligible effect size for android/gynoid body fat in females between the phase angle groups (g = 0.08). Group effects were also observed for cardiometabolic health markers across sexes. Females in the narrow phase angle group had higher glucose (P = .04; large effect size [g = 1.02], Fig. 1D) and resting heart rate values than the wide phase angle group (77 ± 4 vs 68 ± 9 bpm, respectively; P = .02; large effect size [g = 1.24]); there was no significant difference between male phase angle groups for glucose (P = .89; small effect size [g = 0.44]) or heart rate (P = .67; small effect size [g = 0.25]). Conversely, males in the narrow phase angle group had higher triglyceride levels (P < .01; large effect size [g = 1.50], Fig. 1E) and metabolic syndrome risk scores (P = .04; large effect size [g = 0.94], Fig. 1F) than males in the wider phase angle group. Group differences were not observed in females for triglycerides (P = .50; negligible effect size [g < 0.01]) or metabolic syndrome risk scores (P = .17; moderate effect size [g = 0.51]).

In females, Pearson correlations showed strong negative associations between phase angle and body fat percentage (r = −0.62, P = .01; Figs. 2 and 3) as well as resting heart rate (r = −0.73, P < .01; Figs. 2 and 3), but with no statistically significant correlation for glucose (r = −0.32, P = .25). The relationship between phase angle and body fat percentage (adjusted r² = 0.65) and heart rate (adjusted r² = 0.47) remained significant after controlling for BMI (both P < .01). These relationships were not significant in males. Phase angle was negatively correlated with android/gynoid body fat percentage (r = −0.53, P = .04; Figs. 2 and 3) in males, but not females, and there was also a trending negative association with triglyceride levels in the male subgroup (r = −0.46, P = .08; Figs. 2 and 3). The relationship between phase angle and android/gynoid body fat percentage did not reach statistical significance after controlling for BMI (adjusted r² = 0.18, P = .12).

Figure 2.

Correlation matrices of sleep, circadian, and cardiometabolic health markers in males and females with overweight/obesity. The correlation matrix displays Pearson correlation results between each labeled variable at the end of the row/column. *Denotes correlations reaching statistical significance (dark gray boxes) (P < .05). ^†Indicates nonsignificant trending correlations (light gray boxes) (P < .1).

Open in new tabDownload slide

Figure 3.

Associations between sleep, circadian, and cardiometabolic health markers in males and females with overweight/obesity. The solid lines represent the Pearson correlation regression lines and the shaded region represents 95% confidence intervals.

Open in new tabDownload slide

Discussion

There is growing interest in how physiological mechanisms in females and males contribute to sex-related disease risk (19); thus, this investigation sought to identify the influence of circadian alignment on cardiometabolic health outcomes in females and males with overweight/obesity. Sex-specific differences in cardiometabolic health outcomes included (1) males with narrower phase angles having higher levels of abdominal fat distribution, circulating levels of triglycerides, and metabolic syndrome risk scores, and (2) females with narrower phase angles having higher overall percent body fat, circulating levels of glucose, and resting heart rates. Additionally, the degree of circadian alignment (as assessed by phase angle) was negatively associated with cardiometabolic disease risk factors in males, but inversely related to body fat and heart rate in females. These findings suggest that sleep and circadian disruption may not only promote a trajectory of weight gain (20), but could also contribute to negative health consequences in a sex-dependent manner in adults already with overweight/obesity.

Investigations into circadian rhythm disruptions on cardiometabolic outcomes have observed alterations in metabolic pathways, increased blood pressure, and upregulation of inflammatory markers in ostensibly healthy adults (21-23). Evidence further shows that circadian alignment is associated with body fat distribution (24, 25), particularly among in males with healthy weight (8). However, our understanding of how circadian mechanisms affect cardiometabolic regulatory pathways in obesity, particularly in a sex-specific fashion, remains incomplete. Altered circadian patterns of energy metabolism in participants with obesity (26) may increase the susceptibility of metabolic dysregulation when circadian disturbances are experienced (10). Furthermore, sex differences in metabolic regulation (12), especially in circadian misalignment (9), likely affect the complex interplay between obesity, the circadian timing system, and metabolic health. Though sex differences in the circadian timing system have been well documented (27-29), future investigations utilizing in-laboratory protocols are warranted to tease apart potential mechanistic circadian pathways in individuals with overweight/obesity, particularly as it pertains to sex-specific outcomes.

While body fat distribution and its associated metabolic implications for disease risk differs between sexes (11, 12), the present study provides valuable perspective into how the circadian system influences health in a sex-specific manner. Android body fat distribution, specifically, has been shown to be associated with metabolic risk factors (30), including changes in serum lipid profiles and metabolic syndrome risk (31, 32). Thus, circadian disruptions may facilitate metabolic disease risk, in part, through mechanisms that promote central adiposity and increased triglyceride levels, particularly among males. There is a critical need, however, to identify the circadian effects on metabolic regulation in females, as the relationship between shiftwork duration and cardiometabolic disorders is stronger in female shift workers (33, 34). The findings from the present study offer insight into how circadian mechanisms could differentially affect glucose and lipid metabolism in ostensibly healthy females with overweight/obesity. Indeed, less circadian alignment may be an early contributor to metabolic dysregulation, though obesity related outcomes in females may also be affected by the influence of sleep and sex hormones on substrate utilization (35). More work is needed to disentangle the underlying biological mechanisms that contribute to sex disparities in cardiometabolic disease risk, particularly among females as they may be more vulnerable to adverse health outcomes with extreme circadian disturbances, such as that occurring with shiftwork.

Phase angle of entrainment is considered to be a marker of the alignment between the internal circadian timing system (eg, DLMO) and external behavioral markers (eg, sleep onset). In individuals with both controlled (ie, consistent sleep/wake times) and self-selected sleep schedules, the time duration between DLMO and sleep onset is approximately 2 hours on average (36, 37). Indeed, several contributors influence individual differences in phase angle, such as age (38), sex (28, 29), and sleep timing preference (39, 40); however, the current investigation adds to a growing body of evidence showing inverse correlations between phase angle and markers associated with poorer health outcomes (8, 10, 15). While the exact mechanisms contributing to these relationships is unclear, factors such as morning circadian misalignment (10) and/or interactions between circadian and sleep processes (8) have been proposed.

Finally, these findings highlight how circadian strategies should be considered when developing innovative sex-specific approaches for early detection and targeted treatment of obesity-related cardiometabolic disease. Though the assessment of circulating melatonin may not be presently available clinically, at home saliva sampling kits (41) or mathematically modeled melatonin onset from actigraphy (24, 42) could provide opportunity for clinical utility for precision medicine. In the absence of those measures, other proxies such as chronotype (43) or midsleep timing (44) could be utilized, though further work is needed to tease apart sex-specific interactions with those sleep and circadian metrics.

A few limitations should be considered when interpreting these findings. First, the small sample of ostensibly healthy adults with excess body weight may have limited the capacity to identify significant associations between outcomes, potentially underestimating the circadian influences on metabolic disease risk in populations with comorbidities. Furthermore, the sample size restricted the ability to test for interactive effects between sex and circadian alignment on health markers. However, given the need to understand how sex affects disease risk (45) and the notable sex disparities in body composition and disease risk (11, 12), the sex-specific analyses in the current investigation begin to elucidate how circadian mechanisms can differentially affect health outcomes in females and males. Additionally, the current sample likely lacked the statistical power to detect smaller, yet potentially important relationships between circadian factors and cardiometabolic health outcomes independent of BMI. Previous work has suggested that circadian mechanisms play a role in body composition outcomes even in young, primarily lean males (8). However, circadian phase has been shown to be associated with metabolic health markers only in individuals with overweight/obesity and unrelated in individuals with healthy weight (10); thus, individuals with overweight/obesity may be more vulnerable to circadian disruptions. Second, menstrual phase was not assessed, restricting our ability to draw conclusions regarding the independent and combined effects of circadian mechanisms and hormonal fluctuations on cardiometabolic health markers (46). Third, the use of diary-derived sleep onset timing may have yielded less accurate sleep onset times than other methods. Last, the cross-sectional nature of the protocol may have limited our ability to measure behaviors that may be confounding to our outcomes.

Conclusion

This study increases our understanding of how sex assigned at birth and degree of circadian misalignment could affect obesity-related metabolic outcomes. Furthermore, these findings offer valuable insight that can inform precision medicine approaches to enhance personalized disease risk assessment. Optimized circadian alignment should be considered as a potential therapeutic approach to mitigating adverse health outcomes, particularly in individuals with overweight/obesity.

Acknowledgments

The authors thank Diane Elliot, MD, Jonathan Emens, MD, Melanie Gillingham, PhD, Melina Rodriguez, Marykathryn Kordash, Josie Velasco, and the Oregon Clinical and Translational Research Institute staff and their support in conducting these studies.

Funding

This work was supported by National Institutes of Health (NIH) T32HL083808, K01HL146992, R01HL105495, UL1TR000128, UL1TR002369, R35HL155681, and by the Oregon Institute of Occupational Health Sciences at Oregon Health & Science University via funds from the Division of Consumer and Business Services of the State of Oregon (ORS 656.630).

Disclosures

Andrew McHill, PhD, consults for Pure Somni Corporation. The remaining authors have no conflicts of interest to declare.

Data Availability

The data that support the findings presented in the current study will be shared from the corresponding author upon reasonable request.

References

Abbreviations

 
• BMI

  body mass index

 
• DLMO

  dim-light melatonin onset