Changes in sleep duration and risk of metabolic syndrome: the Kailuan prospective study

Using a large longitudinal data set spanning 4 years, we examined whether a change in self-reported sleep duration is associated with metabolic syndrome (MetS). Current analysis included 15,753 participants who were free of MetS during both 2006–2007 and 2010–2011. Sleep duration was categorized into seven groups: ≤5.5 h, 6.0–6.5 h, 7.0 h, 7.5–8.0 h, ≥8.5 h, decrease ≥2 h, and increase ≥2 h. Cox proportional hazards models were used to calculate hazard ratios (HRs) and their confidence intervals (CI) for MetS, according to sleep duration. Compared to the reference group of persistent 7-h sleepers, a decrease of ≥2 h sleep per night was associated with a higher risk of incident MetS (HR = 1.23, 95% CI = 1.05–1.44) in analyses adjusted for age, sex, sleep duration at baseline, marital status, monthly income per family member, education level, smoking status, drinking status, physical activity, body mass index, snoring status and resting heart rate. An increased risk of MetS incidence was also observed in persistent short sleepers (average ≤5.5 h/night; HR = 1.22, 95% CI = 1.01–1.50). This study suggests individuals whose sleep duration decreases ≥2 h per night are at an increased risk of MetS.

shows baseline characteristics according to sleep duration. Significant association was found between sleep duration and age, sex, education level, income level, smoking status, drinking status, physical activity, BMI, SBP, DBP, FBG, TG, HDL-C, RHR, snoring status, history of stroke, myocardial infarction, and cancer (p < 0.001).
During an average 3.4-year follow-up, 6,302 participants (40.0%) developed MetS. Age, sex, sleep duration at baseline, sex, marital status, monthly income per family member, education level, smoking status, drinking status, physical activity, body mass index, snoring status and resting heart rate were designated as confounding factors in Model 2. After adjusting for these confounding factors, participants who slept ≤ 5.5 h per night had an increased risk of developing MetS compared with participants who persistently slept 7 h (HR = 1.22, 95% CI = 1.01-1.50; all p-values < 0.05). Compared with the reference group of persistent 7-h sleepers, a decrease of ≥ 2 h sleep per night was associated with a higher risk of MetS incidence (HR = 1.23, 95% CI = 1.05-1.44; all p-values < 0.05) (Table 3). Moreover, the association between decreased sleep duration and risk of MetS remained significant upon repeating our analysis and excluding individuals with stroke, myocardial infarction and cancer, respectively (Table 4).

Discussion
Our findings suggest persistent short sleep (average < 5.5 h) increases the risk of MetS incidence. In contrast, increased sleep duration was not associated with higher incidence of MetS compared with individuals persistently sleeping 7 h per night. Compared to these individuals, we also found that a decrease of ≥ 2 h in sleep duration over a four-year exposure period was associated with an increased risk of developing MetS in analyses adjusted for age, sex, sleep duration at baseline, sex, marital status, monthly income per family member, education level, smoking status, drinking status, physical activity, body mass index, snoring status and resting heart rate. A set of sensitivity analyses further confirmed these findings.
After 3.4 years of follow-up, the overall incidence of MetS in this study was 40.0%, which was significantly higher than a previously published report about midlife adults comprising a sub-cohort of ARIRANG and Quebec Family Study patients [6][7][8] . A recent meta-analysis and several prospective studies have suggested that short rather than long sleep duration is significantly associated with increased risk of MetS 2,6-8 . Our study results were quite similar to the above-published study. However, some studies have shown that longer sleep duration may also be a risk factor in participants 9,10 . Prior evidence connecting sleep duration to cardiometabolic risk is varied 17 , with a reported U-shaped relationship between sleep duration and MetS 18 , type 2 diabetes 19,20 , and mortality 13,21 . Various factors may contribute to this difference, such as geographic and ethnic variations, varying clinical definitions of MetS, limited confounding factors, and major diseases affecting sleep duration. Thus, in our study, a greater quantity of important influencing factors has been analyzed than in previous research, including RHR and snoring status. First, taking into consideration prior evidence showing RHR is an independent risk factor for existing MetS and a powerful predictor for future incidence of MetS 22-25 , we adjusted for RHR in our full model. Second, poor sleep efficiency, often relating to snoring status, has a significant correlation with over-activity of the sympathetic nervous system, which could result in insulin resistance 26 and increased blood pressure 27 . In light of this, snoring was considered as a confounding factor in our assessment of the relative risks for MetS. Moreover, age, sex, socioeconomic status may contribute to the association between the altered sleep duration and incident MetS, as advanced age is associated with changes in sleep architecture with increased difficulties in sleep initiation and maintenance 28 . And, sleep problems are particularly common in people with anxiety, depression, bipolar disorder, and poor socioeconomic status. Compared to men, women are more likely to develop depression and sleep problems. Therefore, we adjusted for age, sex, income level and education level in the full model. Finally, subjects with stroke, myocardial infarction, and cancer may experience sleep deprivation, which could lead to biased results. Therefore, we further confirmed our finding that short sleep duration increases the risk of MetS incidence by repeating the same analysis and excluding individuals with stroke, myocardial infarction and cancer, respectively. An inherent limitation of previous studies has been the reliance on a single time point by which to assess sleep duration, which may have occurred several decades prior to the event, and is therefore likely to yield biased estimates of the association. Moreover, there has been no consideration of how sleep duration varies within individuals over time, and the subsequent impact this could have on changes in sleep duration and future risk of MetS. As the current study appears to be the first to demonstrate an association between decreased sleep duration and incidence of MetS in a large cohort study from China, our findings thus provide the first evidence that change in sleep duration is not only strongly associated with future risk of MetS, it is also likely to be a more accurate  indicator of the true magnitude of risk as compared with a single measure of sleep duration self-reported several years before the onset of MetS. It is an interesting and new observation that a decrease in sleep duration over a 4-year period is more deleterious than persistent short sleep. This may be related to the shortened sleep duration (≤ 5.5 h) among sleepers with a significant decrease in sleep duration. There are a number of biological mechanisms through which reduced sleep duration may lead to MetS. Experiments have demonstrated short-term sleep deprivation among healthy subjects results in adverse physiological changes, including decreased glucose tolerance and increased insulin resistance, sympathetic tone and blood pressure 29 .
The strengths of our study include a prospective cohort design, large sample size, the Asian ethnicity of our participants, and a broad spectrum of potential confounding parameters. However, there are several potential limitations of our study. First, we only collected information on sleep duration by self-reported questionnaires, without 24-h polysomnography information. However, self-reported measures showed good agreement when compared to quantitative sleep assessment with monitoring 30,31 . Second, our study did not exclude participants with obstructive sleep apnea syndrome, and there is some evidence that sleep apnea is associated with increased risk of MetS 32 . Furthermore, we did not collect sufficient information on the pre-or post-menopause status of women, which appears to be an important determinant of MetS risk in women 8 . Previous studies have reported gender-specific differences in sleep patterns may be influenced by differences in social or household roles, or sex hormones 33 . Finally, most of the participants from the Kailuan coal mine were male; thus, the sex distribution of participants was unbalanced and cannot be viewed as a representative sample of the general Chinese population.
In conclusion, our findings demonstrate an association between reduced sleep duration and increased future risk of MetS. This study also highlights the need to take into consideration change of sleep duration when estimating risk, rather than relying on a single measure of exposure that often precedes the outcome by several decades. Our results should encourage and support individuals to maintain or adopt a 7-h sleep duration each night, as this could have significant beneficial effects in stemming the growing prevalence of MetS.  (Fig. 1). Follow-up evaluations included biennial measurement of laboratory parameters and recording of adverse events. All physicians and nurses had rigorous, unified training before conducting this study. Assessment of Potential Covariates. All participants underwent a clinical examination and standardized interview. Physical activity was evaluated based on individual responses to questions regarding the types and  frequencies of physical activity at work and during leisure time. Physical activity was classified as "≥ 4 times per week and ≥ 20 min at a time", "< 80 min per week", or "none". Smoking and drinking statuses were classified as "never", "former", or "current" according to self-reported information. Monthly income per family member (at baseline) was categorized as "< ¥600", "¥600-799", "¥800-999" and "≥ ¥1,000". Anthropomorphic parameters such as height, weight, and waist circumference were measured. Body mass index (BMI) was calculated as weight/height (kg/m 2 ). Systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured thrice in the seated position using a mercury sphygmomanometer, and the average of three readings was used for analyses.

Methods
Blood samples were collected from the antecubital vein after an overnight fast. Venous blood was obtained for determination of routine chemistry, including fasting blood glucose (FBG), high-density lipoprotein-cholesterol (HDL-C), total cholesterol (TC), and triglycerides (TG). Resting heart rate (RHR) was measured and calculated from electrocardiogram recordings after subjects acclimated to the hospital setting for ≥ 30 min and were in the supine position for ≥ 5 min. In addition, participants were asked, "Do you generally snore when you sleep?" Response alternatives were "yes" and "no. " Follow-Up and Diagnosis of Metabolic Syndrome (MetS). Participants were followed up by face-to-face interviews at every 2-year routine medical examination until December 31, 2015, or to the event of interest or death. Follow-ups were performed by trained physicians who were blinded to baseline data. MetS was diagnosed when a participant had three or more of the following components: 1) Waist circumference ≥ 90 cm for men or ≥ 80 cm for women; 2) TG ≥ 1.7 mmol/l; 3) HDL-C < 1.03 mmol/l for men or < 1.30 mmol/l for women; 4) SBP/DBP ≥ 130/85 mmHg or current use of antihypertensive medications; 5) FBG ≥ 5.6 mmol/l, previous diagnosis of type 2 diabetes, or current use of oral hypoglycemic agents or insulin 36 . Statistical analysis. Continuous variables were expressed as means ± standard deviations; whereas, categorical variables were expressed as percentages. We compared parameters according to each sleep duration group. One-way analysis of variance (ANOVA) was used for non-paired samples of normally distributed parameters and the Kruskal-Waillis test was applied for non-parametric variables. A Chi-squared test was applied to compare categorical variables. A multivariate analysis was performed using two models: Model 1 was adjusted for age, sex, and sleep duration at baseline; Model 2 included Model 1 parameters plus monthly income per family member, education level, marital status, smoking status, drinking status, physical activity, BMI, snoring status and RHR; We used Cox proportional hazards modeling to calculate the hazard ratio (HR) and 95% confidence interval (CI) of MetS, using the group with persistent 7-h sleep duration as a reference category. Person-years were calculated from the date of the 2010 survey was conducted to the date when MetS was detected (depending on the analysis in question), date of death or date of participating in the last interview in this analysis, whichever came first. Further, as individuals with major fatal diseases could impact our assessment of sleep duration and future MetS risk, we conducted three sensitivity analyses to test the robustness of our findings by repeating our aforementioned analysis and excluding individuals with stroke, myocardial infarction and cancer, respectively. Statistical analysis was performed using SAS 9.3 statistical software (SAS Institute, Cary, NC).