Background

The concept of shift work arose from industrial growth and the increase of 24-h workplaces, which required continuous staffing and irregular work schedules [1, 2]. In South Korea, the prevalence of night and shift work is highest in the field of manufacturing, followed by wholesale and retail businesses [3]. Although no consensus has been reached regarding the definition of shift work, this term is often used in reference to work hours outside of the conventional daytime period.

The major difficulties associated with shift work mainly involve work conducted during evening or overnight hours, due to its effects on circadian rhythm. Changes in circadian rhythms can disrupt homeostasis and lead to the desynchronization of enzymatic activity and metabolic function [4]. For example, evidence suggests a correlation between an altered distribution of food intake due to a mismatch in circadian rhythm (e.g., nighttime food ingestion) and increased cholesterol levels [5]. Circadian rhythm disturbances have also been identified as a significant factor related to cardiovascular disease (CVD). For example, an inability of the circadian rhythm governing oxygen supply to adapt promptly to the changing demands of night work will likely lead to myocardial infarction [4]. Furthermore, night workers are more likely to experience fatigue due to a lack of sleep [6]. Although this relationship is poorly understood, sleep deprivation has been identified as a potential risk factor for CVD [7].

CVD is the cause of substantial societal burdens worldwide and is the leading cause of death in South Korea, where the CVD-associated mortality rate has been increasing gradually in recent years. In 2017, diseases of the circulatory system accounted for 21.5% of all deaths in South Korea, second only to neoplasms (28.1%) [8]. The prevalence of dyslipidemia, a major risk factor for CVD [9], is also increasing in South Korea [10], with reported rates ranging from 30 to 60% [10]. Although age, hypertension, and obesity are commonly known risk factors for dyslipidemia, these factors are better controlled and moderated today than in previous periods [11]. Therefore, the increased prevalence of dyslipidemia in South Korea is likely attributable to lifestyle factors.

Previous studies have reported associations between irregular work schedules, particularly night work, and altered lipid profiles [12, 13]. Therefore, preventive measures are needed to mitigate lipid disorders and ensure the well-being of workers during non-standard working hours. Night work appears to serve as barrier to a healthy lifestyle and a threat to well-being, as a circadian rhythm mismatch can disrupt adequate sleeping and eating habits, leading to poor health [14]. We hypothesize that in night workers, insufficient amounts of sleep and irregular eating habits may contribute to the onset of dyslipidemia. In this study, therefore, we aimed to investigate and elucidate the association of dyslipidemia with night work.

Methods

Study participants

We collected data from the 2013 to 2016 Korea National Health and Nutrition Examination Survey (KNHANES), which was conducted by the Korea Centers for Disease Control and Prevention (KCDCP). The KNHANES is a self-reported, nationally representative survey of South Koreans of all ages and is designed to gather annual national data on sociodemographic, economic, and health-related conditions and behaviors. Since 2007, the collected data have been subjected to an annual review and approval by the KCDCP Research Ethics Review Committee. The KNHANES 2013–2016 included 31,908 participants. We excluded 25,285 of these participants for various reasons (Fig. 1). First, participants with a previous clinical diagnosis of dyslipidemia were excluded, as this may have influence the reliability of the outcome (n = 3328). Second, in this study, dyslipidemia was diagnosed via a blood samples collected during the KNHANES. Therefore, people younger than 30 years were excluded because they did not undergo blood testing as part of the survey (n = 9656). Third, our study aimed to examine specific relationships with dyslipidemia. As dyslipidemia and metabolic syndrome share a few diagnostic components, including high-density lipoprotein (HDL) cholesterol and triglycerides levels, participants who met the modified National Cholesterol Education Program Expert Panel Adult Treatment Panel III (NCEP-ATP III) diagnostic criteria for metabolic syndrome with a lower waist circumference were excluded to increase the validity of our study (n = 4163) [15, 16]. Fourth, people deemed ineligible because they were unemployed or were not representative of covariates considered in the study (failure to answer the questionnaires or lack of applicability) were also excluded (n = 8138). Finally, the analyzed sample comprised 5813 participants (men: 2821 and women: 2992).

Fig. 1
figure 1

Flow diagram of subject inclusion and exclusion

Full size image

Variables

Dyslipidemia, the dependent variable in this study, was diagnosed based on the levels of total, high density lipoprotein (HDL), low-density lipoprotein (LDL) cholesterol, and triglycerides in blood samples collected after 9–12 h of fasting. According to the 2015 Korean Guidelines for the Management of Dyslipidemia, one of the following four criteria was required: (a) total cholesterol ≥240 mg/dL, (b) HDL cholesterol ≤40 mg/dL, (c) LDL cholesterol ≥160 mg/dL, or (d) triglycerides ≥200 mg/dL [11].

The main independent variable was the work pattern, which included three categories: day, night, and other shifts. The day shift was defined as between 6 A.M. and 6 P.M., while the night shift merged both evening (6 P.M.–12 A.M.) and overnight work (12 A.M.–8 A.M.). Other shifts included various types of working patterns, such as alternating shifts (e.g., day-night-day), 24-h shifts (a full 24-h shift followed by a day(s) off), and split shifts (≥2 shifts within a day).

Socio-demographic, economic, health-related, and nutritional factors were also assessed. Socio-demographic factors included age (30–39, 40–49, 50–59, and ≥ 60 years), region (metropolitan or rural), educational level (high school or less or college and/or beyond), and marital status (married or unmarried). Economic factors included the household income (low, mid-low, mid-high, or high) and occupational category (white-, pink-, or blue-collar employment). Health-related factors included eating habits (regular consumption of breakfast, lunch, and dinner or skipping meals), physical activity/week (active: ≥150 min of moderate activity, ≥ 75 min of vigorous activity, or a mixture of both for ≥150 min; inactive: < 150 min of moderate activity, < 75 min of vigorous activity, or a mixture of both for < 150 min), sleep duration (0–6 or ≥ 7 h per night), smoking status (current smoker, ex-smoker, or non-smoker), alcohol consumption status (≥2 times/month or never), body mass index (BMI) defined obesity status (in reference to the Korean guidelines for overweight and obesity; underweight/normal: < 23, overweight: 23–24.9, and obese: ≥25) [17], hypertension (in reference to the Korean guideline for normal BP, < 120/80 mmHg; normal: 90–199 mmHg systolic or 60–79 mmHg diastolic; prehypertension: 120–139 mmHg systolic or 80–89 mmHg diastolic; hypertension: ≥140 mmHg systolic or ≥ 90 mmHg diastolic) [18], and menopausal status (yes or no). Nutritional factors included macronutrient intake (total kcal, protein, fat, and carbohydrate). For the continuous variables (macronutrient intakes), the OR was calculated for every 100-kcal increase in calorie intake and every 10-g increase in protein, fat, and carbohydrate intake.

Statistical analysis

All statistical analyses were performed using SAS version 9.4 (SAS Inc., Cary, NC, USA). The chi-square (χ [2]) test was used to evaluate the general characteristics of the study population. For continuous variables (macronutrient intake), a t-test was used to calculate the means and standard deviations. A multiple logistic regression analysis was used to calculate the odds ratios (ORs) with 95% confidence intervals (CIs) in three different models. Model 1 yielded a crude OR, model 2 was adjusted for socio-demographic and economic factors, and model 3 was adjusted for all socio-demographic, economic, health-related, and nutritional factors. Multiple logistic regression analyses of subgroups were also performed to examine the association between night work and dyslipidemia according to occupational category, eating habits, and sleep duration. A general linear model analysis was also used to calculate the mean levels of the four diagnostic determinants (total cholesterol, HDL cholesterol, LDL cholesterol, and triglycerides), and the distributions and percentages of each were calculated. The stratified, clustering, and weight variables developed by the KNHANES were applied to all analyses to improve the representativeness of the sample and account for the limited proportion of participants retained in the final analysis [19]. The significance level was set at p value < 0.05.

Results

Table 1 summarizes the general characteristics of the study population, which included 2821 men and 2992 women. A total of 816 (28.9%) men and 469 (15.7%) women had dyslipidemia. Of the 196 male participants who reported working at night, 76 (38.8%) had dyslipidemia, and the prevalence of dyslipidemia was greater among these night workers compared to those who worked at other times (day: 684/2404, 28.5%; other shifts: 56/221, 25.3%). A similar trend was observed among the female participants, as 70 of the 379 women (18.5%) who reported working at night had dyslipidemia (day: 382/2512, 15.2%; other shifts: 17/101, 16.8%).

Table 1 General characteristics of the study population
Full size table

Table 2 summarizes the results from the multiple logistic analysis of the association between night work and dyslipidemia. In all three models, the association between night work and dyslipidemia remained statistically significant in male participants (model 1: OR = 1.58, 95% CI: 1.12–2.21; model 2: OR = 1.61, 95% CI: 1.13–2.29; model 3: OR = 1.53, 95% CI: 1.05–2.24). By contrast, however, no statistically significant association of night work with dyslipidemia was observed in female participants. However, women aged 50 years or older were more likely to have dyslipidemia, compared to their younger counterparts (50–59 years: OR = 1.61, 95% CI: 1.04–2.50; ≥60 years: OR = 1.66, 95% CI: 0.92–3.01).

Table 2 Odds ratio for dyslipidemia
Full size table

Table 3 summarizes the results from subgroup analyses stratified by occupational categories, eating habits, and sleep duration. Male night workers who reported skipping meals were more likely to have dyslipidemia, compared to their day working counterparts (OR = 1.63, 95% CI: 1.00–2.67). Similarly, male night workers who slept for 0–6 h were more likely to have dyslipidemia, compared to their day working counterparts. Among female participants, a strong significant association was observed between the occupational category and dyslipidemia, as female night workers with white collar jobs had a nearly three-fold risk of dyslipidemia, compared to their day working counterparts (OR = 2.95, 95% CI: 1.68–5.16).

Table 3 The results of subgroup analysis of dyslipidemia to work pattern stratified by occupational categories, eating habit, and sleep duration
Full size table

Table 4 individually summarizes the mean values of the four dyslipidemia diagnostic parameters: (a) total cholesterol, (b) HDL cholesterol, LDL, (c) cholesterol, and (d) triglycerides, as well as the related distributions and percentages of the study sample. Among male subjects, night workers were generally more likely to present with dyslipidemia, compared to their counterparts with other work shift patterns, with 8.2, 16.3, 3.6, and 15.3% meeting the respective criteria of ≥240 mg/dL total cholesterol, ≤40 mg/dL HDL cholesterol, ≥160 mg/dL LDL cholesterol, and ≥ 200 mg/dL triglycerides.

Table 4 Mean values of total cholesterol, HDL cholesterol, LDL cholesterol, and triglycerides
Full size table

Discussion

After controlling for socio-demographic, economic, health-related, and nutritional factors, we found that night work increased the risk of dyslipidemia in the male participants. Physiological activities, such as eating patterns, lipid/carbohydrate/glucose metabolism, and sleep, all operate on day/night rhythms [20] that are controlled by the circadian biological clock [20]. Accordingly, work schedules that extend beyond the standard 9 A.M.–5 P.M. period impair the circadian rhythm [21]. Night work-related disruptions of the biological clock are likely to result in obesity, impaired insulin secretion, and aberrant glucose homeostasis [20, 22]. Notably, overlap has been observed between the mechanisms associated with insulin resistance and atherosclerosis (a consequence of dyslipidemia), including elevated levels of glucose and free acids that cause oxidant stress, the activation of proinflammatory pathways, low levels of HDL, and high levels of triglycerides [23, 24]. The circadian clock is a key regulator of lipid metabolism and therefore, the lipid profile [25, 26], and periodic disruption of circadian rhythm negatively affects lipid metabolism [26, 27]. Accordingly, night work is more strongly associated with dyslipidemia, compared to day or other shift work.

Meal skipping is a common practice in modern society. Commonly, constant changes in the daily routines of night workers lead to irregular meal times. In our subgroup analysis, we observed a significant positive association of night work with dyslipidemia among male participants who reported skipping meals. Several previous studies reported that these workers tend to skip meals and snack more frequently during the night shift [28,29,30]. Additionally, compared with regular eaters, meal skippers have higher average values of mean weight, BMI, and triglycerides, which have all been identified as risk factors for dyslipidemia [31].

.Sleep deprivation negatively affects metabolism and impairs the homeostatic control of energy intake (i.e., protein, fat, and carbohydrate) [28, 32], while also promoting the development of an atherogenic lipid profile [33]. These effects explain the significant association between sleep duration and dyslipidemia in this study. Specifically, night workers who slept for < 7 h per night faced a higher risk of dyslipidemia, compared to their counterparts who reported more sleep. The National Sleep Foundation recommends that adults sleep for 7 h per night [34]. According to previous studies, permanent night workers receive less sleep than day workers [35, 36]. Night workers who sleep during the day will inevitably be exposed to light, which hinders the duration and quality of sleep [37]. Specifically, light is the main environmental regulator of circadian rhythm. As the human brain tends to wake when the environment transitions from darkness to light [38], night workers find it difficult to sleep during the day.

Previous studies have reported higher rates of physical inactivity and obesity among white-collar workers, particularly female workers, than those in other occupations [39, 40]. Furthermore, physical inactivity during working hours negatively affects the health of white-collar workers [41]. Both obesity and physical inactivity have been recognized as risk factors for dyslipidemia. These findings seem relevant to our findings, as our subgroup analysis showed a significant association between night work and dyslipidemia among female white-collar workers. Notably, age also correlated directly with the prevalence of dyslipidemia in women, particularly among menopausal women older than 50 years of age. This may be attributable to lipoprotein changes associated with menopause [42], which include increased levels of total and LDL cholesterol [42, 43].

This study had several limitations. First, the cross-sectional design rendered us unable to determine a causal relationship between night work and dyslipidemia. Second, the durations of day, night, and other shift work could not be determined because of limitations of the KNHANES questionnaire. Finally, the key covariates considered in this study, including the sleep duration and eating habits, were self-reported and may have been subject to recall bias. Despite these limitations, this study also featured strengths. This study involved a large, well-validated dataset collected from a nationally representative sample of the South Korean population. Therefore, the findings will likely support the development of interventions and health policies aimed at the increasing problem of dyslipidemia in this population. The study thus makes a relevant contribution to the fields of cardiovascular medicine and epidemiology. Additionally, the KNHANES questionnaires are updated annually to incorporate changes in the real-life health circumstances of South Koreans. Therefore, KNHANES data have been used widely in health-related studies and have provided meaningful insights to inform health policy development in South Korea.

Conclusions

The findings of previous studies suggest an association of an irregular work schedule, particularly nighttime work, with an altered lipid profile. Accordingly, in this study, we examine the association between night work and dyslipidemia in a nationally representative sample of South Korean adults aged ≥30 years who participated in the KNHANES 2013–2016. In the overall analysis, we found a significant association of night work with dyslipidemia only among male workers. Additionally, subgroup analyses of male workers who reported skipping meals or receiving < 7 h of sleep per night revealed associations of night work with dyslipidemia. Among female participants, a subgroup analysis of white-collar workers found that those who worked at night faced higher risk of dyslipidemia, compared to their day working counterparts.

However, our study was unable to determine a causal relationship between the onset of dyslipidemia and night work, and further investigations are needed to validate the findings of our study. Given the increasing prevalence of dyslipidemia in South Korea and the association of this condition with cardiovascular disease, we also suggest the development of future interventions intended to alleviate dyslipidemia among night workers and ease the burden of CVD in South Korea.