Poor Sleep Quality and Unhealthy Blood Fats Are Unrelated in Underground Miners, but Night Shifts Increase High Cholesterol Risk

This cross-sectional study recruited male underground miners from the Kailuan Group in Tangshan, China. The participants underwent routine occupational health examinations. Inclusion criteria were male miners with at least 1 year of consistent work history in their current schedule. Exclusion criteria included current use of lipid-lowering medications, clinically diagnosed hepatic or renal failure, or incomplete data on sleep or work schedules. The final analytical sample consisted of 921 participants. The study protocol adhered to the principles of the Declaration of Helsinki and was approved by the Ethics Committee of our hospital.

Work schedules were retrieved from administrative records and verified by participant interviews. Participants were dichotomized into two groups based on the timing of their shifts to ensure distinct circadian phenotypes and avoid exposure misclassification (19):

1. Night Shift Group (n =326): Workers performing duties during the biological night, defined here as shifts covering the interval from 0:00 to 7:59.

2. Day/Morning Shift Group (n =595): Workers performing duties exclusively between 8:00 and 23:59. This binary classification emphasizes exposure to light-at-night and nocturnal activity, key disruptors of the circadian system (20).

Job tenure (years in current shift schedule) and average weekly work hours were also recorded to control for cumulative exposure and workload intensity.

Subjective sleep quality was assessed using the Pittsburgh Sleep Quality Index (PSQI) (21), a widely validated 19-item questionnaire evaluating sleep quality over the preceding month. The PSQI generates a global score ranging from 0 to 21, aggregating seven components: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication, and daytime dysfunction. Consistent with clinical standards, a global PSQI score >5 was categorized as “Poor Sleep Quality,” while a score ≤ 5 was categorized as “Good Sleep Quality.”

Potential confounders were selected a priori based on literature review (22). ^**Demographic variables^** included age (years) and education level (years of schooling). ^**Lifestyle factors^** included cigarette smoking (Current smoker: Yes/No) and alcohol consumption (Current drinker: Yes/No). Occupational covariates included job tenure and weekly work hours. BMI was measured but treated as a potential mediator rather than a primary confounder in the main models, though it was explored in sensitivity analyses.

Venous blood samples were collected in the morning (7:00–9:00 a.m.) after an overnight fast of at least 8 h relative to the worker's biological day. Serum Total Cholesterol (TC), Triglycerides (TG), High-Density Lipoprotein Cholesterol (HDL-C), and Low-Density Lipoprotein Cholesterol (LDL-C) were measured using an auto-analyzer (Hitachi 747; Hitachi, Tokyo, Japan) with enzymatic methods. Dyslipidemia was defined according to the 2016 Chinese Guidelines for the Management of Dyslipidemia in Adults (23).

Data analyses were performed using SPSS software (Version 26.0, IBM Corp, Armonk, NY, USA). Continuous variables are presented as mean ± standard deviation (SD), and categorical variables as frequencies and percentages. Differences between the Night Shift and Day Shift groups were assessed using Student's t-test for continuous variables and Chi-square (χ2) tests for categorical variables. Pearson correlation coefficients were calculated to assess univariate associations between lipid biomarkers and risk factors (Age, Education, Smoking, Alcohol, PSQI). To test the independent effect of shift work, General Linear Models (GLM) in the form of Analysis of Covariance (ANCOVA) were employed, adjusting for age, smoking, drinking, and occupational tenure. Finally, to explore the interaction between shift work and sleep, participants were stratified into four subgroups (e.g., Night Shift + Good Sleep), and differences in TC were compared. A two-sided P-value < 0.05 was considered statistically significant. We also calculated the E-value to assess the robustness of the observed associations to potential unmeasured confounding (24).

The baseline characteristics of the 921 participants are presented in Table 1. The mean age was 38.36 ± 1.00 years. Night shift workers accounted for 35.4% (n = 326) of the cohort. While age, BMI, smoking, and drinking habits were comparable between groups, night shift workers had significantly lower education levels compared to day workers (12.11 vs. 12.78 years, P < 0.001). Notably, night shift work was strongly associated with poorer sleep quality; the mean PSQI score was significantly higher in the night shift group (4.70 ± 3.06) compared to the day shift group (3.59 ± 2.71, P < 0.001). Accordingly, the prevalence of poor sleep (PSQI > 5) was nearly double in the night shift group (34.96% vs. 19.16%, P < 0.001). Additional analysis of occupational factors showed no significant difference in weekly work hours between groups (Supplementary Table S1), minimizing the likelihood that workload intensity confounded the results.

Correlation analysis was performed to identify univariate associations between potential risk factors and lipid profiles (Table 2). Surprisingly, the PSQI total score showed virtually no correlation with TC (r = −0.024), TG (r = −0.015), HDL (r = −0.027), or LDL (r = −0.041), indicating that perceived sleep quality is not a linear predictor of lipid levels in this cohort. Age showed a weak positive correlation with HDL (r = 0.041), while smoking and alcohol consumption showed weak associations with HDL and TC.

Despite the lack of difference in overall dyslipidemia prevalence, quantitative analysis revealed specific lipid alterations. As shown in Table 3, night shift workers exhibited significantly higher levels of Total Cholesterol (TC) compared to day workers (5.36 vs. 5.09 mmol/L; t = 4.262, P = 0.039). This association remained significant after adjusting for age, smoking, and drinking status. No statistically significant differences were observed for TG, HDL, or LDL between the two shift groups (P > 0.05). This specific elevation in TC is visualized in Figure 1. The E-value for the association between night shift and TC was 1.58, indicating that an unmeasured confounder (e.g., diet) would need to be associated with both shift work and TC by a risk ratio of 1.58 to explain away the observed effect.

To further test whether the effect of night work was mediated by poor sleep, we stratified the population into four groups based on combinations of shift status and sleep quality (Table 4). Strikingly, the highest mean TC levels were observed in the “Night Shift + Good Sleep” group (5.51 ± 3.43 mmol/L). This level was significantly higher than that of the “Day Shift + Good Sleep” group (4.96 ± 1.06 mmol/L, P = 0.039). Conversely, when analyzing sleep quality in isolation (ignoring shift work), “Good Sleepers” (n = 693) had a mean TC of 5.24 ± 2.11 mmol/L compared to 5.02 ± 1.04 mmol/L in “Poor Sleepers” (n = 228), a difference that was not statistically significant (P = 0.153). This interaction is illustrated in Figure 2, highlighting that good subjective sleep does not mitigate the hypercholesterolemic effect of night work.

While shift work was strongly correlated with poor sleep scores (Table 1), consistent with extensive literature on hospital night workers (27), our correlation analysis failed to link Pittsburgh Sleep Quality Index (PSQI) scores with any lipid parameter. Furthermore, our stratified analysis (Table 4) revealed a paradoxical finding: the subgroup with the highest cholesterol burden was not the sleep-deprived night workers, but rather the night workers who reported good sleep according to the Good Sleeper Scale-15 (28). This aligns with emerging evidence from controlled laboratory studies suggesting that “circadian misalignment” and “sleep loss” have distinct metabolic consequences (29). Specifically, Leproult et al. (30) demonstrated that circadian misalignment reduces insulin sensitivity and increases inflammation independently of sleep loss.

We propose that the “Good Sleeper” night workers may represent a phenotype of workers who successfully maintain sleep duration (hence the low PSQI) but fail to resynchronize their peripheral biological clocks. The central circadian clock in the suprachiasmatic nucleus (SCN) adapts slowly to night shifts (31), while peripheral clocks in the liver (which regulate cholesterol synthesis via HMG-CoA reductase) are rapidly reset by food intake (32). When night workers eat during the biological night (chrononutrition misalignment), their liver clocks desynchronize from their central SCN clock (32). This internal desynchrony has been shown to upregulate hepatic lipogenesis and inhibit bile acid synthesis, leading to elevated serum total cholesterol, regardless of how well the individual sleeps during the day (15, 33). Therefore, subjective sleep quality may offer a false sense of security regarding metabolic health in shift workers.

Interestingly, while Total Cholesterol (TC) was significantly elevated, we did not observe significant differences in Triglycerides (TG) or HDL-C between shift groups. This contrasts with meta-analyses that report broad dyslipidemia in shift workers (34, 35). The discrepancy may be attributed to the specific occupational nature of mining. Unlike sedentary office shift work, underground mining involves heavy physical labor. Physical activity is a potent inducer of lipoprotein lipase (LPL) activity, which clears triglycerides and raises HDL (36, 37). It is plausible that the high physical demand of mining exerts a “protective” effect on the TG/HDL axis, masking the deleterious effects of shift work (34, 36). However, Total Cholesterol, which is more heavily influenced by endogenous hepatic synthesis than by acute energy expenditure, remains elevated, highlighting it as a more sensitive marker of circadian disruption in physically active cohorts.

Current occupational health guidelines heavily prioritize “sleep hygiene” education for shift workers (38). While essential for cognitive safety, our results imply that optimizing sleep alone is insufficient to mitigate cardiovascular risk. Future research should prioritize longitudinal designs utilizing objective circadian markers (e.g., actigraphy, salivary melatonin, continuous glucose monitoring) to validate these findings and overcome the limitations of cross-sectional design. Furthermore, interventional studies exploring chrononutrition strategies, such as time-restricted eating (39), are warranted to determine if aligning food intake with circadian rhythms can mitigate lipid risks in shift workers. Workplace policies should consider regulating meal timing during night shifts rather than focusing solely on sleep duration (40).

The original contributions presented in the study are included in the article/, further inquiries can be directed to the corresponding authors. Supplementary material