Abstract
Background
Lipid metabolism is influenced by multiple factors, and its association with sleep duration may vary across lipid fractions. Lipid levels and their genetic susceptibility have been associated with sleep duration; however, these associations are heterogeneous, and it remains unclear whether specific lipid components exhibit more consistent associations with sleep duration.
Methods
This study included 410,493 participants from the UK Biobank. Multivariable linear regression (MLR) was used to assess the associations of triglycerides (TG), total cholesterol (TC), and low-density lipoprotein (LDL) with sleep duration. Polygenic risk scores (PRSs) were incorporated to evaluate genetic susceptibility. Interaction terms between lipid levels and PRSs were included in multivariable models. Cox proportional hazards models were used to examine the associations of lipid levels and PRSs with sleep disorders. Joint exposure analyses of lipid levels and PRSs were performed, and interaction effects were assessed. Restricted cubic spline models were further applied to evaluate potential non-linear associations. Proteomic analyses were conducted to explore proteomic features associated with both lipid profile and sleep duration.
Results
MLR analyses showed that TG levels were positively associated with sleep duration (β = 0.027, P < 0.001), whereas TC levels were inversely associated with sleep duration (β = -0.005, P = 0.005). No significant association was observed for LDL. In contrast, PRSs for TG, TC, and LDL were consistently inversely associated with sleep duration. A significant interaction was observed between TG and their PRS (β = -0.009, p < 0.001). TG and PRS-TG were associated with higher risks of sleep disorders (6.9% and 3.2% per unit increase), with a significant interaction (p = 0.014). Elevated risks were observed in high TG groups, while TC and LDL were not significant. Both lipid traits and sleep duration were mainly enriched in immune and inflammatory pathways.
Conclusions
Triglycerides showed the most consistent associations with sleep duration, and TG-related genetic susceptibility may influence TG levels and their associations with sleep duration and sleep disorders.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12944-026-02998-z.
Keywords: Lipid, Triglycerides, Sleep duration, Sleep disorders
Introduction
Recent estimates indicate that approximately 16.2% of adults worldwide have insomnia, affecting more than 850 million individuals. In the certain population, the prevalence of sleep problems or poor sleep quality is about 36.7%-40%, rising to nearly 50% among older adults, placing a substantial burden on individuals and public health [1, 2]. Sleep duration is a quantifiable sleep phenotype and is associated with genetic, metabolic, immune, and proteomic features [3, 4]. Abnormal sleep duration is associated with common conditions including obesity, type 2 diabetes, cardiovascular disease, stroke, and dementia [5–7], while these conditions, particularly obesity, cardiovascular disease, and neurodegeneration, can in turn worsen sleep disturbances [8, 9]. In addition, sleep is influenced by several modifiable behavioral and metabolic factors [10].
Lipid metabolism is linked not only to sleep duration but also to a range of sleep-related problems, particularly obstructive sleep apnea (OSA) and poor sleep quality [11, 12]. OSA severity is closely associated with an adverse lipid profile, most consistently characterized by elevated triglycerides (TG), whereas changes in total cholesterol (TC) and low-density lipoprotein (LDL) are less consistent in severe cases [13, 14]. Lipid levels are closely associated with sleep duration, although the patterns differ across lipid fractions. TG exhibit a U-shaped relationship with sleep duration, with higher levels observed in both short and long sleepers, and this association shows sex differences [15, 16]. For TC and LDL, the associations with sleep duration are relatively inconsistent; some studies have reported lower levels with longer sleep duration, but these findings are not consistent across populations [17–19].
Evidence suggests that the relationship between sleep duration and lipid metabolism is mediated by multiple interconnected pathways, including neuroendocrine regulation, insulin resistance, behavioral factors, and systemic inflammation [20, 21]. However, whether specific lipid fractions show consistent associations across both sleep duration and clinically defined sleep disorders remains unclear.
In addition, lipid phenotypes are influenced by multiple factors, including age, sex, and the use of lipid-lowering medications, which may introduce residual confounding and reverse causation [22, 23]. To mitigate these biases, we incorporated polygenic risk scores (PRSs) as stable indicators of lifelong genetic susceptibility. This study aimed to examine the associations of lipid profile and lipid related genetic susceptibility with sleep duration, evaluate incident sleep disorders as a secondary clinical outcome, and investigate proteomic signatures associated with both lipid traits and sleep duration.
Methods
Participants
This prospective cohort study utilized data from the UK Biobank, a large population-based cohort comprising over 500,000 participants recruited from 22 assessment centers across the United Kingdom between 2006 and 2010. Participants were included if data on plasma lipids, lipid-related PRSs, self-reported sleep duration, and covariates were available; those with missing data were excluded.
Sleep and lipid measurements
Sleep duration was obtained from self-reported data, based on participants’ responses to the question regarding the average number of hours slept within a 24-hour period. According to established cutoffs, sleep duration was categorized as short sleep (≤ 6 h), normal sleep (6–8 h), and long sleep (> 8 h).
Sleep disorders were defined using the International Classification of Diseases, Tenth Revision (ICD-10) code G47, derived from linked hospital inpatient records in the UK Biobank. This category was treated as a composite sleep disorder outcome, including disorders of initiating and maintaining sleep/insomnias, disorders of excessive somnolence/hypersomnias, disorders of the sleep-wake schedule, sleep apnoea, narcolepsy and cataplexy, other sleep disorders, and unspecified sleep disorder.
Serum lipid concentrations were measured using centralized and standardized laboratory procedures in the UK Biobank. Blood samples were processed according to uniform protocols and analyzed on automated biochemical analyzers for the quantification of TC, LDL, and TG.
Hyperlipidemia was defined according to established thresholds as TC ≥ 5.2 mmol/L, LDL ≥ 3.4 mmol/L, or TG ≥ 1.7 mmol/L [24].
Polygenic risk score
Lipid-related PRSs were constructed using summary statistics from external genome-wide association studies independent of the UK Biobank. Summary statistics were harmonized to the GRCh37 reference genome and integrated via inverse-variance-weighted meta-analysis. Imputation of summary statistics was conducted using the 100,000 Genomes Project as the reference panel. Variant effect sizes were estimated within a Bayesian framework and used as weights for PRS construction. Quality control retained variants with high imputation quality (INFO > 0.8), common allele frequency (MAF > 0.05), and no substantial deviation from Hardy-Weinberg equilibrium (p > 1 × 10− 10), while variants with potential cross-ancestry discrepancies were excluded. PRSs were standardized to enable interpretation per standard deviation increase.
In multivariable linear regression models, each lipid PRS was significantly associated with its corresponding circulating lipid level. Specifically, PRS-TG was positively associated with TG levels (β = 0.318, p < 0.001), PRS-TC with TC (β = 0.347, p < 0.001), and PRS-LDL with LDL (β = 0.365, p < 0.001) (sFigure 1). For categorical analyses, PRSs were categorized into tertiles, with the highest tertile defined as high genetic risk and the lower two tertiles as the reference group [25, 26].
Covariates
Covariates were selected a priori based on clinical relevance. Demographic and socioeconomic factors included age, sex, Townsend deprivation index, and educational attainment. Clinical and anthropometric variables comprised body mass index, systolic blood pressure, diastolic blood pressure, and C-reactive protein levels. Medication use was additionally accounted for by including indicators of antihypertensive, lipid-lowering, and glucose-lowering therapies. Stroke, anxiety, and depression were ascertained using International Classification of Diseases, 10th Revision (ICD-10) codes and linked medical records. Field IDs for these covariates are provided in the supplementary file.
Proteomic analyses
Proteomic analyses began with quality control, excluding proteins with missingness > 50%, leaving 2,921 proteins for downstream analyses. Each protein was analyzed using complete-case samples and was inverse-rank normalized to reduce distributional skewness. Two-step linear regression analyses were then performed. First, standardized lipid levels were modeled as exposures and protein levels as outcomes to assess lipid-protein associations. Second, standardized sleep duration was modeled as the outcome, with each protein as the main predictor while adjusting for the corresponding lipid level and covariates, to examine protein-sleep duration associations after accounting for measured lipid levels. Multiple testing was controlled using the false discovery rate (FDR), with FDR_p < 0.05 considered statistically significant.
Proteins that were significant in both the lipid-protein and protein-sleep analyses were defined as overlapping proteins and were carried forward for downstream functional enrichment and protein-protein interaction analyses.
For functional annotation, overlapping proteins were subjected to Gene Ontology (GO) biological process and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses to identify relevant biological processes and pathways. A protein-protein interaction (PPI) network was constructed based on the STRING database, and network topology was analyzed to identify potential hub proteins.
To assess whether the associations between network-central proteins and sleep duration were mainly explained by lipid levels and lipid related genetic susceptibility, sensitivity analyses were performed for the top 20 hub proteins in each lipid-specific protein-protein interaction network. For each hub protein, three models were fitted: Model 1 adjusted for covariates; Model 2 adjusted for covariates and the corresponding lipid trait; and Model 3 adjusted for covariates, the corresponding lipid trait, and the corresponding lipid-related PRS.
Statistical analysis
Using means ± standard deviations (SD) for continuous variables and counts (percentages) for categorical variables.
Multivariable linear regression was used to assess the association between lipid and sleep duration. Joint exposure was evaluated by classifying lipid levels and PRSs into categories and combining them into four groups, and differences in sleep duration across these groups were compared. Subgroup analyses were performed according to age, sex, use of lipid-lowering medication, and stroke status to assess robustness and heterogeneity. Interaction analyses were also conducted by including cross-product terms to test whether lipid levels and lipid PRSs jointly modified sleep duration.
Cox proportional hazards models were used to evaluate the associations of lipid levels and lipid-related PRSs with the risk of sleep disorders. To reduce potential reverse causation, participants diagnosed with sleep disorders before recruitment or within the first year after recruitment were excluded, resulting in the exclusion of 8,671 participants. The proportional hazards assumption was assessed using proportional hazards (PH) tests. For joint exposure analyses, lipid levels were categorized into normal and elevated groups based on clinical thresholds, and PRSs into low and high groups, with participants classified into four combined exposure groups. The group with normal lipid levels and low PRS was used as the reference. Restricted cubic spline (RCS) models were applied to investigate potential non-linear relationships.
Sensitivity analyses were further performed to evaluate the potential influence of neurological and psychiatric conditions on sleep duration and sleep disorders. The main analyses were repeated after separately excluding participants with stroke, depression, or anxiety.
All statistical analyses were conducted using R version 4.5.2, with statistical significance set at p < 0.05 for all analyses.
Results
Comparisons of lipid across sleep groups
A total of 410,493 participants were included in the analysis, with a mean age of 56.50 ± 8.08 years, of whom 54.1% were female.
Significant differences in lipid levels and lipid PRSs were observed across sleep duration groups (p < 0.001) (Table 1). TG levels were lowest in the normal sleep group, whereas TC and LDL levels were lowest in the long sleep group. In addition, all three lipid-related PRSs were lowest among individuals with normal sleep duration.
Table 1.
Baseline characteristics of participants
| Overall | ≤ 6 hours | 6–8 hours | > 8 hours | p | |
|---|---|---|---|---|---|
| Number | 410,493 | 100,152 | 279,252 | 31,089 | |
| Age (years) | 56.50 (8.08) | 56.33 (7.86) | 56.34 (8.14) | 58.52 (8.02) | < 0.001 |
| TDI | -1.39 (3.03) | -0.98 (3.22) | -1.56 (2.92) | -1.18 (3.15) | < 0.001 |
| Sex (%) | < 0.001 | ||||
| Female | 222,168 (54.1) | 53,201 (53.1) | 151,423 (54.2) | 17,544 (56.4) | |
| Male | 188,325 (45.9) | 46,951 (46.9) | 127,829 (45.8) | 13,545 (43.6) | |
| Drinking (%) | < 0.001 | ||||
| Current | 379,380 (92.4) | 90,885 (90.7) | 260,655 (93.3) | 27,840 (89.5) | |
| Never | 16,899 (4.1) | 4928 (4.9) | 10,304 (3.7) | 1667 (5.4) | |
| Previous | 14,214 (3.5) | 4339 (4.3) | 8293 (3.0) | 1582 (5.1) | |
| Smoking (%) | < 0.001 | ||||
| No | 163,871 (39.9) | 39,069 (39.0) | 112,757 (40.4) | 12,045 (38.7) | |
| Yes | 246,622 (60.1) | 61,083 (61.0) | 166,495 (59.6) | 19,044 (61.3) | |
| Education (%) | < 0.001 | ||||
|
Advanced level qualifications and above |
181,912 (44.3) | 39,736 (39.7) | 131,390 (47.1) | 10,786 (34.7) | |
|
Ordinary level qualifications and below |
139,258 (33.9) | 35,642 (35.6) | 92,985 (33.3) | 10,631 (34.2) | |
| Other | 89,323 (21.8) | 24,774 (24.7) | 54,877 (19.7) | 9672 (31.1) | |
| Use of antihypertensive medications (%) | < 0.001 | ||||
| No | 326,125 (79.4) | 78,128 (78.0) | 225,592 (80.8) | 22,405 (72.1) | |
| Yes | 84,368 (20.6) | 22,024 (22.0) | 53,660 (19.2) | 8684 (27.9) | |
| Use of lipid-lowering medications (%) | < 0.001 | ||||
| No | 339,913 (82.8) | 81,887 (81.8) | 234,618 (84.0) | 23,408 (75.3) | |
| Yes | 70,580 (17.2) | 18,265 (18.2) | 44,634 (16.0) | 7681 (24.7) | |
| Use of glucose-lowering medications (%) | < 0.001 | ||||
| No | 406,126 (98.9) | 98,934 (98.8) | 276,706 (99.1) | 30,486 (98.1) | |
| Yes | 4367 (1.1) | 1218 (1.2) | 2546 (0.9) | 603 (1.9) | |
| Stroke (%) | < 0.001 | ||||
| No | 393,888 (96.0) | 95,727 (95.6) | 269,038 (96.3) | 29,123 (93.7) | |
| Yes | 16,605 (4.0) | 4425 (4.4) | 10,214 (3.7) | 1966 (6.3) | |
| Depression (%) | < 0.001 | ||||
| No | 356,840 (86.9) | 84,786 (84.7) | 247,476 (88.6) | 24,578 (79.1) | |
| Yes | 53,653 (13.1) | 15,366 (15.3) | 31,776 (11.4) | 6511 (20.9) | |
| Anxiety (%) | < 0.001 | ||||
| No | 371,535 (90.5) | 89,286 (89.2) | 255,205 (91.4) | 27,044 (87.0) | |
| Yes | 38,958 (9.5) | 10,866 (10.8) | 24,047 (8.6) | 4045 (13.0) | |
| DBP (mmHg) | 82.19 (10.68) | 82.43 (10.73) | 82.08 (10.64) | 82.47 (10.79) | < 0.001 |
| SBP (mmHg) | 139.67 (19.64) | 139.65 (19.33) | 139.53 (19.68) | 141.02 (20.26) | < 0.001 |
| BMI (Kg/m2) | 27.37 (4.73) | 27.92 (5.04) | 27.09 (4.55) | 28.04 (5.09) | < 0.001 |
| CRP (mg/L) | 2.56 (4.29) | 2.73 (4.47) | 2.43 (4.12) | 3.17 (5.10) | < 0.001 |
| TG (mmol/L) | 1.74 (1.02) | 1.77 (1.04) | 1.71 (0.99) | 1.88 (1.09) | < 0.001 |
| TC (mmol/L) | 5.70 (1.14) | 5.69 (1.14) | 5.70 (1.13) | 5.65 (1.22) | < 0.001 |
| LDL (mmol/L) | 3.56 (0.87) | 3.56 (0.87) | 3.56 (0.86) | 3.53 (0.92) | < 0.001 |
| PRS-TG | -0.31 (1.01) | -0.29 (1.01) | -0.32 (1.00) | -0.29 (1.00) | < 0.001 |
| PRS-TC | -0.20 (1.00) | -0.18 (1.01) | -0.20 (1.00) | -0.19 (1.00) | < 0.001 |
| PRS-LDL | -0.18 (1.05) | -0.17 (1.05) | -0.19 (1.05) | -0.18 (1.05) | < 0.001 |
Data are mean (SD), n (%)
TDI Townsend deprivation index, BMI Body mass index, SBP Systolic blood pressure, DBP Diastolic blood pressure, CRP C-reactive protein, LDL Low-density lipoprotein cholesterol, TG Triglycerides, TC Total cholesterol, PRS Polygenic risk score
Associations of lipid levels and PRSs with sleep duration
Multivariable linear regression analyses showed that TG levels were positively associated with sleep duration (β = 0.027, p < 0.001), whereas TC levels were inversely associated with sleep duration (β = -0.005, p = 0.005). No significant association was observed between LDL levels and sleep duration. In addition, all three lipid-related PRSs were inversely associated with sleep duration (Fig. 1A-F).
Fig. 1.
Associations of lipid levels and polygenic risk scores with sleep duration. A–C Scatter plots showing the associations of triglycerides, total cholesterol, and low-density lipoprotein with sleep duration. D–F Associations of polygenic risk scores (PRS-TG, PRS-TC, and PRS-LDL) with sleep duration. G–I Interaction effects between lipid levels and corresponding polygenic risk scores on sleep duration. Solid lines represent estimated effects, and shaded areas indicate 95% confidence intervals
Using normal sleep duration as the reference, both short and long sleep durations were positively associated with TC and LDL levels, regardless of covariate adjustment. In contrast, TG were positively associated only with long sleep duration. After covariate adjustment, all three lipid-related PRSs were positively associated with short sleep duration (Table 2).
Table 2.
Associations of lipid levels and polygenic risk scores with sleep duration categories
| Exposure | Group | Unadjusted | Adjusted | ||
|---|---|---|---|---|---|
| OR(95%CI) | p | OR(95%CI) | p | ||
| Triglyceride | 6–8 hours | Reference | Reference | ||
| ≤ 6 hours | 1.056 (1.049, 1.064) | < 0.001 | 0.995 (0.988, 1.003) | 0.238 | |
| > 8 hours | 1.164 (1.152, 1.176) | < 0.001 | 1.102 (1.089, 1.115) | < 0.001 | |
|
Total cholesterol |
6–8 hours | Reference | Reference | ||
| ≤ 6 hours | 0.986 (0.979, 0.993) | < 0.001 | 1.024 (1.015, 1.032) | < 0.001 | |
| > 8 hours | 0.956 (0.945, 0.968) | < 0.001 | 1.017 (1.004, 1.031) | 0.011 | |
|
Low-density lipoprotein cholesterol |
6–8 hours | Reference | Reference | ||
| ≤ 6 hours | 0.993 (0.986, 1.000) | 0.059 | 1.013 (1.005, 1.021) | 0.002 | |
| > 8 hours | 0.963 (0.951, 0.974) | < 0.001 | 1.019 (1.005, 1.032) | 0.006 | |
| PRS-TG | 6–8 hours | Reference | Reference | ||
| ≤ 6 hours | 1.034 (1.027, 1.042) | < 0.001 | 1.018 (1.010, 1.025) | < 0.001 | |
| > 8 hours | 1.037 (1.024, 1.049) | < 0.001 | 1.009 (0.997, 1.022) | 0.125 | |
| PRS-TC | 6–8 hours | Reference | Reference | ||
| ≤ 6 hours | 1.020 (1.013, 1.028) | < 0.001 | 1.021 (1.013, 1.028) | < 0.001 | |
| > 8 hours | 1.013 (1.001, 1.025) | 0.030 | 1.004 (0.992, 1.016) | 0.487 | |
| PRS-LDL | 6–8 hours | Reference | Reference | ||
| ≤ 6 hours | 1.017 (1.010, 1.025) | < 0.001 | 1.017 (1.009, 1.024) | < 0.001 | |
| > 8 hours | 1.007 (0.995, 1.019) | 0.241 | 0.998 (0.987, 1.011) | 0.801 | |
Sleep duration was categorized as ≤ 6 h (short sleep), 6–8 h (reference), and > 8 h (long sleep). Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated using logistic regression models. The reference group was participants with 6–8 h of sleep. Models were adjusted for age, sex, Townsend deprivation index, educational attainment, body mass index, systolic blood pressure, diastolic blood pressure, C-reactive protein, use of antihypertensive, lipid-lowering, and glucose-lowering medications, as well as stroke, anxiety, and depression
Interaction analyses revealed a significant interaction between TG and their PRS (β = -0.009, p < 0.001). Specifically, the positive association between TG and sleep duration was attenuated with increasing PRS-TG levels. In contrast, no significant interactions were observed between TC or LDL and their corresponding PRSs (Fig. 1G-I; sTable 1).
Subgroup analyses
Across subgroups stratified by age, sex, use of lipid-lowering medication, and stroke status, TG levels were consistently positively associated with sleep duration. In age-stratified analyses, LDL was positively associated with sleep duration across all age groups, whereas TC was associated with sleep duration only among participants aged < 65 years. PRS-TC and PRS-LDL showed consistent inverse associations, while no consistent significant association was observed for PRS-TG (sTable 2).
In sex-stratified analyses, both LDL and TC were positively associated with sleep duration in males and females, whereas PRS-TC and PRS-LDL were inversely associated with sleep duration (sTable 3).
In analyses stratified by lipid-lowering medication use, TG and LDL were inversely associated with sleep duration among participants using lipid-lowering medications, whereas LDL showed a positive association with sleep duration among those not using such medications. Corresponding PRSs were inversely associated with sleep duration regardless of medication use (sTable 4).
In analyses stratified by stroke status, the overall direction of associations remained largely unchanged; TG was positively associated with sleep duration, while TC and LDL were inversely associated. No significant associations of PRSs were observed among participants with stroke (sTable 5).
Associations of lipid levels and PRSs with sleep disorders
The mean follow-up time was 15.7 years, during which 8,934 incident sleep disorder cases were identified. PH tests showed no major violation of the proportional hazards assumption, except for PRS-LDL (sTable 6). In both unadjusted and adjusted models, only TG were significantly associated with an increased risk of sleep disorders as levels increased. Each one-unit increase in PRS-TG and PRS-TC was associated with a 3.2% (HR 1.032[1.011,1.054], p = 0.003) and 2.9% (HR 1.029[1.008,1.051], p = 0.008) higher risk, respectively (sTable 7). RCS analyses showed that the risk of sleep disorders increased with higher TG levels, while both PRS-TC and PRS-TG exhibited linear positive associations with sleep disorder risk (Fig. 2). Interaction analyses further revealed a significant interaction between TG and PRS-TG. (p = 0.010) (sTable 8).
Fig. 2.
Restricted cubic spline analyses of lipid levels and lipid-related polygenic risk scores in relation to sleep disorder risk. P-overall and P-nonlinear values are shown in each panel. Models were adjusted for age, sex, Townsend deprivation index, educational attainment, body mass index, systolic and diastolic blood pressure, C-reactive protein, use of antihypertensive, lipid-lowering, and glucose-lowering medications, as well as history of stroke, anxiety, and depression
Joint associations of lipid levels and PRSs with sleep outcomes
Participants were categorized according to combinations of lipid levels and PRSs, with the group characterized by normal lipid levels and normal PRS serving as the reference.
For TG, the group with normal TG but high PRS (β = -0.020, p < 0.001) was inversely associated with sleep duration, whereas both the high TG + normal PRS group (β = 0.051, p < 0.001) and the high TG + high PRS group (β = 0.027, p < 0.001) were positively associated with sleep duration. For TC, both the high TC with normal PRS group (β = -0.012, p = 0.005) and the high TC + high PRS group (β = -0.024, p < 0.001) showed inverse associations with sleep duration. For LDL, an inverse association with sleep duration was observed only in the group with both elevated LDL and high PRS (β = -0.100, p = 0.024), whereas other combinations were not statistically significant (Table 3).
Table 3.
Joint associations of lipid levels and polygenic risk scores with sleep duration
| Group | Unadjusted | Adjusted | ||
|---|---|---|---|---|
| β | p | β | p | |
| Normal TG+Normal PRS-TG | Reference | Reference | ||
| Normal TG+High PRS-TG | -0.020 | < 0.001 | -0.020 | < 0.001 |
| High TG+Normal PRS-TG | 0.031 | < 0.001 | 0.051 | < 0.001 |
| High TG+High PRS-TG | 0.012 | 0.006 | 0.027 | < 0.001 |
| Normal TC+Normal PRS-TC | Reference | Reference | ||
| Normal TC+High PRS-TC | 0.014 | 0.024 | -0.002 | 0.723 |
| High TC+Normal PRS-TC | -0.002 | 0.574 | -0.012 | 0.005 |
| High TC+High PRS-TC | -0.019 | < 0.001 | -0.024 | < 0.001 |
| Normal LDL+Normal PRS-LDL | Reference | Reference | ||
| Normal LDL+High PRS-LDL | 0.015 | 0.008 | -0.001 | 0.854 |
| High LDL+Normal PRS-LDL | 0.006 | 0.096 | 0.007 | 0.080 |
| High LDL+High PRS-LDL | -0.014 | < 0.001 | -0.100 | 0.024 |
Participants were categorized into four groups according to lipid levels (normal vs. high) and polygenic risk scores (PRS; normal vs. high), with the group characterized by normal lipid levels and normal PRS as the reference. β coefficients represent the change in sleep duration compared with the reference group. Models were adjusted for age, sex, Townsend deprivation index, educational attainment, body mass index, systolic blood pressure, diastolic blood pressure, C-reactive protein, use of antihypertensive, lipid-lowering, and glucose-lowering medications, as well as stroke, anxiety, and depression
For TG, both the high TG + normal PRS group (HR 1.092[1.009,1.183], p = 0.029) and the high TG+high PRS group (HR = 1.155[1.090,1.224], p < 0.001) were significantly associated with an increased risk of sleep disorders. No significant associations were observed for TC or LDL across all joint exposure groups (Table 4).
Table 4.
Joint associations of lipid levels and polygenic risk scores with sleep disorders risk
| Group | Unadjusted | Adjusted | ||
|---|---|---|---|---|
| HR(95%CI) | p | HR(95%CI) | p | |
| Normal TG+Normal PRS-TG | Reference | Reference | ||
| Normal TG+High PRS-TG | 1.017(0.956,1.081) | 0.593 | 1.041(0.979,1.107) | 0.197 |
| High TG+Normal PRS-TG | 1.710(1.581,1.850) | < 0.001 | 1.092(1.009,1.183) | 0.029 |
| High TG+High PRS-TG | 1.733(1.637,1.835) | < 0.001 | 1.155(1.090,1.224) | < 0.001 |
| Normal TC+Normal PRS-TC | Reference | Reference | ||
| Normal TC+High PRS-TC | 1.133(1.062,1.208) | < 0.001 | 1.012(0.947,1.081) | 0.729 |
| High TC+Normal PRS-TC | 0.710(0.660,0.763) | < 0.001 | 0.929(0.861,1.001) | 0.054 |
| High TC+High PRS-TC | 0.757(0.714,0.803) | < 0.001 | 0.993(0.935,1.055) | 0.824 |
| Normal LDL+Normal PRS-LDL | Reference | Reference | ||
| Normal LDL+High PRS-LDL | 1.125(1.060,1.194) | < 0.001 | 1.023(0.962,1.088) | 0.464 |
| High LDL+Normal PRS-LDL | 0.918(0.853,0.987) | 0.021 | 1.043(0.967,1.124) | 0.273 |
| High LDL+High PRS-LDL | 0.842(0.795,0.892) | < 0.001 | 0.990(0.933,1.050) | 0.735 |
Participants were categorized into four groups according to lipid levels (normal vs. high) and polygenic risk scores (PRS; normal vs. high), with the group characterized by normal lipid levels and normal PRS as the reference. Models were adjusted for age, sex, Townsend deprivation index, educational attainment, body mass index, systolic blood pressure, diastolic blood pressure, C-reactive protein, use of antihypertensive, lipid-lowering, and glucose-lowering medications, as well as stroke, anxiety, and depression
Shared proteomic signatures of lipid traits and sleep duration
Proteomic analyses identified several proteins significantly associated with TG and sleep duration, including ABL1, ACRN, and AFM for TG, and ITGAV and ITGA11 for sleep duration (Fig. 3A-B). Functional enrichment analysis showed that these proteins were predominantly involved in immune and inflammatory processes, such as leukocyte proliferation, immune regulation, and cell adhesion (Fig. 3C). KEGG pathway analysis showed enrichment in cytokine-cytokine receptor interaction, complement and coagulation cascades, and the PI3K-Akt signaling pathway (Fig. 3D). PPI network analysis showed a highly interconnected network, with TNF, IL6, IL1B, EGFR, FN1, and CXCL8 identified as network-central proteins (Fig. 3E-F) The associations between TG-related top hub proteins and sleep duration were generally stable after adjustment for TG levels and TG-related PRS (sTable 9).
Fig. 3.
Proteomic characterization of proteins associated with triglycerides and sleep duration. A Volcano plot showing proteins associated with sleep duration. B Volcano plot showing proteins associated with triglycerides. Red dots indicate positive associations and blue dots indicate negative associations. C Gene Ontology biological process enrichment analysis of overlapping proteins. D KEGG pathway enrichment analysis of overlapping proteins. E Protein-protein interaction (PPI) network of overlapping proteins. Node colors represent associations with triglycerides, sleep duration, or both. F Identification of key proteins based on network topology, ranked by degree
For TC and LDL, proteomic analyses showed broadly similar enrichment patterns, mainly involving immune and inflammatory pathways; however, only a few hub proteins remained associated with sleep duration after further adjustment for the corresponding lipid trait and lipid-related PRS (sFigure 2–3; sTable 9).
Sensitivity analyses
In sensitivity analyses excluding participants with stroke, depression, or anxiety, TG remained positively associated with sleep duration (sTable 5, sTable 10). For incident sleep disorders, both continuous TG levels and high TG status remained associated with higher risk across the three sensitivity populations (sTable 11). TC and LDL did not show stable associations with incident sleep disorders after adjustment. Joint exposure analyses were generally consistent with the main findings (sTable 12).
Discussion
This study showed that TG and TC were positively associated with sleep duration, whereas PRSs for all three lipid traits were inversely associated. TG demonstrated consistent associations across subgroups, while associations for TC and LDL were mainly observed in individuals younger than 65 years, men, and those receiving lipid-lowering therapy. Only TG was significantly associated with sleep disorder risk, and this association was modified by genetic susceptibility. Joint analyses further showed that elevated TG was associated with increased risk regardless of PRS level. Proteins associated with both TG and sleep duration were enriched in cytokine signaling, complement and coagulation cascades, and the PI3K-Akt signaling pathway.
In the present study, plasma TG levels were positively associated with sleep duration, whereas PRS-TG showed an inverse association, suggesting that phenotypic TG and genetic susceptibility may have different roles in sleep regulation. The positive association between TG and sleep duration is more likely driven by reverse causation and confounding by health status. A longitudinal cross-lagged analysis in a Chinese population showed that the positive association between TG and subsequent sleep duration was mainly observed among older individuals or those with higher body mass index, indicating that this relationship is more likely to occur in populations with obesity [27]. At the same time, the association between sleep duration and dyslipidemia is typically non-linear or U-shaped rather than a simple linear relationship [28]. In addition, self-reported long sleep duration, apart from recall bias, does not necessarily indicate good sleep quality. Instead, it is often accompanied by a range of adverse health conditions, including underlying chronic diseases, reduced physical activity, increased fatigue, sleep fragmentation, and sleep-disordered breathing [7, 29]. For example, OSA is closely associated with elevated TG and reduced HDL levels, and the triglyceride-glucose index is significantly higher in patients with OSA. This suggests that some individuals with longer sleep duration may actually experience poor sleep quality, intermittent nocturnal hypoxia, and daytime fatigue, leading to an apparent prolongation of sleep duration [30]. These conditions may promote elevated TG levels through multiple metabolic mechanisms. Taking inflammation as an example, individuals with longer sleep duration have higher levels of C-reactive protein and interleukin-6, while IL-6 itself has sleep-promoting effects, suggesting that inflammation-related sleep prolongation and inflammation and insulin resistance-related increases in TG may occur simultaneously [31]. In addition, inflammation and sympathetic activation associated with sleep disorders may further exacerbate abnormalities in lipid metabolism, thereby contributing to an interaction among sleep disturbance, inflammation, and dyslipidemia [32].
PRS-TG may be inversely associated with sleep duration due to several underlying mechanisms. Individuals with higher genetic susceptibility to elevated TG may carry a greater predisposition to disrupted energy homeostasis and circadian rhythm imbalance, which are more likely to manifest as shorter sleep duration, delayed sleep timing, or unstable sleep patterns. Circadian misalignment and insufficient sleep have been shown to reduce insulin sensitivity, disrupt energy metabolism, and impair lipid metabolic regulation [33–36]. In addition, genetic effects on lipid traits are modified by sleep duration rather than acting independently. Loci associated with sleep duration are enriched in pathways related to neuronal development, synaptic transmission, and circadian regulation, indicating that sleep duration is primarily governed by central nervous system and biological clock mechanisms. Consistently, genome-wide association studies have demonstrated that sleep duration-related variants are highly expressed in brain tissues and are involved in neuronal signaling pathways, further supporting a central regulatory basis of sleep duration [4, 37].
Lipid-lowering medications, particularly statins, may modify the observed lipid–sleep association. Although recent high-level evidence suggests that statins have little overall effect on sleep at the population level, lipophilic statins can cross the blood–brain barrier, and subjective sleep complaints such as insomnia or nightmares have been reported in some individuals [38, 39]. Meta-analyses of randomized placebo-controlled trials have shown no significant effect of statins on total sleep duration or sleep efficiency, although some studies have reported modest changes in nocturnal awakenings or subjective sleep complaints [40]. At the same time, observational and pharmacovigilance studies indicate that some individuals may report insomnia, nightmares, or reduced sleep duration, particularly with lipophilic statins [41]. Another important explanation is confounding by indication. Individuals receiving lipid-lowering therapy typically have higher cardiovascular risk, more severe metabolic abnormalities, and a greater burden of comorbidities, all of which are independently associated with sleep disturbances.
In individuals not using lipid-lowering medications, the positive association between LDL and sleep duration may be related to the physiological role of cholesterol in the central nervous system. Adequate cholesterol levels are essential for normal synaptic function and neuronal signaling, processes that are closely involved in sleep regulation [41]. In addition, brain cholesterol homeostasis contributes to membrane stability, synaptic transmission, and myelin function, all of which are relevant to sleep-wake regulation [42, 43].
Proteomic analyses indicated that although proteins associated with both lipid traits and sleep duration were commonly enriched in immune and inflammatory processes, the underlying mechanisms are not entirely consistent. Overlapping proteins related to triglycerides were mainly enriched in cytokine-cytokine receptor interactions, complement and coagulation cascades, and the PI3K-Akt signaling pathway, with TNF, IL6, and IL1B located at the core of the PPI network. These findings suggest that systemic inflammatory responses and immune activation may provide biological context for the association between TG and sleep duration [44]. Although proteins associated with both TC or LDL and sleep duration were enriched in immune and inflammatory processes, their functional profiles differed. TC-related overlapping proteins were more involved in immune effector regulation, lymphocyte and monocyte proliferation, and leukocyte adhesion. Previous studies have shown that lipids, particularly cholesterol, can directly regulate immune cell activation, differentiation, and proliferation, thereby shaping inflammatory phenotypes and maintaining immune homeostasis [45]. Moreover, dysregulated cholesterol metabolism has been linked to increased monocyte and neutrophil production and enhanced inflammatory responses, which are closely associated with chronic low-grade inflammation [46]. LDL-related proteins, beyond general inflammatory processes, were more prominently enriched in host-microbe interactions, phagocytosis, and wound healing pathways. Mechanistically, LDL and its modified forms can be recognized and internalized by macrophages, leading to activation of Toll-like receptor signaling and inflammasomes such as NLRP3, thereby amplifying inflammatory responses [46]. In addition, the complement system, as a key component of innate immunity, not only mediates pathogen clearance and phagocytosis but also acts as a critical link between lipid metabolism and immune responses in lipid-related disorders. These processes are closely involved in both tissue repair and the regulation of inflammation [47].
In this study, several strengths should be noted. First, by jointly analyzing lipid levels and PRSs, we were able to partially disentangle environmental influences from genetically determined susceptibility. Second, stratified analyses by lipid-lowering medication use revealed heterogeneity in the associations between lipids and sleep, indicating that these relationships are context-dependent.
Several limitations should be considered. First, sleep duration was based on self-reported data, which may be subject to recall bias and may not accurately capture sleep quality or objective sleep patterns, moreover, sleep duration alone cannot fully represent the multidimensional features of sleep health, such as sleep quality, timing, efficiency, and insomnia symptoms. Second, due to the observational design, residual confounding and reverse causation cannot be fully excluded despite comprehensive adjustments. Third, the biological mechanisms linking lipids, genetic susceptibility, and sleep regulation remain largely speculative, and further experimental and mechanistic studies are required to validate the proposed pathways. Finally, the predominantly White UK Biobank cohort may limit generalizability.
Conclusion
Among the lipid profiles examined, TG showed the most consistent associations with both sleep duration and incident sleep disorders, and these associations were modified by genetic susceptibility. These findings suggest that the relationship between lipid levels and sleep reflects not only current metabolic status but also underlying genetic background.
Supplementary Information
Acknowledgements
This research has been conducted using the UK Biobank Resource under Application Number 1029284. We are grateful to the UK Biobank participants for their invaluable contributions to this research resource.
Authors’ contributions
WQ-Z contributed to the conception and design of the work. ZH-Z drafted the manuscript. All authors were responsible for the acquisition, analysis and interpretation of data. Critical revision of the manuscript for important intellectual content was performed by all authors. All coauthors have reviewed and approved the contents of the final version of the manuscript.
Funding
This work was partially supported by the startup package and developmental funds of the First Affiliated Hospital of Nanchang University (#500021001, #500021002), National Natural Science Foundation (NSFC) (82471499) to WQZ, and Jiangxi Key Laboratory of Neurological Diseases (2024SSY06072) to DH and WQZ.
Data availability
This study primarily used data obtained from the UK Biobank Resource, a public repository, under the application number 1029284.
Declarations
Ethics approval and consent to participate
The UK Biobank received ethical approval from the National Health Service National Research Ethics Service (11/NW/0382; 16/NW/0274). All procedures were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Zihao Zhang, Yuanyuan Zhang and Ziyi Peng contributed equally to this study.
Contributor Information
Daojun Hong, Email: hdj@ncu.edu.cn.
Wen-Quan Zou, Email: wenquanzou@ncu.edu.cn.
References
- 1.Benjafield AV, Sert Kuniyoshi FH, Malhotra A, Martin JL, Morin CM, Maurer LF, et al. Estimation of the global prevalence and burden of insomnia: a systematic literature review-based analysis. Sleep Med Rev. 2025;82:102121. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lim DC, Najafi A, Afifi L, Bassetti C, Buysse DJ, Han F, et al. The need to promote sleep health in public health agendas across the globe. Lancet Public Health. 2023;8:e820–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Svensson T, Svensson AK, Kitlinski M, Engström G, Nilsson J, Orho-Melander M, et al. Very short sleep duration reveals a proteomic fingerprint that is selectively associated with incident diabetes mellitus but not with incident coronary heart disease: a cohort study. BMC Med. 2024;22:173. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Dashti HS, Jones SE, Wood AR, Lane JM, van Hees VT, Wang H, et al. Genome-wide association study identifies genetic loci for self-reported habitual sleep duration supported by accelerometer-derived estimates. Nat Commun. 2019;10:1100. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Yang C, Yan P, Wu X, Zhang W, Cui H, Zhang L, et al. Associations of sleep with cardiometabolic risk factors and cardiovascular diseases: An umbrella review of observational and mendelian randomization studies. Sleep Med Rev. 2024;77:101965. [DOI] [PubMed] [Google Scholar]
- 6.Livingston G, Huntley J, Liu KY, Costafreda SG, Selbæk G, Alladi S, et al. Dementia prevention, intervention, and care: 2024 report of the Lancet standing Commission. Lancet. 2024;404:572–628. [DOI] [PubMed] [Google Scholar]
- 7.St-Onge MP, Grandner MA, Brown D, Conroy MB, Jean-Louis G, Coons M, et al. Sleep Duration and Quality: Impact on Lifestyle Behaviors and Cardiometabolic Health: A Scientific Statement From the American Heart Association. Circulation. 2016;134:e367–86. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Jaspan VN, Greenberg GS, Parihar S, Park CM, Somers VK, Shapiro MD, et al. The Role of Sleep in Cardiovascular Disease. Curr Atheroscler Rep. 2024;26:249–62. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Figorilli M, Velluzzi F, Redolfi S. Obesity and sleep disorders: A bidirectional relationship. Nutr Metab Cardiovasc Dis. 2025;35:104014. [DOI] [PubMed] [Google Scholar]
- 10.Koemel NA, Biswas RK, Ahmadi MN, Teixeira-Pinto A, Hamer M, Rezende LFM, et al. Minimum combined sleep, physical activity, and nutrition variations associated with lifeSPAN and healthSPAN improvements: a population cohort study. EClinicalMedicine. 2026;92:103741. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Drager LF, Polotsky VY. Lipid metabolism: a new frontier in sleep apnea research. Am J Respir Crit Care Med. 2011;184:288–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Huang LY, Tan L, Tan CC, Zhang XY, Lin X, Zou JJ, et al. Sleep quality, APOE ε4, and Alzheimer’s disease: associations from two prospective cohort studies and mechanisms by plasma proteomic analysis. BMC Med. 2025;23:462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Wang Y, Zhou B, Yue W, Wang M, Hu K. Glucose and lipid metabolism in non-diabetic, non-obese patients with obstructive sleep apnea: sex differences. Front Nutr. 2025;12:1619371. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Yang Y, Wu G, Ren L. Correlation between TG/HDL-C ratio and obstructive sleep apnea in the adult US population: a cross-sectional study. Front Neurol. 2025;16:1594875. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chen Z, Zhang X, Duan Y, Mo T, Liu W, Ma Y, et al. The Relationship Between Sleep Duration and Blood Lipids Among Chinese Middle-Aged and Older Adults: Cross-Lagged Path Analysis From CHARLS. Front Public Health. 2022;10:868059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Du J, Chen Y, Zhou N, Song Y, Wang W, Hong X. Associations between self-reported sleep duration and abnormal serum lipids in eastern China: a population-based cross-sectional survey. Sleep Med. 2022;95:1–8. [DOI] [PubMed] [Google Scholar]
- 17.Yamada T, Yamamoto Y, Toragai R, Watanabe Y, Kuroshima T, Arimoto M, Nakamura M, Ito Y, Ai M. Association of sleep duration and quality with small dense low-density lipoprotein cholesterol. J Clin Lipidol. 2026;20(3):578–86. [DOI] [PubMed]
- 18.Barragán R, Zuraikat FM, Cheng B, Scaccia SE, Cochran J, Aggarwal B, et al. Paradoxical Effects of Prolonged Insufficient Sleep on Lipid Profile: A Pooled Analysis of 2 Randomized Trials. J Am Heart Assoc. 2023;12:e032078. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kruisbrink M, Robertson W, Ji C, Miller MA, Geleijnse JM, Cappuccio FP. Association of sleep duration and quality with blood lipids: a systematic review and meta-analysis of prospective studies. BMJ Open. 2017;7:e018585. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Korostovtseva L, Alieva A, Rotar O, Bochkarev M, Boyarinova M, Sviryaev Y, Konradi A, Shlyakhto E. Sleep Duration, Lipid Profile and Insulin Resistance: Potential Role of Lipoprotein(a). Int J Mol Sci. 2020;21(13). [DOI] [PMC free article] [PubMed]
- 21.Broussard J, Brady MJ. The impact of sleep disturbances on adipocyte function and lipid metabolism. Best Pract Res Clin Endocrinol Metab. 2010;24:763–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Liu HH, Li JJ. Aging and dyslipidemia: a review of potential mechanisms. Ageing Res Rev. 2015;19:43–52. [DOI] [PubMed] [Google Scholar]
- 23.Mach F, Baigent C, Catapano AL, Koskinas KC, Casula M, Badimon L, et al. 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk. Eur Heart J. 2020;41:111–88. [DOI] [PubMed] [Google Scholar]
- 24.Third Report of the National Cholesterol Education Program (NCEP). Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143–421. [PubMed] [Google Scholar]
- 25.Kim J, Kim D, Bae HJ, Park BE, Kang TS, Lim SH, et al. Associations of combined polygenic risk score and glycemic status with atrial fibrillation, coronary artery disease and ischemic stroke. Cardiovasc Diabetol. 2024;23:5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zhang Z, Huang J, Wang X, Li Q, Yang Y, Ning K, Wang Y, Chen T, Xiong R, Liu X, et al. Joint associations of grip strength, lipid levels, and lipid-related genetic risk with atrial fibrillation. Lipids Health Dis. 2026;25(1). [DOI] [PMC free article] [PubMed]
- 27.Syauqy A, Hsu CY, Rau HH, Kurniawan AL, Chao JC. Association of Sleep Duration and Insomnia Symptoms with Components of Metabolic Syndrome and Inflammation in Middle-Aged and Older Adults with Metabolic Syndrome in Taiwan. Nutrients. 2019;11(8). [DOI] [PMC free article] [PubMed]
- 28.Grandner MA, Sands-Lincoln MR, Pak VM, Garland SN. Sleep duration, cardiovascular disease, and proinflammatory biomarkers. Nat Sci Sleep. 2013;5:93–107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Grandner MA, Drummond SP. Who are the long sleepers? Towards an understanding of the mortality relationship. Sleep Med Rev. 2007;11:341–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Drager LF, Lopes HF, Maki-Nunes C, Trombetta IC, Toschi-Dias E, Alves MJ, et al. The impact of obstructive sleep apnea on metabolic and inflammatory markers in consecutive patients with metabolic syndrome. PLoS ONE. 2010;5:e12065. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Yeo BSY, Koh JH, Tan BKJ, Ding Y, Teo YH, Alkan U, et al. Improved Inflammatory and Cardiometabolic Profile After Soft-Tissue Sleep Surgery for Obstructive Sleep Apnea: A Systematic Review and Meta-analysis. JAMA Otolaryngol Head Neck Surg. 2022;148:862–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Souza SP, Santos RB, Santos IS, Parise BK, Giatti S, Aielo AN, et al. Obstructive Sleep Apnea, Sleep Duration, and Associated Mediators With Carotid Intima-Media Thickness: The ELSA-Brasil Study. Arterioscler Thromb Vasc Biol. 2021;41:1549–57. [DOI] [PubMed] [Google Scholar]
- 33.Eckel RH, Depner CM, Perreault L, Markwald RR, Smith MR, McHill AW, et al. Morning Circadian Misalignment during Short Sleep Duration Impacts Insulin Sensitivity. Curr Biol. 2015;25:3004–10. [DOI] [PubMed] [Google Scholar]
- 34.Depner CM, Stothard ER, Wright KP. Jr. Metabolic consequences of sleep and circadian disorders. Curr Diab Rep. 2014;14:507. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.McHill AW, Melanson EL, Higgins J, Connick E, Moehlman TM, Stothard ER, et al. Impact of circadian misalignment on energy metabolism during simulated nightshift work. Proc Natl Acad Sci U S A. 2014;111:17302–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Potter GD, Skene DJ, Arendt J, Cade JE, Grant PJ, Hardie LJ. Circadian Rhythm and Sleep Disruption: Causes, Metabolic Consequences, and Countermeasures. Endocr Rev. 2016;37:584–608. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Lane JM, Jones SE, Dashti HS, Wood AR, Aragam KG, van Hees VT, et al. Biological and clinical insights from genetics of insomnia symptoms. Nat Genet. 2019;51:387–93. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Assessment of adverse. effects attributed to statin therapy in product labels: a meta-analysis of double-blind randomised controlled trials. Lancet. 2026;407:689–703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Zeng W, Deng H, Luo Y, Zhong S, Huang M, Tomlinson B. Advances in statin adverse reactions and the potential mechanisms: A systematic review. J Adv Res. 2025;76:781–97. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Broncel M, Gorzelak-Pabiś P, Sahebkar A, Serejko K, Ursoniu S, Rysz J, et al. Sleep changes following statin therapy: a systematic review and meta-analysis of randomized placebo-controlled polysomnographic trials. Arch Med Sci. 2015;11:915–26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Salari S, Moeinipour Y, Elahabadi G, Moeinipour A, Mohammadpour AH. Sleep effects of hydrophilic and lipophilic statins: a comparative narrative review of clinical and experimental evidence. Behav Pharmacol. 2026;37:71–84. [DOI] [PubMed] [Google Scholar]
- 42.Zhang J, Liu Q. Cholesterol metabolism and homeostasis in the brain. Protein Cell. 2015;6:254–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Shin KC, Ali Moussa HY, Park Y. Cholesterol imbalance and neurotransmission defects in neurodegeneration. Exp Mol Med. 2024;56:1685–90. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Zhang Y, Chen X, Tang Y, Wang Y, Zhang K, Wu J. Dietary inflammatory potential exacerbates sleep disturbances in hyperlipidemia: mediation by BMI and neuro-metabolic pathways. BMC Nutr. 2025;11:182. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Garcia C, Andersen CJ, Blesso CN. The Role of Lipids in the Regulation of Immune Responses. Nutrients. 2023;15(18). [DOI] [PMC free article] [PubMed]
- 46.Tall AR, Yvan-Charvet L. Cholesterol, inflammation and innate immunity. Nat Rev Immunol. 2015;15:104–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Alic L, Dendinovic K, Papac-Milicevic N. The complement system in lipid-mediated pathologies. Front Immunol. 2024;15:1511886. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
This study primarily used data obtained from the UK Biobank Resource, a public repository, under the application number 1029284.