Abstract
Background
Evidence is limited regarding the relationship between sleep factors (self-reported sleep disorder diagnosis, subjective sleep difficulties, and sleep duration), sleep patterns, and risk of type 2 diabetes mellitus (T2D). Thus, this study aims to investigate the relationship between sleep factors, sleep patterns, and the risk of T2D using data from the National Health and Nutrition Examination Survey (NHANES).
Methods
A total of 14,652 individuals aged ≥18 years from the NHANES (2005–2014) were enrolled with complete data on sleep factors, T2D, and covariates. Information on self-reported sleep disorder diagnosis, subjective sleep difficulties, and sleep duration was collected during in-home visits by trained interviewers using the Computer-Assisted Personal Interviewing system. The sleep pattern was derived from scoring three mentioned factors: no self-reported sleep disorder diagnosis, no subjective sleep difficulties, and sleep duration of 7–9 h were classified as low-risk (score 0), while the presence of self-reported sleep disorder diagnosis, subjective sleep difficulties, or sleep duration <7 or >9 h were classified as high-risk (score 1). Cumulative scores range from 0 to 3, with 0 indicating a healthy sleep pattern, 1 an intermediate sleep pattern, and 2–3 a poor sleep pattern, respectively. Weighted logistic regression was conducted to assess the association of sleep factors and sleep patterns with the risk of T2D.
Results
Self-reported sleep disorder diagnosis (odds ratio (OR) = 1.32, P = 0.01), subjective sleep difficulties (OR = 1.29, P = 0.001), and sleep deprivation (<7 h; OR = 1.20, P = 0.01) were significantly positive with T2D. Poor sleep pattern also significantly increased T2D risk (OR = 1.52, P < 0.0001). Moreover, subgroup analyses stratified by age and BMI (body mass index) further confirmed that the positive association between sleep patterns and T2D was consistent and robust across groups.
Conclusion
Our findings indicate that poorer sleep patterns are associated with an increased risk of T2D. These results emphasize the importance of sleep management in T2D prevention. Further prospective studies are needed to investigate the causal or bidirectional relationship between sleep and T2D risk, as well as the underlying molecular mechanisms.
Keywords: Type 2 diabetes mellitus, Self-reported sleep disorder diagnosis, Subjective sleep difficulties, Sleep duration, Sleep patterns, NHANES
Highlights
-
•
We investigated the association between three sleep factors, sleep patterns, and the risk of T2D.
-
•
A positive association was found between self-reported diagnosis of sleep disorders, subjective sleep difficulties, sleep duration< 7 h, poor sleep patterns, and the risk of T2D.
-
•
Subgroup analyses revealed that poor sleep patterns were associated with a significantly higher risk of T2D in individuals with BMI <25 kg/m2 and BMI ≥30 kg/m2.
1. Introduction
Sleep is a highly complicated process occupying approximately one-third of the human lifespan (Regier et al., 2012). Although sleep is regarded as a passive state, it is a highly dynamic and active process (Besedovsky et al., 2019). According to the National Heart, Lung, and Blood Institute, an estimated 50 to 70 million U.S. adults suffer from a sleep disorder or reported sleep deprivation (St-Onge et al., 2016). Common sleep disorders, such as insomnia, sleep apnea, and restless legs syndrome, require professional diagnosis by a physician or healthcare provider. Beyond sleep disorders, self-reported trouble sleeping is also a critical indicator of sleep quality. A healthy sleep pattern is not only about having the appropriate sleep duration but also involves the entire sleep process, such as falling asleep within a proper time, without snoring, without obstructive sleep apnea, and restless legs syndrome (Petrov et al., 2014; Nelson et al., 2022). Sleep is vital for restoring brain function and growing evidence suggests that sleep is associated with type 2 diabetes mellitus (T2D)(Magnavita et al., 2017). However, previous studies have mainly focused on the relationship between sleep duration and T2D. There has yet to be a comprehensive evaluation of how single sleep factors and sleep patterns (encompassing self-reported sleep disorder diagnosis, self-reported sleep difficulties, and sleep duration), relate to T2D.
T2D is a global health problem of great severity and rising prevalence. It is estimated that by 2045, approximately 700 million people will have diabetes (Wu et al., 2014). One study indicated that sleep insufficiency may contribute to the development of T2D (McNeil et al., 2013). In addition, a prospective study demonstrated significant associations between both excessive and insufficient sleep duration and T2D (Shan et al., 2015). Besides sleep duration, sleep disorders and trouble sleeping may also as risk factors for diabetes (Anothaisintawee et al., 2016). Mechanistically, insufficient sleep disrupts the normal secretion of various hormones, including leptin, ghrelin, and glucagon, among others. This disruption not only affects the secretion and function of insulin but also directly impairs insulin sensitivity, thereby exacerbating insulin resistance. Furthermore, sleep deprivation leads to elevated levels of inflammatory cytokines such as interleukin-6 (IL-6) and C-reactive protein (CRP), further increasing the risk of insulin resistance. (Reutrakul et al., 2018; Tang et al., 2014). However, most studies focus on single sleep factors (particularly sleep duration) rather than combined sleep behaviors, and the findings are often inconsistent due to differences in sample size and methodologies.
The present study investigated the association between single sleep factors (self-reported sleep disorder diagnosis, subjective sleep difficulties, and sleep duration), sleep patterns, and the risk of T2D in a representative sample of the US population. Additionally, we aimed to explore whether these relationships remain consistent across subgroups stratified by age and BMI.
2. Methods
2.1. Study subjects
The National Health and Nutrition Examination Survey (NHANES) is a complex multistage sample survey being conducted by the U.S. government. It determines the physical health status of the entire U.S. population by administering health questionnaires, physical examinations, and laboratory tests to a representative and specific group of people to guide people toward healthier lifestyles.
As illustrated in Fig. 1, we analyzed NHANES data from 2005 to 2014. We excluded 20,670 participants who were under 18 years old, as our study focused on adults. Additionally, 1067 participants without Mobile Examination Center (MEC) data were excluded due to incomplete health assessments. From the sleep questionnaire data, we excluded 43 participants missing self-reported sleep disorder diagnosis data, 7 participants missing subjective sleep difficulties data, and 50 participants missing sleep duration data. Furthermore, we excluded 2773 participants who lacked a confirmed T2D diagnosis. Finally, 11,703 participants with missing critical covariate data, such as BMI, dietary quality, or physical activity, were also excluded. After these exclusions, 14,652 participants were included in the final analysis.
Fig. 1.
Participants selection flowchart.
2.2. Sleep factors and sleep patterns
Self-reported sleep disorder diagnosis, subjective sleep difficulties, and sleep duration were evaluated based on responses to the questions on the sleep questionnaire (SLQ). In the current study, to identify whether participants have self-reported sleep disorder diagnosis, whether they have subjective sleep difficulties, and their sleep duration, we used the NHANES variables SLQ060 (“Have you ever been told by a doctor or other health professional that you have a sleep disorder?“), SLQ050 (“Have you ever told a doctor or other health professional that you have trouble sleeping?“), and SLD010h (“How much sleep do you usually get at night on weekdays or workdays?“), respectively. Sleep duration was divided into three groups: <7, 7–9, and >9 h of sleep. No self-reported sleep disorder diagnosis, no subjective sleep difficulties, and sleep duration of 7–9 h are considered low-risk sleep factors and assigned a score of 0; the rest are considered high-risk sleep factors and assigned a score of 1. The three cumulative scores generate a total score ranging from 0 to 3. The sleep patterns can be determined based on this total score: a score of 2–3 indicates a poor sleep pattern, 1 indicates an intermediate sleep pattern, and 0 indicates a healthy sleep pattern.
2.3. T2D
T2D was diagnosed by a doctor. Medications to reduce blood glucose levels were provided. Fasting blood glucose was equal to or greater than 7.0 mmol/L, glycosylated hemoglobin was equal to or greater than 6.5%, and a 2-h blood glucose level was equal to or greater than 11.1 mmol/L from an oral glucose tolerance test.
2.4. Covariates
The covariates in this study were age, gender, ethnicity, educational level, marital status, Healthy Eating Index (HEI) 2015, poverty-to-income ratio (PIR), physical activity, body mass index (BMI), energy intake, smoking status, alcohol drinking status, hypertension, immune inflammation index (SII), and Patient Health Questionnaire (PHQ-9). Alcohol drinking status was classified into four groups: no, mild, moderate, and heavy. Smoking status was categorized as either smoking or nonsmoking. PIR was categorized with a cutoff of 2. BMI was classified into three groups: <25.0, 25.0–29.9, and ≥30.0 kg/m2. Marital status was categorized as living with a partner or married, separated or divorced or widowed, and never married. Education was categorized as less than high school, high school or General Education Development (GED), and more than high school. Ethnicity was classified as non-Hispanic Black, non-Hispanic White, Mexican Americans, and other races. Age was divided into three groups: 20–39, 40–59, and ≥60 years. Physical activity was measured using metabolic equivalent and classified into four groups. SII is a novel marker derived from lymphocyte, neutrophil, and platelet counts, and is commonly used to assess systemic inflammation. Due to the SII distribution not conforming to the normal distribution we have log-transformed. In addition, we included the PHQ-9 as a covariate to assess depressive symptoms.
2.5. Statistical analysis
In the current study, NHANES survey weights, strata, and primary sampling units were considered to ensure that nationally representative estimates were calculated. Sample weights for the mobile examination center were used. Continuous variables such as age, HEI-2015 score, energy intake, and SII were presented as mean ± standard error (SE). Prior to statistical analysis, tests for normality and homogeneity of variance were conducted. If these assumptions were met, t-tests (for comparisons between groups with and without T2D) or one-way ANOVA (for comparisons among the three sleep patterns) were performed. Categorical variables were expressed as N (%) and assessed for statistical differences using the χ2 test. To assess associations between variables and T2D, multivariate logistic regression was performed, and results were reported as odds ratios (OR) with 95% confidence intervals (CI). Before the regression analysis, the linear relationship between the logit-transformed continuous independent variables and T2D was confirmed, and multicollinearity among predictors was assessed. Three models were constructed for the logistic regression analysis: Model 1 was the unadjusted model. Model 2 was adjusted for gender, age, and ethnicity. Model 3 was adjusted for gender, age, ethnicity, marital status, BMI, PIR, smoking, alcohol consumption, physical activity, HEI-2015 index, hypertension, SII, PHQ-9. P value < 0.05 was considered statistically significant. All analyses in this study were performed using R software (version 4.2.2).
3. Results
3.1. Demographics of the study population
The characteristics of the participants, classified according to T2D status, are presented in Table 1. Of the total 14,652 participants, 6895 (47.06%) were female and 7757 (52.94%) were male, with a mean age of 44.95 ± 0.31 years. The overall weighted prevalence of T2D was 15.86%. Age, BMI, physical activity, energy intake, ethnicity, education, marital status, smoking, alcohol consumption, hypertension, PIR, SII, PHQ-9, self-reported sleep disorder diagnosis, subjective sleep difficulties, and sleep duration were significantly statistically different (P < 0.0001) between those with and without T2D.
Table 1.
Demographics of participants aged ≥20 years according to T2D.
| Characteristics | Total (N = 14,652) | Without T2D (N = 12,328) | With T2D (N = 2324) | P-value |
|---|---|---|---|---|
| Age, (years) | 44.95 ± 0.31 | 43.34 ± 0.32 | 57.29 ± 0.36 | <0.0001 |
| Gender, N (%) | 0.05 | |||
| Female | 6895 (47.06) | 5862 (48.73) | 1033 (45.83) | |
| male | 7757 (52.94) | 6466 (51.27) | 1291 (54.17) | |
| Ethnicity, N (%) | <0.0001 | |||
| Non-Hispanic Black | 3004 (20.5) | 2411 (9.56) | 593 (14.15) | |
| Non-Hispanic White | 7178 (48.99) | 6248 (73.21) | 930 (65.00) | |
| Mexican American | 2042 (13.94) | 1642 (6.95) | 400 (8.77) | |
| Others | 2428 (16.57) | 2027 (10.28) | 401 (12.07) | |
| Education, N (%) | <0.0001 | |||
| Less than high school | 2994 (20.43) | 2323(12.82) | 671 (18.74) | |
| High school or GED | 3323 (22.68) | 2750 (21.54) | 573 (26.63) | |
| More than high school | 8335 (56.89) | 7255 (65.64) | 1080 (54.63) | |
| Marital status, N (%) | <0.0001 | |||
| Living with partner or Married | 8791 (60) | 7352 (63.53) | 1439 (65.95) | |
| Separated or Divorced or Widowed | 2889 (19.72) | 2216 (15.52) | 673 (25.87) | |
| Never married | 2972 (20.28) | 2760 (20.95) | 212 (8.18) | |
| HEI-2015 score | 51.23 ± 0.23 | 51.16 ± 0.25 | 51.72 ± 0.41 | 0.17 |
| PIR, N (%) | <0.0001 | |||
| < 2 | 6460 (44.09) | 5311 (30.96) | 1149 (36.30) | |
| ≥ 2 | 8192 (55.91) | 7017 (69.04) | 1175 (63.70) | |
| Physical activity (MET-h/week) | <0.0001 | |||
| Q1 (≤560) | 371 5(25.35) | 2997 (24.39) | 718 (30.38) | |
| Q2 (560, 1680] | 3875 (26.45) | 3179 (26.17) | 696 (31.33) | |
| Q3 (1680, 4800] | 3552 (24.24) | 3041 (25.47) | 511 (21.72) | |
| Q4 (>4800) | 3510 (23.96) | 3111 (23.97) | 399 (16.57) | |
| BMI, N (%), (kg/m2) | <0.0001 | |||
| < 25 | 4666 (31.85) | 4346 (36.22) | 320 (12.69) | |
| 25–29.9 | 4973 (33.94) | 4266 (34.98) | 707 (27.93) | |
| ≥ 30 | 5013 (34.21) | 3716 (28.80) | 1297 (59.38) | |
| Energy intake, (kcal) | 2254.37 ± 12.32 | 2279.53 ± 12.57 | 2061.12 ± 27.40 | <0.0001 |
| Smoke, N (%) | <0.0001 | |||
| No | 11,446 (78.12) | 9501 (77.98) | 1945 (84.85) | |
| Yes | 3206 (21.88) | 2827 (22.02) | 379 (15.15) | |
| Alcohol consumption, N (%) | <0.0001 | |||
| No | 4073 (27.8) | 3076 (20.88) | 997 (36.92) | |
| Mild | 4961 (33.86) | 4202 (36.15) | 759 (37.10) | |
| Hypertension, N (%) | <0.0001 | |||
| No | 9207 (62.84) | 8472 (71.69) | 735 (32.88) | |
| Yes | 5445 (37.16) | 3856 (28.31) | 1589 (67.12) | |
| SII | 6.17 ± 0.01 | 6.16 ± 0.01 | 6.22 ± 0.01 | <0.0001 |
| PHQ-9, N (%) | <0.0001 | |||
| [0,9] | 13550 (93.71) | 11482 (94.25) | 2068 (90.75) | |
| [10,27] | 1068 (6.14) | 818 (5.75) | 250 (9.25) | |
| Self-reported sleep disorder diagnosis, N (%) | <0.0001 | |||
| No | 13,509 (92.2) | 11,507 (93.31) | 2002 (84.72) | |
| Yes | 1143 (7.8) | 821(6.69) | 322 (15.28) | |
| Subjective sleep difficulties, N (%) | <0.0001 | |||
| No | 11,133 (75.98) | 9577 (75.92) | 1556 (64.57) | |
| Yes | 3519 (24.02) | 2751 (24.08) | 768 (35.43) | |
| Sleep time, hours N (%) | <0.0001 | |||
| 7-9 | 8634 (58.93) | 7382 (63.39) | 1252 (56.19) | |
| < 7 | 5741 (39.18) | 4721 (35.02) | 1020 (41.72) | |
| ≥ 10 | 2,77 (1.89) | 225 (1.59) | 52 (2.08) |
Continuous variables were presented as mean ± SE; Categorical variables expressed as N (%); t-test and χ2 test were performed to assess statistical differences between participants without T2D and with T2D.
BMI: body mass index; PIR: poverty income ratio, HEI-2015: Healthy Eating Index-2015; MET: metabolic equivalent; SII: immune inflammation index; PHQ-9: Patient Health Questionnaire; Q: quartile.
To examine the demographic characteristics of participants with different sleep patterns, we categorized them into three groups: healthy sleep, intermediate sleep, and poor sleep. Individuals with poor sleep patterns appeared to be older, more likely to be female, non-Hispanic White, have lower educational levels, live alone, and have poorer dietary quality, lower incomes, trends of obesity, and hypertension (Table S1). Furthermore, T2D prevalence was significantly higher in participants with poor sleep pattern (19.93%) compared to intermediate (10.75%) and healthy sleep patterns (9.31%) (P < 0.0001) (Table S1).
3.2. Link between sleep factors and risk of T2D
As presented in Table 2, in the unadjusted model (Model 1), participants with self-reported sleep disorder diagnosis, subjective sleep difficulties, and sleep duration <7 h had 2.51 (95% CI: 2.12–2.98), 1.73 (95% CI: 1.53–1.96) and 1.34-fold (95% CI: 1.20–1.50) more likely to have T2D, respectively. In the fully adjusted model (Model 3), self-reported sleep disorder diagnosis (OR = 1.32, P = 0.01) and subjective sleep difficulties (OR = 1.29, P = 0.001) remained significantly associated with the risk of T2D compared with the corresponding reference (Table 2). Compared to a sleep duration of 7–9 h, sleeping <7 h (OR = 1.20, P = 0.01) was associated with a higher risk of T2D, while sleeping ≥9 h (OR = 1.47, P = 0.13) showed no statistically significant association (fully adjusted model). These findings suggest that sleep deprivation may pose a greater risk for T2D and should be given more attention.
Table 2.
Logistic regression analysis of sleep factors and T2D.
| Model | Sleep factors | OR (95% CI) | P-value |
|---|---|---|---|
| Model 1 | Self-reported sleep disorder diagnosis | ||
| No | ref | ref | |
| Yes | 2.51 (2.12, 2.98) | <0.0001 | |
| Subjective sleep difficulties | |||
| No | ref | ref | |
| Yes | 1.73 (1.53, 1.96) | <0.0001 | |
| Sleep time | |||
| 7–9 | ref | ref | |
| <7 | 1.34 (1.20, 1.50) | <0.0001 | |
| ≥10 | 1.48 (0.99, 2.21) | 0.05 | |
| Model 2 | Self-reported sleep disorder diagnosis | ||
| No | ref | ref | |
| Yes | 2.32 (1.92, 2.80) | <0.0001 | |
| Subjective sleep difficulties | |||
| No | ref | ref | |
| Yes | 1.66 (1.45, 1.90) | <0.0001 | |
| Sleep time | |||
| 7–9 | ref | ref | |
| <7 | 1.40 (1.24, 1.59) | <0.0001 | |
| ≥10 | 1.51 (0.94, 2.42) | 0.09 | |
| Model 3 | Self-reported sleep disorder diagnosis | ||
| No | ref | ref | |
| Yes | 1.32 (1.06, 1.63) | 0.01 | |
| Subjective sleep difficulties | |||
| No | ref | ref | |
| Yes | 1.29 (1.11, 1.49) | 0.001 | |
| Sleep time | |||
| 7–9 | ref | ref | |
| <7 | 1.20 (1.05, 1.38) | 0.01 | |
| ≥10 | 1.47 (0.89, 2.43) | 0.13 | |
Model 1: crude model.
Model 2: adjusted for gender, age and ethnicity.
Model 3: adjusted for Mode 2 plus education, marital status, BMI, PIR, smoke, alcohol consumption, physical activity, Hei-2015 index, hypertension, SII, PHQ-9.
Ref: references.
3.3. Association between sleep patterns and the risk of T2D
As demonstrated in Table 3, in the crude model (Model 1), a poor sleep pattern (OR = 2.42, P < 0.0001) and a intermediate sleep pattern (OR = 1.17, P = 0.02) were positively associated with the risk of T2D compared with a healthy sleep pattern. In the fully adjusted model (Model 3), this relationship between a poor sleep pattern and T2D remained robust (OR = 1.52, P < 0.0001), but no significant differences were observed between a intermediate sleep pattern and T2D.
Table 3.
Logistic regression analysis of sleep patterns and T2D.
| Model | Sleep patterns | OR (95% CI) | P-value |
|---|---|---|---|
| Model 1 |
Healthy sleep pattern | ref | ref |
| Intermediate sleep pattern | 1.17 (1.02, 1.35) | 0.02 | |
| Poor sleep pattern |
2.42 (2.07, 2.84) |
<0.0001 |
|
| Model 2 |
Healthy sleep pattern | ref | ref |
| Intermediate sleep pattern | 1.20 (1.04, 1.38) | 0.01 | |
| Poor sleep pattern |
2.31 (1.97, 2.71) |
<0.0001 |
|
| Model 3 | Healthy sleep pattern | ref | ref |
| Intermediate sleep pattern | 1.13 (0.97, 1.32) | 0.10 | |
| Poor sleep pattern | 1.52 (1.27, 1.82) | <0.0001 |
Healthy sleep pattern: score 0, no high-risk sleep factors. Intermediate sleep pattern: score 1, presence of 1 high-risk sleep factor. Poor sleep pattern: score 2–3, presence of 2–3 high-risk sleep factors.
Model 1: crude model.
Model 2: adjusted for gender, age and ethnicity.
Model 3: adjusted for Mode 2 plus education, marital status, BMI, PIR, smoke, alcohol consumption, physical activity, HEI-2015, hypertension, SII, PHQ-9.
Ref: reference.
3.4. Link between sleep patterns with risk of T2D by age and BMI
When categorized by age, poor sleep pattern were consistently associated with a significantly increased risk of constipation. Among individuals aged 40–59 years, the risk was highest (OR = 1.852, P < 0.0001). The P-trend analysis indicated that the risk of constipation increased significantly with worsening sleep patterns in both the 40–59 years and ≥60 years age groups (P-trend <0.0001) (Table 4). Stratified by BMI, poor sleep pattern were associated with a significantly increased risk of constipation only in individuals with BMI of 25–29.9 (OR = 1.475, P = 0.040) and BMI ≥30 kg/m2 (OR = 1.659, P < 0.0001). The P-trend analysis further revealed that the risk of constipation increased significantly with worsening sleep patterns in individuals with a BMI ≥30 (P-trend <0.0001) (Table 5).
Table 4.
Logistic regression analysis of sleep patterns and constipation by age.
| Age, years | Healthy sleep pattern |
Intermediate sleep pattern |
Poor sleep pattern |
P-trend | ||
|---|---|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | P-value | OR (95% CI) | P-value | ||
| Model 1 | ||||||
| 20-39 | ref | 1.071 (0.715, 1.602) | 0.737 | 2.759 (1.699, 4.452) | <0.0001 | 0.002 |
| 40-59 | ref | 1.391 (1.113, 1.737) | 0.004 | 2.296 (1.819, 2.898) | <0.0001 | <0.0001 |
| ≥ 60 | ref | 1.112 (0.882, 1.401) | 0.365 | 2.044 (1.606, 2.600) | <0.0001 | <0.0001 |
| Model 2 | ||||||
| 20-39 | ref | 1.060 (0.706, 1.592) | 0.776 | 2.837 (1.755, 4.588) | <0.0001 | 0.001 |
| 40-59 | ref | 1.331 (1.064, 1.665) | 0.013 | 2.364 (1.873, 2.983) | <0.0001 | <0.0001 |
| ≥ 60 | ref | 1.083 (0.846, 1.386) | 0.521 | 2.038 (1.597, 2.600) | <0.0001 | <0.0001 |
| Model 3 | ||||||
| 20-39 | ref | 0.951 (0.626, 1.444) | 0.810 | 1.797 (1.036, 3.118) | 0.037 | 0.110 |
| 40-59 | ref | 1.296(1.012, 1.634) | 0.040 | 1.852 (1.433, 2.394) | <0.0001 | <0.0001 |
| ≥ 60 | ref | 1.076 (0.822, 1.379) | 0.577 | 1.802 (1.386, 2.342) | <0.0001 | <0.001 |
Healthy sleep pattern: score 0, no high-risk sleep factors. Intermediate sleep pattern: score 1, presence of 1 high-risk sleep factor. Poor sleep pattern: score 2–3, presence of 2–3 high-risk sleep factors.
Model 1: without excluding any confounders.
Model 2: adjusted for gender, age and ethnicity.
Model 3: adjusted for Mode 2 plus education, marital status, BMI, PIR, smoke, alcohol consumption, physical activity, HEI-2015, hypertension, SII, PHQ-9.
Ref: reference.
Table 5.
Logistic regression analysis of sleep patterns and constipation by BMI.
| BMI, Kg/m2 | Healthy sleep |
Intermediate sleep |
Poor sleep |
P-trend | ||
|---|---|---|---|---|---|---|
| OR (95% CI) | OR (95% CI) | P-value | OR (95% CI) | P-value | ||
| Model 1 | ||||||
| < 25 | ref | 1.248 (0.908, 1.714) | 0.169 | 1.935 (1.198, 3.125) | 0.008 | 0.006 |
| 25–29.9 | ref | 1.002 (0.796, 1.261) | 0.986 | 1.715 (1.253, 2.346) | <0.001 | 0.006 |
| ≥ 30 | ref | 1.104 (0.889, 1.371) | 0.368 | 2.133 (1.777, 2.560) | <0.0001 | <0.0001 |
| Model 2 | ||||||
| < 25 | ref | 1.198 (0.865, 1.658) | 0.272 | 2.042 (1.263, 3.302) | 0.004 | 0.006 |
| 25–29.9 | ref | 0.968 (0.768, 1.219) | 0.779 | 1.745 (1.266, 2.405) | <0.001 | 0.008 |
| ≥ 30 | ref | 1.099 (0.883, 1.369) | 0.394 | 2.154 (1.792, 2.588) | <0.0001 | <0.0001 |
| Model 3 | ||||||
| < 25 | ref | 1.187 (0.855, 1.649) | 0.300 | 1.701 (0.967, 2.992) | 0.065 | 0.053 |
| 25–29.9 | ref | 0.959 (0.743, 1.239) | 0.747 | 1.475 (1.019, 2.134) | 0.040 | 0.100 |
| ≥ 30 | ref | 1.152 (0.912, 1.456) | 0.231 | 1.659 (1.349, 2.041) | <0.0001 | <0.0001 |
Healthy sleep pattern: score 0, no high-risk sleep factors. Intermediate sleep pattern: score 1, presence of 1 high-risk sleep factor. Poor sleep pattern: score 2–3, presence of 2–3 high-risk sleep factors.
Model 1: without excluding any confounders.
Model 2: adjusted for gender, age and ethnicity.
Model 3: adjusted for Mode 2 plus education, marital status, BMI, PIR, smoke, alcohol consumption, physical activity, HEI-2015, hypertension, SII, PHQ-9.
Ref: reference.
4. Discussion
The present study investigated the correlation between sleep factors, sleep patterns, and T2D in a nationally representative U.S. population aged 20 years or older. We demonstrated that self-reported sleep disorder diagnosis, subjective sleep difficulties, sleep duration <7 h, and poor sleep patterns were associated with an increased risk of T2D. Then we performed subgroup analyses and observed that poor sleep pattern was associated with a significantly higher risk of T2D in all age groups and in individuals with BMI <25 and BMI≥ 30 kg/m2.
T2D presents with impaired glucose metabolism in the liver and functional defects in the number and function of β-cells, resulting in inadequate insulin secretion and insulin resistance (Butler et al., 2003; Skyler et al., 2017). We can improve individuals' quality of life by addressing the risk factors for T2D to prevent its onset and reduce its effects. Sleep plays a crucial role in restoring endocrine, immune, and neurological functions and largely reflects overall human health (Chaput et al., 2023; Krueger et al., 2016; Irwin, 2019). We observed that individuals who slept <7 h exhibited a significantly higher risk of developing T2D compared to those who slept 7–9 h, which is consistent with previous findings. Holliday EG et al. reported that individuals sleeping <6 h per night had a 30% increased risk of T2D compared to those sleeping 7 h per night (Holliday et al., 2013). A meta-analysis of 447,124 participants also confirmed the association between short sleep duration and T2D incidence (RR = 1.33, 95% CI: 1.20–1.48) (Cappuccio et al., 2010). However, the relationship between long sleep duration and T2D risk remains controversial. In our study, no significant association was observed between long sleep duration and T2D risk (OR = 1.47, 95% CI: 0.89–2.43). Conversely, other studies have suggested that long sleep duration may increase T2D risk. For instance, a Finnish population-based cross-sectional study found that ≥8 h of sleep was associated with increased T2D risk in middle-aged women (Tuomilehto et al., 2008). Additionally, a recent meta-analysis indicated that long sleep duration (≥9 h) was associated with an elevated T2D risk (RR = 1.26, 95% CI: 1.15–1.39, P < 0.001) (Lu et al., 2021). Interestingly, a study of UK children reported a negative correlation between sleep duration and T2D (Rudnicka et al., 2017). Furthermore, a recent study found that sleep efficiency (<80%) was associated with T2D prevalence in individuals without sleep-disordered breathing (OR = 1.894, 95% CI: 1.187–3.022, P = 0.007) (Yan et al., 2020). These inconsistencies may be due to differences in sleep assessment criteria, and variations in study populations regarding age, ethnicity, and environmental factors. Self-reported sleep duration may also be influenced by demographic characteristics, socioeconomic status, employment conditions, and health behaviors. Therefore, further research is needed to elucidate the relationship between long sleep duration and T2D.
T2D is often comorbid with hypertension, obesity, cardiovascular disease and psychiatric disorders. Studies have shown that sleep deprivation is associated with obesity, potentially mediated by disruptions in leptin and ghrelin regulation (Spiegel et al., 2004). Obstructive sleep apnoea, a common sleep disorder, is particularly prevalent among obese individuals with T2D and its treatment has shown significant benefits in improving T2D symptoms (Kurnool et al., 2023). In our subgroup analysis, individuals with a BMI ≥30 and poor sleep patterns had a higher risk of T2D compared to those with healthy sleep patterns. In addition, maintaining a lifestyle that includes healthy sleep patterns may reduce the risk of T2D in hypertensive patients (Song et al., 2021). Depression is the most common mental disorder among individuals with T2D, with a prevalence of approximately 10%–15%, which is about twice that observed in non-diabetic individuals (Sartorius, 2018). Depression can lead to impaired self-management of the disease, worsened glycemic control, and increased complications and mortality (Holt et al., 2014). Furthermore, individuals with depression often experience poor sleep quality. In the current study, after adjusting for depressive symptoms as a covariate in multivariable logistic regression and subgroup analyses, we still observed an increased risk of T2D for self-reported sleep disorder diagnoses, subjective sleep difficulties, short sleep duration, and poor sleep patterns, suggesting these factors are independent risk factors for T2D. A NHANES-based study identified a nomogram model combining nine predictors, including sleep duration and sleep disorders, that accurately predicted the risk of depression in T2D patients with strong predictive performance (AUC >0.75) (Yu et al., 2024). However, in a small clinical trial, Safari M et al. found that treatment with Melissa officinalis extract improved symptoms of depression and anxiety in T2D patients with depression but had no significant impact on sleep quality after 12 weeks of intervention (Safari et al., 2023). Thus, further research is needed to explore the interrelationships between sleep factors, depression, and T2D.
Mechanistically, sleep restriction may increase T2D risk through multiple pathways. Studies have shown that sleep restriction significantly increases insulin resistance, reduces insulin sensitivity, and impairs glucose tolerance (Broussard et al., 2012), (Dracup et al., 1990). In a randomized crossover trial, restricting sleep to <5 h for 5 consecutive nights resulted in over-activation of the sympathetic nervous system and elevated serum norepinephrine levels (Dettoni et al., 2012). Sleep deprivation also elevates pro-inflammatory markers, such as IL-1β, IL-6, and C-reactive protein (Frey et al., 2007). Furthermore, in healthy individuals, restricting sleep to 4 h for 5 consecutive nights significantly increased the expression of IL-6, IL-1β, and IL-17 (van Leeuwen et al., 2009).
Based on these findings, we recommend strengthening the focus on sleep health in public health. First, health organizations should include evidence-based sleep recommendations in their guidelines to promote health. Second, public health promotion should focus on the importance of sleep duration and screening for sleep disorders. Finally, healthcare providers should regularly screen T2D patients for sleep disorders and implement targeted interventions to manage sleep disorders.
Our study utilized a robust and representative database, employing multiple models to minimize confounding factors. However, several limitations merit consideration. Firstly, reliance on self-reported sleep measures introduces potential recall and reporting biases, compromising the accuracy of sleep factor assessment. Objective methods, such as actigraphy or polysomnography, are necessary to validate our findings. Secondly, the cross-sectional design of our study prevents causal inferences between sleep factors and T2D risk, highlighting the need for longitudinal studies. Finally, this study did not explore potential molecular mechanisms underlying the association between sleep factors, sleep patterns, and T2D.
5. Conclusions
This nationally representative study emphasises self-reported diagnoses of sleep disorders, subjective sleep difficulties, sleep duration of <7 h, poor sleep patterns and T2D risk. Further prospective studies should be performed to investigate the causal relationship between sleep and T2D risk. In particular, exploring the underlying molecular mechanisms between sleep and T2D is also essential for the effective management of T2D.
CRediT authorship contribution statement
Jijun Zhang: Writing – original draft, Formal analysis, Data curation. Chuanli Yang: Writing – original draft, Formal analysis, Data curation. Jie An: Writing – review & editing. Yunhe Fan: Writing – review & editing. Xiushan Dong: Methodology.
Ethics approval and consent to participate
All NHANES surveys were approved by the National Centre for Health Statistics Research Ethics.
Funding
This study was funded by the Shanxi Province “136” Revitalization Medical Project Construction Fund, the Key R&D Program of Shanxi Province (202102130501015), and Research and Innovation Team Project for Scientific Breakthroughs at Shanxi Bethune Hospital (2024ZHANCHI06).
Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgments
We are grateful for the NHANES database for the freely accessible data for research analysis.
Footnotes
Supplementary data to this article can be found online at https://doi.org/10.1016/j.bbih.2025.100953.
Abbreviations:
- BMI
body mass index
- T2D
Type 2 diabetes mellitus
- MET
metabolic equivalent
- HEI-2015
Healthy Eating Index-2015
- NHANES
National Health and Nutrition Examination Survey
- OR
odds ratio
- SLQ
sleep questionnaire
- PIR
Poverty-to-income ratio
Appendix A. Supplementary data
The following is the Supplementary data to this article:
Data availability
The datasets used and analysed during the current study are available in the NHANES (http://www.cdc.gov/nchs/nhanes.htm).
References
- Anothaisintawee T., et al. Sleep disturbances compared to traditional risk factors for diabetes development: systematic review and meta-analysis. Sleep Med. Rev. 2016;30:11–24. doi: 10.1016/j.smrv.2015.10.002. [DOI] [PubMed] [Google Scholar]
- Besedovsky L., et al. The sleep-immune crosstalk in health and disease. Physiol. Rev. 2019;99:1325–1380. doi: 10.1152/physrev.00010.2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Broussard J.L., et al. Impaired insulin signaling in human adipocytes after experimental sleep restriction: a randomized, crossover study. Ann. Intern. Med. 2012;157:549–557. doi: 10.7326/0003-4819-157-8-201210160-00005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Butler A.E., et al. Beta-cell deficit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes. 2003;52:102–110. doi: 10.2337/diabetes.52.1.102. [DOI] [PubMed] [Google Scholar]
- Cappuccio F.P., et al. Quantity and quality of sleep and incidence of type 2 diabetes: a systematic review and meta-analysis. Diabetes Care. 2010;33:414–420. doi: 10.2337/dc09-1124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chaput J.-P., et al. The role of insufficient sleep and circadian misalignment in obesity. Nat. Rev. Endocrinol. 2023;19:82–97. doi: 10.1038/s41574-022-00747-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dettoni J.L., et al. Cardiovascular effects of partial sleep deprivation in healthy volunteers. J. Appl. Physiol. 2012;113:232–236. doi: 10.1152/japplphysiol.01604.2011. 1985. [DOI] [PubMed] [Google Scholar]
- Dracup K., et al. Readers respond with ideas for improving diagnostic abilities. Heart Lung. 1990;19:3. [PubMed] [Google Scholar]
- Frey D.J., et al. The effects of 40 hours of total sleep deprivation on inflammatory markers in healthy young adults. Brain Behav. Immun. 2007;21:1050–1057. doi: 10.1016/j.bbi.2007.04.003. [DOI] [PubMed] [Google Scholar]
- Holliday E.G., et al. Short sleep duration is associated with risk of future diabetes but not cardiovascular disease: a prospective study and meta-analysis. PLoS One. 2013;8 doi: 10.1371/journal.pone.0082305. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Holt R.I.G., et al. Diabetes and depression. Curr. Diabetes Rep. 2014;14:491. doi: 10.1007/s11892-014-0491-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Irwin M.R. Sleep and inflammation: partners in sickness and in health. Nat. Rev. Immunol. 2019;19:702–715. doi: 10.1038/s41577-019-0190-z. [DOI] [PubMed] [Google Scholar]
- Krueger J.M., et al. Sleep function: toward elucidating an enigma. Sleep Med. Rev. 2016;28:46–54. doi: 10.1016/j.smrv.2015.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kurnool S., et al. Sleep apnea, obesity, and diabetes - an intertwined trio. Curr. Diabetes Rep. 2023;23:165–171. doi: 10.1007/s11892-023-01510-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lu H., et al. A meta-analysis of a cohort study on the association between sleep duration and type 2 diabetes mellitus. J. Diabetes Res. 2021;2021 doi: 10.1155/2021/8861038. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Magnavita N., et al. Sleep, health and wellness at work: a scoping review. Int. J. Environ. Res. Publ. Health. 2017;14:1347. doi: 10.3390/ijerph14111347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- McNeil J., et al. Inadequate sleep as a contributor to obesity and type 2 diabetes. Can. J. Diabetes. 2013;37:103–108. doi: 10.1016/j.jcjd.2013.02.060. [DOI] [PubMed] [Google Scholar]
- Nelson K.L., et al. Sleep quality: an evolutionary concept analysis. Nurs. Forum. 2022;57:144–151. doi: 10.1111/nuf.12659. [DOI] [PubMed] [Google Scholar]
- Petrov M.E., et al. Prevalence of sleep disorders by sex and ethnicity among older adolescents and emerging adults: relations to daytime functioning, working memory and mental health. J. Adolesc. 2014;37:587–597. doi: 10.1016/j.adolescence.2014.04.007. [DOI] [PubMed] [Google Scholar]
- Regier D.A., et al. Research planning for the future of psychiatric diagnosis. Eur. Psychiatr. 2012;27:553–556. doi: 10.1016/j.eurpsy.2009.11.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Reutrakul S., et al. Sleep influences on obesity, insulin resistance, and risk of type 2 diabetes. Metabolism. 2018;84:56–66. doi: 10.1016/j.metabol.2018.02.010. [DOI] [PubMed] [Google Scholar]
- Rudnicka A.R., et al. Sleep duration and risk of type 2 diabetes. Pediatrics. 2017;140 doi: 10.1542/peds.2017-0338. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Safari M., et al. The effects of melissa officinalis on depression and anxiety in type 2 diabetes patients with depression: a randomized double-blinded placebo-controlled clinical trial. BMC Comple. Med. Ther. 2023;23:140. doi: 10.1186/s12906-023-03978-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Sartorius N. Depression and diabetes. Dialogues Clin. Neurosci. 2018;20:47–52. doi: 10.31887/DCNS.2018.20.1/nsartorius. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Shan Z., et al. Sleep duration and risk of type 2 diabetes: a meta-analysis of prospective studies. Diabetes Care. 2015;38:529–537. doi: 10.2337/dc14-2073. [DOI] [PubMed] [Google Scholar]
- Skyler J.S., et al. Differentiation of diabetes by pathophysiology, natural history, and prognosis. Diabetes. 2017;66:241–255. doi: 10.2337/db16-0806. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Song Z., et al. Association of healthy lifestyle including a healthy sleep pattern with incident type 2 diabetes mellitus among individuals with hypertension. Cardiovasc. Diabetol. 2021;20:239. doi: 10.1186/s12933-021-01434-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Spiegel K., et al. Brief communication: sleep curtailment in healthy young men is associated with decreased leptin levels, elevated ghrelin levels, and increased hunger and appetite. Ann. Intern. Med. 2004;141:846–850. doi: 10.7326/0003-4819-141-11-200412070-00008. [DOI] [PubMed] [Google Scholar]
- St-Onge M.-P., 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–e386. doi: 10.1161/CIR.0000000000000444. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Tang Y., et al. Interaction of sleep quality and sleep duration on glycemic control in patients with type 2 diabetes mellitus. Chin. Med. J. 2014;127:3543–3547. [PubMed] [Google Scholar]
- Tuomilehto H., et al. Sleep duration is associated with an increased risk for the prevalence of type 2 diabetes in middle-aged women - the FIN-D2D survey. Sleep Med. 2008;9:221–227. doi: 10.1016/j.sleep.2007.04.015. [DOI] [PubMed] [Google Scholar]
- van Leeuwen W.M.A., et al. Sleep restriction increases the risk of developing cardiovascular diseases by augmenting proinflammatory responses through IL-17 and CRP. PLoS One. 2009;4 doi: 10.1371/journal.pone.0004589. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wu Y., et al. Risk factors contributing to type 2 diabetes and recent advances in the treatment and prevention. Int. J. Med. Sci. 2014;11:1185–1200. doi: 10.7150/ijms.10001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Yan B., et al. The association between sleep efficiency and diabetes mellitus in community-dwelling individuals with or without sleep-disordered breathing. J. Diabetes. 2020;12:215–223. doi: 10.1111/1753-0407.12987. [DOI] [PubMed] [Google Scholar]
- Yu X., et al. Construction of a depression risk prediction model for type 2 diabetes mellitus patients based on NHANES 2007-2014. J. Affect. Disord. 2024;349:217–225. doi: 10.1016/j.jad.2024.01.083. [DOI] [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
The datasets used and analysed during the current study are available in the NHANES (http://www.cdc.gov/nchs/nhanes.htm).