Do the Long-Term Cardiometabolic Effects of Night Shift Work Differ Between Men and Women? A 17-Year Cohort Study of the U.S. Workforce

Abstract

Background

Shift work, particularly in night shifts, has been linked to adverse health outcomes, including disruptions in circadian rhythms that may contribute to metabolic disorders. However, sex-related differences in the impact of night shift work on cardiometabolic health remain understudied. This study sought to examine the relationship between night shift work and cardiometabolic health in young adults, identifying potential sex-related differences.

Methods

In this prospective cohort study, a total of 3201 participants (mean age: 22.08 years, 59.9% women) were analyzed over a 17-year period. Relative risks (RRs) with 95% confidence intervals (CIs) were calculated to evaluate the impact of night shift work on cardiometabolic health outcomes, including obesity, hyperlipidemia, hypertension, type 2 diabetes, and metabolic syndrome (MetS).

Results

At Wave V, 74.7% of participants were diagnosed with overweight/obesity, 51.8% with abdominal obesity, 15.5% with hyperlipidemia, 29.1% with hypertension, 8.3% with type 2 diabetes, and 20.9% with MetS. Night shift work was associated with an increased risk of obesity and abdominal obesity in women (RR = 1.99, 95% CI: 1.08 to 3.65 and RR = 1.60, 95% CI: 1.07 to 2.56, respectively) but not in men. Women also showed an elevated risk of diabetes (RR = 2.69, 95% CI: 1.49 to 4.86), whereas no significant relationship was identified in men. Night shift work was not significantly associated with hypertension and MetS in either men or women.

Conclusion

Women working night shifts may benefit from targeted interventions for cardiometabolic health, particularly in managing weight and preventing diabetes, while men did not exhibit similar associations.

1. Introduction

Cardiometabolic conditions, including obesity, hyperlipidemia, hypertension, type 2 diabetes, and metabolic syndrome (MetS), have become increasingly prevalent and are recognized as leading contributors to global morbidity and mortality [1,2]. These conditions not only are major public health concerns but also elevate the risk of developing more severe cardiovascular complications, including heart disease and stroke [3]. Addressing the factors that contribute to cardiometabolic health is therefore critical for public health strategies aimed at reducing the burden of chronic disease [2].

One factor that has gained increasing attention due to its impact on health is night shift work [4]. This type of nonstandard working schedule not only is necessary for sustaining critical services and improving industrial productivity but also remains indispensable for sectors like tourism, entertainment, and customer service, where continuous operation is crucial [5]. As a result, this has led to a significant rise in the number of individuals working outside traditional daytime hours [6]. Night shift work, however, disrupts the body’s natural circadian rhythms, leading to disturbances in sleep patterns, social life, eating behaviors, and stress [4], and these disruptions are suggested to have a profound impact on cardiometabolic health [7]. Several studies have linked night shift work with adverse health outcomes, particularly those related to metabolic and cardiovascular function, including a higher risk for developing obesity [8], insulin resistance, hypertension, and hyperlipidemia [9]. However, while some studies have considered the overall impact of shift work on cardiometabolic health, few have explored how these effects may differ between men and women [10,11]. Given the cultural, physiological, and hormonal differences between sexes, there is reason to believe that the cardiometabolic consequences of night shift work may vary significantly. Understanding the cardiometabolic effects of night shift work on general workers, and the potential inequalities between men and women, is essential for identifying individuals at a higher risk and for developing targeted prevention and intervention strategies.

For these reasons, the present study sought to examine the relationship between night shift work and cardiometabolic health—specifically obesity, hyperlipidemia, hypertension, type 2 diabetes, and MetS —among young adults, identifying potential sex-related differences.

2. Methods

2.1. Population sample and study design

This study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines and utilized data from the Add Health study, a prospective cohort study including a nationally representative sample of U.S. adolescents in grades 7–12, tracked from adolescence into adulthood. We specifically analyzed data from waves III (2001–2002), IV (2008–2009), and V (2016–2018), during which information on work-shift type was available. After excluding cases with missing data on relevant parameters, the final sample consisted of 3201 adults followed for 17 years.

The Add Health study was approved by the Institutional Review Board at the University of North Carolina at Chapel Hill, and permission for secondary analysis was granted by the Ethics Committee of the University Hospital of Navarra (PI_2020/143).

2.2. Shift work (an independent variable)

Information on shift work was gathered from employed participants using the question: “Which of these categories best describes the hours you work at this job?” Response options included “regular day shift, regular evening shift, regular night shift, rotating shift (changing periodically between day, evening, and night), split shift (two distinct periods each day), or irregular schedule.” For analysis purposes, work schedules were grouped into two categories (1): regular daytime shift and (2) night shift or rotating night shift. Participants were classified as night-shift workers if they reported working night shifts in both Wave III and Wave IV. Participants reporting split or irregular shifts were excluded from the analysis as it was not possible to determine with certainty whether their schedule involved night work (n = 1065).

2.3. Cardiometabolic health (dependent variables)

For the assessment of cardiometabolic markers, we utilized biological and clinical data from Wave V, including measurements of blood pressure (BP), lipid profiles [total cholesterol, low-density lipoprotein cholesterol, high-density lipoprotein cholesterol (HDL-c), and triglycerides], height, weight, waist circumference (WC), and diabetes indicators (fasting or nonfasting glucose and/or glycated hemoglobin (HbA1c)]. Following established clinical guidelines, these markers were dichotomized as either 0 (no risk) or 1 (risk present) based on the following criteria:

• a)Obesity: Defined as a body mass index (BMI) ≥30 kg/m². This was measured through height and weight, which were recorded during Wave V, and BMI was calculated by dividing weight (in kilograms) by height (in m²). WC was measured with a precision of 0.5 cm at the iliac crest, and abdominal obesity was defined as a WC ≥102 cm for men and ≥88 cm for women [12].
• b)Hypertension: Classified as either stage 1 (a BP of 140–159/80–89 mmHg) or stage 2 (BP ≥ 160/≥100 mmHg) according to the Seventh Joint National Committee (JNC 7) guidelines [13] or self-reported physician-diagnosed hypertension or use of antihypertensive medications.
• c)Hyperlipidemia: Defined by self-reported high total cholesterol or triglycerides, recent use of lipid-lowering medications, fasting triglycerides ≥500 mg/dL, or low-density lipoprotein cholesterol levels ≥190 mg/dL, based on the American College of Cardiology and the American Heart Association classifications [14].
• d)Type 2 diabetes: Identified by fasting glucose levels ≥126 mg/dL, nonfasting glucose levels ≥200 mg/dL, HbA1c ≥6.5%, self-reported diabetes (excluding cases during pregnancy), or the use of antidiabetic medication [15].
• e)MetS: Diagnosed when three or more of the following criteria were met: abdominal obesity, elevated triglycerides (≥500 mg/dL, nonfasting), hypertension, elevated glucose (≥126 mg/dL, nonfasting), and low HDL-c (<40 mg/dL in men and <50 mg/dL in women) [14].
• f)Blood pressure: Trained professional field staff measured BP using a calibrated oscillometric device (BP3MC1-PC-IB, Micro-Life USA; Dunedin, FL) with a cuff that was properly sized for each participant. After a 5-minute seated rest, three measurements of systolic and diastolic BP were taken from the right arm at 30-second intervals, and the average of the last two readings was utilized for analysis. Following the 2017 guidelines set by the American College of Cardiology and the American Heart Association [16], hypertension was defined as having a systolic BP ≥130 mmHg and/or a diastolic BP ≥80 mmHg, a prior diagnosis of hypertension by a physician, or current use of antihypertensive medications.

2.4. Covariates

Age, biological sex, race/ethnicity, and alcohol consumption were all measured through in-home questionnaires during Wave III. Race/ethnicity was determined using the question “Which one category best describes your racial background?“ Based on the participants’ responses, individuals were categorized into four groups: White, Black or African American, American Indian or Alaska Native, and Asian. Alcohol consumption was evaluated by asking the question “In the past 30 days, on how many days did you consume alcohol (beer, wine, or liquor)?“ Based on their responses, participants were grouped into the following categories: “None,” “1 or 2 days in the past 12 months,” “Once a month or less (3 to 12 times in the past 12 months),” “2 or 3 days a month,” “1 or 2 days a week,” “3 to 5 days a week,” and “Every day or almost every day.”

Lastly, job type and weekly hours were analyzed through the following questions “How many hours a week do you usually spend at work?” and “What do you do in your [current] job?“

2.5. Statistical analysis

We described the analytic sample and performed tests to identify significant differences based on work shift. Descriptive statistics are presented as frequencies and percentages for categorical variables and as means and standard deviations for continuous variables. All model assumptions, including normality and homoscedasticity, were verified. When examining the interaction between night shifts and sex across various dependent variables (e.g., abdominal obesity and MetS, p < 0.001), a significant interaction was found, prompting separate analyses for men and women.

To assess mean differences (MDs) in continuous cardiometabolic risk factors (e.g., BMI, WC, BP, cholesterol levels, and glucose metabolism) between participants working night shifts and other shifts at waves III and IV, we applied generalized linear models with a Gaussian distribution, adjusting for relevant covariates. Although all model assumptions (e.g., normality and homoscedasticity) were evaluated, residuals were not normally distributed. Therefore, to obtain robust estimates, we used a nonparametric bias-corrected and accelerated bootstrap method with 5000 replicates, resampling the dependent variable with replacement.

For binary outcomes, specifically, the presence or absence of obesity, abdominal obesity, hyperlipidemia, hypertension, type 2 diabetes, or MetS at Wave V, we used generalized linear models with a binomial distribution and a logit link function (i.e., logistic regression) to evaluate the risk of developing obesity, abdominal obesity, hyperlipidemia, hypertension, type 2 diabetes, or MetS in relation to night shift work compared to other shifts. Night shift work status at waves III to V was included as the primary predictor. All models were adjusted for age, race/ethnicity, alcohol consumption, weekly hours worked, and job type, all measured at Wave III.

All analyses were performed using R (Version 4.3.2; R Foundation for Statistical Computing, Vienna, Austria) and RStudio (Version 2023.09.1 + 494; Posit, PBC, Boston, MA, USA), with statistical significance set at a two-sided p value < 0.05.

3. Results

The baseline characteristics of participants are presented in Table 1. The study included 3201 participants, with a mean age of 22.08 years, of whom 59.9% were women. In terms of race, a higher percentage of men were White (75.9% vs. 72.4%, p < 0.001), with similar proportions of Black or African American and American Indian or Alaska Native individuals across both groups. Alcohol consumption patterns varied significantly, with more men consuming alcohol frequently (3.5% reporting daily or almost daily intake vs. 0.7% for women, p = 0.002). Lastly, night shift work was more prevalent among men (5.3%) than among women (3.9%) (p < 0.001). At Wave V, 74.7% of participants were diagnosed with overweight/obesity, 51.8% with abdominal obesity, 15.5% with hyperlipidemia, 29.1% with hypertension, 8.3% with type 2 diabetes, and 20.9% with MetS.

Table 1.

Demographic characteristics of the participating subjects during Wave III according to sex

	Men (n = 1282)	Women (n = 1919)	p
Age, years	22.21 (1.74)	21.96 (1.72)	<0.001
Race/ethnicity, n (%)	3251 (75.9)	3446 (72.4)	<0.001
White	719 (16.8)	1011 (21.2)
Black or African American	43 (1.0)	41 (0.9)
American Indian or Alaska Native	271 (6.3)	261 (5.5)
Alcohol consumption, n (%)			0.002
None	151 (3.6)	194 (4.1)
1 or 2 days in the past 12 months	382 (9.0)	657 (13.9)
Once a month or less (3 to 12 times in the past 12 months)	548 (13.0)	1057 (22.3)
2 or 3 days a month	700 (16.6)	876 (18.5)
1 or 2 days a week	1020 (24.1)	720 (15.2)
3 to 5 days a week	441 (10.4)	190 (4.0)
Every day or almost every day	150 (3.5)	32 (0.7)
Night shift, n (%)	227 (5.3)	188 (3.9)	<0.001

Table 2 illustrates differences in cardiometabolic risk factors at Wave V between night shift’s and other shifts’ workers. Among women, night shift work was associated with a higher BMI [MD = 2.99 kg/m², 95% confidence interval (CI): (0.61 to 5.58), p = 0.016] and WC [MD = 5.13 cm, 95% CI: (0.05 to 9.82), p = 0.044], whereas no significant differences were observed in men. Additionally, women night-shift workers had significantly higher triglyceride levels [MD = 39.04 mg/dL, 95% CI: (8.23 to 73.89), p = 0.016] and a worsened cholesterol-to-HDL ratio [MD = 0.29, 95% CI: (0.02 to 0.62), p = 0.041], along with lower HDL levels [MD = -4.77 mg/dL, 95% CI: (-7.73 to -1.42), p = 0.005]. Fasting glucose levels were also significantly elevated in women working night shifts [MD = 5.95 mg/dL, 95% CI: (0.66 to 16.25), p = 0.045], while men did not show significant changes in these parameters.

Table 2.

Differences in cardiometabolic risk factors at Wave V between participants working night shifts and day shifts

	Men			Women
	MD	95%bca CI	p	MD	95%bca CI	p
Anthropometric variables
Body mass index, kg/m2	-0.49	-2.23 to 1.33	0.540	2.99	0.61 to 5.58	0.016
Waist circumference, cm	0.83	-4.96 to 3.24	0.705	5.13	0.05 to 9.82	0.044
Lipid profile
Total cholesterol, mg/dL	-3.47	-12.96 to 6.01	0.473	-2.04	-10.12 to 6.03	0.620
Triglycerides, mg/dL	-17.56	-35.68 to 2.14	0.077	39.04	8.23 to 73.89	0.016
High-density lipoprotein, mg/dL	0.97	-2.89 to 5.48	0.650	-4.77	-7.73 to -1.42	0.005
Low-density lipoprotein, mg/dL	-1.62	-6.66 to 9.15	0.705	-3.88	-11.16 to 3.28	0.289
Total cholesterol-to-high-density lipoprotein ratio	-0.04	-.36 to 0.42	0.836	0.29	0.02 to 0.62	0.041
Blood pressure
Systolic blood pressure, mmHg	4.32	0.59 to 1.18	0.026	2.08	-1.18 to 5.57	0.218
Diastolic blood pressure, mmHg	2.12	-0.95 to 5.25	0.169	1.84	-0.39 to 4.18	0.127
Glucose metabolism
Fasting glucose, mg/dL	-2.14	-5.47 to 9.10	0.572	5.95	0.66 to 16.25	0.045
Glycated hemoglobin, %	0.01	-1.49 to 0.19	0.929	0.20	-0.01 to 0.42	0.064

Abbreviations: bca, bias corrected accelerated (mean difference estimate is based on the median of the bootstrap distribution); CI, confidence interval; MD, mean difference.

Analyses were adjusted for age, race/ethnicity, alcohol intake, weekly hours worked, and job type, all measured at Wave III.

Reference: other shifts.

Bold values indicate statistical significance (p < 0.05).

Table 3 presents the relative risk (RR) for cardiometabolic health conditions at Wave V among night-shift workers. Women working night shifts had an increased risk of obesity [RR = 1.99, 95% CI: (1.08 to 3.65), p = 0.027], abdominal obesity [RR = 1.60, 95% CI: (1.07 to 2.56), p = 0.045], and type 2 diabetes [RR = 2.69, 95% CI: (1.49 to 4.86), p = 0.001]. In men, no significant associations were observed for these conditions, although there was a nonsignificant trend toward a higher risk of type 2 diabetes [RR = 1.70, 95% CI: (0.86 to 3.40), p = 0.129]. Both men and women exhibited an increased, though nonsignificant, risk for MetS [RR = 1.50, 95% CI: (0.72 to 3.13), p = 0.277 for men; RR = 1.59, 95% CI: (0.76 to 3.31), p = 0.216 for women].

Table 3.

Relative risks and 95% confidence intervals for cardiometabolic health conditions at Wave V for night-shift workers by sex

	Men			Women
	RR	95% CI	p	RR	95% CI	p
Anthropometric variables
Obesity	0.82	0.46 to 1.47	0.822	1.99	1.08 to 3.65	0.027
Abdominal obesity	1.29	0.79 to 2.11	0.306	1.60	1.07 to 2.56	0.045
Lipid profile
Hyperlipidemia	1.07	0.59 to 1.93	0.828	1.54	0.82 to 2.92	0.181
Blood pressure
Hypertension	1.24	0.76 to 2.02	0.388	1.53	0.92 to 2.56	0.102
Glucose metabolism
Type 2 diabetes	1.70	0.86 to 3.40	0.129	2.69	1.49 to 4.86	0.001
Metabolic syndrome
3 or more conditions	1.50	0.72 to 3.13	0.277	1.59	0.76 to 3.31	0.216

Abbreviations: CI, confidence intervals; RR, relative risk.

Analyses were adjusted for age, race/ethnicity, alcohol intake, weekly hours worked, and job type, all measured at Wave III.

Reference (RR = 1.00): other shifts.

Bold values indicate statistical significance (p < 0.05).

4. Discussion

This study investigated the long-term association between night shift work and various cardiometabolic conditions, with a particular focus on differences between men and women. Over a 17-year follow-up period, our results showed that women working night shifts exhibited increased risks of obesity, abdominal obesity, and type 2 diabetes. In contrast, these associations were not observed in men, where night shift work did not significantly affect cardiometabolic health. These results suggest that women may be more vulnerable to the negative cardiometabolic impacts of night shift work, highlighting the importance of considering sex-specific factors when assessing the health risks of non-standard work schedules.

There exist several reasons underlying the association between night shift work and cardiometabolic disturbances. Night shifts are known to disrupt the body’s natural circadian rhythms, a well-established cause of metabolic issues [4]. Circadian misalignment has been shown to increase insulin resistance, promote inflammation, and alter hormone levels such as leptin and cortisol, which play key roles in glucose metabolism [17]. Moreover, melatonin, a potent antioxidant and hormone essential for regulating circadian rhythms, is typically produced by the pineal gland in darkness but is suppressed by light exposure during night shifts, further exacerbating these disturbances [18]. Furthermore, sleep deprivation—common among shift workers—has been widely recognized as a contributing factor to obesity [19]. In addition to this, shift workers are reported to have poorer-quality diets [20], irregular eating patterns [21], higher alcohol consumption [22], higher smoking rates [23], and less physical activity [24]. Likewise, evidence from a recent umbrella review including 33 systematic reviews has also shown that the overall risk of having overweight is about 25% higher for shift workers and that this could reach up to 38% among night-shift workers specifically [25]. However, despite this context, additional explanations are needed to understand why women who work in night shifts are at a higher risk of cardiometabolic disturbances and not men.

Our findings revealed that women working night shifts exhibited significantly higher BMI, WC, triglycerides, total cholesterol-to-HDL-c ratio, reduced HDL, and higher fasting glucose levels than their day-shift counterparts. Similarly, the risks of developing obesity, abdominal obesity, and type 2 diabetes were significantly higher in women who worked night shifts than in those working standard hours. In agreement with our results, previous studies based exclusively on cohorts of women have reported that those working night shifts face a high risk of type 2 diabetes [[26], [27], [28]]. Studies analyzing sex differences in the association between night shift and obesity [10], and night shift with type 2 diabetes [11,29] also found stronger associations in women than in men, with a study highlighting that the risk of type 2 diabetes among men might be mitigated by high work demands, an active job, and high-strain work environments [11]. Moreover, although it was not measured in our study, the aforementioned studies found that the association between night shift work and diabetes incidence intensifies with prolonged exposure, suggesting that the risk increases with more years spent working night shifts.

A combination of biological, social, and environmental factors may explain why women are disproportionately affected by the metabolic consequences of night shift work, while men are not significantly impacted. Although chronotype was not evaluated in our study, women are generally more likely to be “morning type” and tend to prefer morning activities [30], which may intensify the negative impact of working night shifts. Additionally, their naturally shorter circadian rhythms—such as earlier melatonin release and body temperature regulation than in men [31]—make them more susceptible to disturbances caused by night work. Hormonal fluctuations, particularly in estrogen and progesterone, may further increase their vulnerability to the harmful effects of disrupted sleep and circadian misalignment [32]. Sex-related differences in glucose metabolism and insulin sensitivity may also contribute to this association. Women tend to have proportionally more adipose mass, higher circulating free fatty acids, higher intramyocellular lipid content, and less skeletal muscle mass than men [33], and when fitness levels are lower in women, they are more likely to exhibit insulin resistance than men [34]. Similarly, although not measured in our study, women with higher levels of free testosterone, as seen in conditions like polycystic ovary syndrome, are frequently associated with glucose intolerance and insulin resistance [35], suggesting a potential unexplored factor influencing our results. This, combined with women’s greater sensitivity to sleep disruption [36], can make them more vulnerable to metabolic disturbances. From a cultural and social perspective, women often bear a heavier load regarding balancing night shift and family responsibilities [37] and the societal pressure to meet caregiving expectations, even after demanding shifts [38]. This increased strain may trigger a stress response, and for women who are more sensitive to stress, this response may be linked to greater food consumption [39], further contributing to the sex-specific impact of night shift work on metabolic health.

4.1. Clinical and public health implications

These findings have significant clinical and public health implications, especially given the increasing prevalence of night shift work. There is a pressing need for targeted interventions to reduce the negative health effects of shift work, particularly for women. As an attempt to reduce the health effects of night shift work, various interventions, including controlled light exposure, shift schedule, and behavioral and pharmacological approaches have been shown to yield positive overall effects on chronic disease outcomes [40]. Workplace strategies should prioritize promoting work-life balance and addressing modifiable factors such as light exposure. For instance, improving sleep hygiene is important, and efforts should focus on educating workers about healthy sleep practices, providing opportunities for naps, and creating environments that facilitate restful sleep [41]. Encouraging healthy eating habits and promoting physical activity through workplace initiatives can further support cardiometabolic health. Additionally, aligning shifts with individual chronotypes and offering stress management resources could also help to prevent cardiometabolic disturbances among night-shift workers.

4.2. Strengths and limitations

The main strengths of this study include the use of a large sample size, a long follow-up period, and the stratification of data by sex. However, certain limitations must be considered. As an observational study, it cannot establish causality, and despite efforts to adjust for confounders, the possibility of residual confounding remains. Moreover, the use of self-reported data for variables like alcohol consumption and night shift work introduces the potential for reporting bias. Since the sample was predominantly White and relatively young, generalization may be limited across different demographic groups and occupational contexts. Moreover, we did not collect data on participants’ gender identity or the use of gender-affirming hormone therapy or surgery, which represents a potential limitation in interpreting our results. Finally, several important potential confounders—such as dietary patterns, use of exogenous hormones (e.g., contraceptives), reproductive history (e.g., the number of pregnancies, live births, and breastfeeding), and perimenopausal status or symptoms—were unavailable and thus not accounted for in the analyses. The absence of these variables may influence the observed associations and should be considered when interpreting the results.

Future research should investigate the mechanisms driving the sex differences observed in this study. It would also be valuable to assess the effectiveness of interventions aimed at reducing the health risks associated with night shift work, such as modifying work schedules, improving sleep hygiene, or implementing nutritional guidance.

In conclusion, our study highlights significant sex-specific differences in the association between night shift work and cardiometabolic health outcomes. Women appear to be particularly vulnerable to the adverse effects of night shift work, with higher risks of obesity, abdominal obesity, and diabetes. These findings highlight the need for targeted interventions and workplace policies that address the unique health challenges faced by night-shift workers, with a focus on preventing cardiometabolic conditions in women.

CRediT authorship contribution statement

María Romero-Parra: Writing – original draft, Methodology, Conceptualization. Óscar Caballero: Methodology, Data curation. Antonio García-Hermoso: Supervision, Formal analysis, Data curation. José Francisco López-Gil: Writing – review & editing, Data curation. Jacqueline Páez-Herrera: Methodology. Rodrigo Yáñez-Sepúlveda: Conceptualization. Yasmin Ezzatvar: Writing – review & editing, Supervision, Conceptualization.

Data availability statement

Due to our data protection agreements with the participating cohort study, we are unable to share individual-level data with third parties. According to Add Health’s data access policy, researchers can submit data requests to the steering committee. These requests will be reviewed promptly for confidentiality, data protection, and intellectual property considerations and will not be unreasonably denied. Researchers registered with Add Health can apply for access to its database by submitting an application (https://data.cpc.unc.edu/projects/2/view).

Statement on the Use of AI Tools

The authors declare that no artificial intelligence tools were utilized in any stage of the preparation of this manuscript.

Conflicts of interest

The authors declare that there are no conflicts of interest related to this study. This research was conducted independently, and no financial or personal relationships influenced the outcomes or interpretations presented in the paper. All authors have contributed to the research and have approved the final version of the manuscript for submission.