Summary
Background
Despite emerging evidence, the causal relationship between insulin resistance and depression remains controversial. This study aimed to investigate whether insulin resistance is associated with increased risk of incident depression and whether the association is affected by potential moderators.
Methods
This multi-centered prospective cohort study analyzed health screening data from 233,452 Korean adults participating in the Kangbuk Samsung Health Study from 2011 to 2022. At baseline, all participants indicated no major psychiatric or neurologic disorders and had not used antidepressant or diabetes medications. Insulin resistance was assessed using the homeostasis model assessment of insulin resistance (HOMA-IR). Incident depression was defined as having a Center for Epidemiologic Studies Depression Scale score of ≥16.
Findings
Participants (age = 36.1 ± 8.6 years, 54.1% male) were followed up for 4.8 ± 2.9 years. During the 1,124,268 person-years of follow-up duration, 38,801 cases of incident depression were identified. Multivariate Cox proportional hazards analysis revealed a positive dose-dependent association between HOMA-IR level and the risk of incident depression (hazard ratio [HR] for highest vs. lowest quartile = 1.15, 95% confidence interval [CI] = 1.11–1.19). This association was particularly strong in younger adults under 40 years and in individuals with euglycemia, overweight, and low muscle-to-fat ratio.
Interpretation
Insulin resistance may be a modifiable risk factor for depression, underscoring the importance of early screening and management of insulin resistance to potentially reduce the burden of depression, especially among at-risk subgroups.
Funding
None.
Keywords: Depression, Insulin resistance, Metabolic syndrome, Cohort study
Research in context.
Evidence before this study
We searched PubMed, Scopus, Web of Science, and Embase for studies published until December 2024 investigating the relationship between insulin resistance and depression risk. The search terms included “insulin resistance,” “diabetes,” “prediabetes,” “metabolic syndrome,” “depression,” “depressive,” “longitudinal,” and “cohort study.” Several studies have suggested a link between insulin resistance and depression, but findings have been inconsistent. Cross-sectional studies demonstrated a positive association, but longitudinal studies provided conflicting results, possibly due to variations in population characteristics, definitions of depression, and failure to account for potential effect modifiers such as glycemic status and adiposity.
Added value of this study
This large-scale prospective cohort study analyzed data from 233,452 Korean adults followed for 1,124,268 person-years. This study identified a significant, dose-dependent relationship between insulin resistance and the risk of incident depression. The association was notably pronounced in younger adults, individuals with euglycemia, and those with higher adiposity. These findings highlight the importance of considering demographic factors, metabolic status, and anthropometric factors in understanding the insulin resistance–depression link.
Implications of all the available evidence
These findings suggest that insulin resistance is a modifiable risk factor for depression and support early metabolic screening and intervention as a potential strategy to mitigate depression risk. Future studies should further investigate the underlying mechanisms and assess whether targeted interventions to improve insulin sensitivity can reduce the incidence of depression.
Introduction
Depression is a leading cause of disability particularly in the Western Pacific region, where the burden of mental health disorders is increasing. According to the Global Burden of Disease Study 2021, depressive disorders accounted for approximately 10.5 million disability-adjusted life years (DALYs) in the Western Pacific region, ranking 12th among all non-communicable diseases and 1st among all mental disorders in terms of disease burden.1 Early identification of individuals at risk for developing depression is essential to reduce the disease burden and improve quality of life. In addition to psychosocial risk factors, further research into modifiable biological factors associated with depression is required to devise preventive strategies for at-risk groups.
Emerging evidence suggests that depression may also be associated with metabolic dysregulations, particularly insulin resistance.2 Insulin resistance is defined as a diminished sensitivity to the actions of insulin on glucose uptake in insulin-targeted tissues.3 Along with skeletal muscle and adipose tissue, the brain—a major consumer of glucose—is sensitive to insulin’s effects. Insulin plays a critical role in regulating glucose uptake and metabolism in the brain, which, in turn, influences essential brain functions such as mood regulation, cognition, and neuroplasticity.4,5 Animal studies have revealed that insulin resistance could induce increased neuroinflammation, decreased neurogenesis, and impaired mitochondrial function, as well as increased depression-like behaviors.6
Previous studies have demonstrated a positive cross-sectional association between insulin resistance and depression.7, 8, 9 However, longitudinal studies have shown inconsistent findings regarding the correlation between insulin resistance and depression risk; some studies have linked insulin resistance to increased depression risk,10, 11, 12 while others have not.13, 14, 15 These discrepancies may stem from several limitations. First, many studies focus on specific demographics, such as youth,11 middle-aged or older adults,10,13, 14, 15 or men only,10,15 despite a significant variation in insulin resistance and depression risk profiles by age and sex.16, 17, 18 Additionally, previous studies have tended to overlook the potential moderating effect of the glycemic status. In the stages of prediabetes or type 2 diabetes mellitus (T2DM), insulin resistance may attain a plateau potentially obscuring the relationship with depression risk.19 Finally, the specific role of obesity or adiposity in the relationship between insulin resistance and depression has not been thoroughly analyzed. As obesity or adiposity can be a shared risk factor for both conditions, accounting for these factors is essential to isolate the unique impact of insulin resistance on depression.20
Addressing these limitations could provide greater clarity on the potential relationship between insulin resistance and depression. Using prospective data from a large cohort of Korean adults with diverse demographics, glycemic statuses, and anthropometric profiles, this study aims to determine whether insulin resistance is associated with an increased risk of developing depression. Additionally, the study examines whether the association between insulin resistance and depression risk is moderated by age, sex, glycemic status, and obesity/adiposity.
Methods
Study design, setting, and participants
As part of the Kangbuk Samsung Health Study, this prospective cohort study gathered data from Korean adults who underwent annual or biannual health screenings at the Kangbuk Samsung Total Healthcare Centers in Seoul and Suwon, Republic of Korea.21,22 We initially included 270,977 participants who underwent a comprehensive health examination including assessments for mood status from January 2011 to December 2021 and at least one follow-up assessment before the end of 2022. We excluded 33,715 participants who had Center for Epidemiologic Studies Depression Scale (CES-D) scores of ≥16 and exhibited major psychiatric/neurologic disorders, antidepressant use, and diabetes under medication at the baseline assessment. Additionally, we excluded 3810 participants who had missing baseline data on laboratory variables such as fasting glucose, fasting insulin, glycated hemoglobin (HbA1c), homeostasis model assessment of insulin resistance (HOMA-IR), and lipid panel, as well as on health-related behaviors (alcohol consumption, current smoking, physical activity), medical history (history of cardiovascular diseases, previous history of depression), and anthropometric measures including body mass index (BMI) and muscle-to-fat ratio (MFR). Finally, 233,452 eligible participants (126,387 male and 107,065 female) were included (see Fig. 1).
Fig. 1.
Flowchart of the study participants’ selection process. CES-D, Center for Epidemiologic Studies Depression Scale; HOMA-IR, homeostasis model assessment of insulin resistance; BMI, body mass index.
Assessment of glycemic status and insulin resistance
Venous blood samples were collected after fasting at least 12 h. Fasting serum glucose and HbA1c were measured using a hexokinase method on a Cobas Integra 800 apparatus (Roche Diagnostics, Basel, Switzerland) and an immunoturbidimetric assay with the Cobas Integra 800 automatic analyzer (Roche Diagnostics, Basel, Switzerland), respectively. Fasting serum insulin was measured using an immunoradiometric assay (Biosource, Nivelles, Belgium). Diabetes was defined as (a) fasting serum glucose ≥126 mg/dL, (b) serum HbA1c level ≥6.5%, (c) self-reported history of diabetes diagnosed by physician, or (d) current use of any diabetes medication.23 As noted above, participants having any diabetes medication were excluded from analyses. Among the participants without diabetes, prediabetes was defined as either (a) fasting serum glucose 100–125 mg/dL, or (b) serum HbA1c level 5.7%–6.4%.23
Insulin resistance at the baseline assessment was assessed using HOMA-IR obtained via the following formula: HOMA-IR = fasting serum insulin (uIU/mL) × fasting serum glucose (mg/dl)/405.24 The Laboratory Medicine Department has been accredited and participates annually in inspections and surveys by the Korean Association of Quality Assurance for Clinical Laboratories. Participants were categorized into quartile groups according to the cut-off values for HOMA-IR as follows: 1st quartile, HOMA-IR <0.8628; 2nd quartile, HOMA-IR 0.8628–1.3004; 3rd quartile, HOMA-IR 1.3005–1.9386; 4th quartile, HOMA-IR ≥1.9387.
Assessment of depression
Depression was assessed at the baseline and follow-up assessments using the 20-item CES-D, a self-reported questionnaire to assess the frequency and severity of depressive symptoms. The CES-D has been widely utilized and validated in community-based and primary care settings.25 Among participants without depression at baseline, incident depression was defined as the first occurrence of a CES-D score ≥16 during follow-up, indicating elevated risk for clinically significant depression.26
Assessment of covariates
All participants were asked to respond to self-reported questionnaires about health-related behaviors and medical histories. Health-related behaviors included alcohol consumption, current smoking, and physical activity. Alcohol consumption was assessed based on the average amount of alcohol intake per week. Average alcohol consumption (g/day) was calculated using self-reported drinking frequency and both the type and amount of alcohol consumed per drinking day, based on the grams of ethanol contained in a standard drink—for example, one jan (a typical unit of soju, the most commonly consumed alcoholic beverage in Korea) was considered to contain 8 g of ethanol.27 Physical activity was evaluated using the metabolic equivalent of task (MET) derived from the Korean-validated version of the International Physical Activity Questionnaire Short Form.28 Serum triglycerides and high-density lipoprotein (HDL) cholesterol were determined with an enzymatic colorimetric assay via the venous blood collection after 12-h fasting. Trained nurses measured the systolic and diastolic blood pressure under sitting position using a standard sphygmomanometer (53000-E2, Welch Allyn, USA) after a 5-min seated rest. The presence of hypertension was defined as (a) measured blood pressure ≥140/90 mmHg, (b) self-reported history of hypertension diagnosed by physician, or (c) current use of any antihypertensive medication.
Trained nurses measured the height and weight of all participants. BMI was calculated as a weight (kg) divided by the square of height (m2). Based on the Korean Society for the Study of Obesity guideline,29 we categorized the level of BMI as follows: normal, BMI 18.5–22.9 kg/m2; underweight, BMI < 18.5 kg/m2; overweight, BMI 23.0–24.9 kg/m2; obesity, BMI ≥ 25.0 kg/m2. Body composition was measured based on the bioimpedance analysis (InBody 3.0 and Inbody 720, Biospace Co., Seoul, Republic of Korea). MFR was calculated by dividing total muscle mass (kg) by total fat mass (kg). Participants were categorized in the low MFR group if their MFR was below the cut-off values for the lowest quartile (male < 2.57; female < 1.85).
Statistical analysis
We compared the baseline characteristics of participants using the chi-squared test for categorical variables, followed by post-hoc pairwise comparisons with Bonferroni correction, and one-way analysis of variance (ANOVA) for continuous variables, followed by Tukey’s post-hoc test, which adjusts for all pairwise comparisons using the studentized range distribution. Participants were censored at the date of incident depression (CES-D ≥16) or the date of the last follow-up assessment, whichever came first. We examined the association of insulin resistance, measured using HOMA-IR as a marker of insulin resistance severity, with the risk of incident depression using multivariable Cox proportional hazards models. HOMA-IR was assessed both by quartiles based on predefined cutoffs, with the first quartile as the reference, and by per one standard deviation (SD) increase. Models were adjusted for age (years), sex (male or female), alcohol consumption (g/day), smoking (never, former, or current smoker), physical activity (MET, minutes/week), triglycerides (mg/dL), HDL cholesterol (mg/dL), hypertension (yes or no), history of coronary artery disease (yes or no), history of stroke (yes or no), previous history of depression (yes or no), glycemic status (euglycemia, prediabetes, or diabetes), BMI (kg/m2), and MFR (normal or low).
To examine the moderating effects of age, sex, glycemic status, and anthropometric measures on the association between insulin resistance and incident depression risk, we entered the interaction terms of those variables with HOMA-IR into the multivariate Cox proportional hazard analyses. For interaction analyses, age (in years) and BMI were modeled as continuous variables, while sex (male or female), glycemia status (euglycemia or dysglycemia), and MFR (normal or low) were treated as binary variables. Next, we performed analyses with the stratification of participants according to those variables. As the interaction between sex and HOMA-IR was not statistically significant, sex-stratified analysis was not conducted. To account for multiple comparisons in the subgroup analyses, we applied the Bonferroni correction. Given that 13 subgroup analyses were performed—based on combinations of age (18–39 years or ≥40 years), sex (male or female), glycemic status (euglycemia, prediabetes, or diabetes), body mass index (normal, underweight, overweight, or obesity), and muscle-to-fat ratio (normal or low)—a Bonferroni-adjusted threshold of p < 0.0038 (0.05/13) was used to determine statistical significance in these analyses. As a sensitivity analysis for missing data, multiple imputation was performed using chained equations, generating 10 imputed datasets. Comparisons of participants’ baseline characteristics by the glycemic status (eTable 1), BMI (eTable 2), and MFR (eTable 3) are discussed elsewhere. All statistical analyses were performed using IBM SPSS Statistics, version 28.0 (IBM Corporation).
Ethics approval
This study was approved by the Institutional Review Board of the Kangbuk Samsung Hospital (IRB no. KBSMC 2023-12-048) and exempted from the requirement of informed consent owing to the use of anonymized de-identified data routinely collected during health screening examinations.
Role of the funding source
There was no funding source for this study.
Results
Table 1 presents the baseline characteristics of the 233,452 participants according to the quartile groups of HOMA-IR. The higher HOMA-IR quartile groups were more likely to be male and older, and exhibited greater alcohol consumption, smoking, physical activity, higher triglyceride, lower HDL-cholesterol, and higher rates of coronary heart disease and stroke than the lower HOMA-IR quartile groups. Compared with the lower quartile groups, the higher HOMA-IR quartile group had higher rates of prediabetes and T2DM, higher fasting glucose/insulin, HbA1c, BMI, and MFR. The mean CES-D total score was slightly higher in the lowest HOMA-IR quartile group than in the others quartile groups.
Table 1.
Participants’ characteristics at the baseline assessment.e
| Total participants (n = 233,452) | By the quartiles of HOMA-IR levelf |
||||||
|---|---|---|---|---|---|---|---|
| 1st quartilea (n = 58,370) | 2nd quartileb (n = 58,325) | 3rd quartilec (n = 58,384) | 4th quartiled (n = 58,373) | p | post-hoc | ||
| Age, years, mean (SD) | 36.1 (8.6) | 36.1 (8.5) | 35.8 (8.4) | 35.9 (8.6) | 36.7 (9.0) | <0.001 | b,c < a < d |
| Male, n (%) | 126,387 (54.1) | 26,432 (45.3) | 29,323 (50.3) | 32,501 (55.7) | 38,131 (65.3) | <0.001 | a < b < c < d |
| Alcohol consumption, g/week, mean (SD) | 272.5 (1059.9) | 222.5 (898.6) | 244.3 (981.7) | 280.2 (1141.9) | 342.3 (1186.2) | <0.001 | a < b < c < d |
| Current smoking, n (%) | 41,453 (18.4) | 8878 (15.9) | 9418 (16.7) | 10,372 (18.3) | 12,785 (22.4) | <0.001 | a < b < c < d |
| Physical activity, MET, mean (SD) | 1548.3 (3015.2) | 1709.5 (3353.9) | 1553.2 (2975.3) | 1480.3 (2904.3) | 1450.7 (2791.8) | <0.001 | a > b > c,d |
| Triglyceride, mg/dL, mean (SD) | 107.4 (76.3) | 73.1 (37.2) | 91.8 (50.6) | 111.0 (67.5) | 153.7 (105.9) | <0.001 | a < b < c < d |
| HDL-cholesterol, mg/dL, mean (SD) | 60.4 (15.9) | 66.5 (15.6) | 62.8 (15.4) | 59.3 (15.2) | 52.9 (14.1) | <0.001 | a > b > c > d |
| Hypertension, n (%) | 22,007 (9.4) | 2666 (4.6) | 3777 (6.5) | 5230 (9.0) | 10,334 (17.7) | <0.001 | a < b < c < d |
| History of coronary heart disease, n (%) | 1338 (0.5) | 247 (0.4) | 245 (0.4) | 278 (0.5) | 368 (0.6) | <0.001 | a,b,c < d |
| History of stroke, n (%) | 752 (0.3) | 171 (0.3) | 168 (0.3) | 190 (0.3) | 223 (0.4) | 0.017 | b < d |
| Previous history of depression, n (%) | 1867 (0.8) | 499 (0.9) | 445 (0.8) | 441 (0.8) | 482 (0.8) | 0.159 | – |
| CES-D total score, mean (SD) | 4.90 (4.19) | 4.98 (4.19) | 4.91 (4.19) | 4.85 (4.19) | 4.85 (4.17) | <0.001 | a > b,c,d |
| Glycemic status | |||||||
| Prediabetes, n (%) | 73,558 (31.5) | 11,229 (19.2) | 14,331 (24.6) | 19,386 (33.2) | 28,612 (49.0) | <0.001 | a < b < c < d |
| Diabetes, n (%) | 6710 (2.9) | 423 (0.7) | 669 (1.1) | 1056 (1.8) | 4562 (7.8) | <0.001 | a < b < c < d |
| Fasting serum glucose, mg/dL, mean (SD) | 94.2 (13.5) | 87.2 (8.0) | 92.0 (7.7) | 95.0 (9.1) | 102.5 (20.0) | <0.001 | a < b < c < d |
| HbA1c, %, mean (SD) | 5.50 (0.46) | 5.42 (0.30) | 5.44 (0.31) | 5.50 (0.36) | 5.67 (0.71) | <0.001 | a < b < c < d |
| Fasting serum insulin, uIU/mL, mean (SD) | 6.61 (4.84) | 2.79 (0.80) | 4.77 (0.63) | 6.80 (0.91) | 12.06 (6.65) | <0.001 | a < b < c < d |
| HOMA-IR, mean (SD) | 1.58 (1.32) | 0.60 (0.18) | 1.08 (0.13) | 1.59 (0.18) | 3.06 (1.87) | <0.001 | a < b < c < d |
| BMI, kg/m2, mean (SD) | 23.3 (3.5) | 21.4 (2.5) | 22.4 (2.8) | 23.4 (3.1) | 25.9 (3.8) | <0.001 | a < b < c < d |
| MFR, mean (SD) | 2.92 (1.17) | 3.36 (1.41) | 3.05 (1.17) | 2.82 (0.99) | 2.45 (0.83) | <0.001 | a > b > c > d |
HOMA-IR, homeostasis model assessment of insulin resistance; MET, metabolic equivalent task; HDL, high-density lipoprotein; CES-D, center for epidemiologic studies depression scale; HbA1c, glycated hemoglobin; BMI, body mass index; MFR, muscle-to-fat ratio.
HOMA-IR 1st quartile group.
HOMA-IR 2nd quartile group.
HOMA-IR 3rd quartile group.
HOMA-IR 4th quartile group.
Comparison between quartile groups using chi-squared test for categorical variables and analysis of variance for continuous variables.
1st quartile, HOMA-IR < 0.8628; 2nd quartile, HOMA-IR 0.8628–1.3004; 3rd quartile, HOMA-IR 1.3005–1.9386; 4th quartile, HOMA-IR ≥ 1.9387.
During the 1,124,268 person-years (4.8 ± 2.9 years) of follow-up duration, 38,801 cases of incident depression were identified. The incidence rates of depression showed a dose-dependent relationship with the HOMA-IR levels; 33.4 per 1000 person-years in the 1st HOMA-IR quartile group, 34.3 per 1000 person-years in the 2nd HOMA-IR quartile group, 33.3 per 1000 person-years in the 3rd HOMA-IR quartile group, and 34.6 per 1000 person-years in the 4th HOMA-IR quartile group.
Multivariate Cox proportional hazard analyses showed that the 2nd, 3rd, and 4th HOMA-IR quartile groups had 5%, 6%, and 15% increased risks of incident depression, respectively, compared with the lowest quartile group. A 1 SD-increase of HOMA-IR was associated with an 8% increased risk of incident depression (Table 2). The association between HOMA-IR and the risk of depression was moderated by age (HR = 1.001, 95% CI = 1.000–1.002, p = 0.004 for the age ∗ HOMA-IR interaction term) but not by sex. Compared with the lowest quartile, the highest quartile of HOMA-IR had 14% increased risk of incident depression in young adults. Compared with the lowest quartile, the highest quartile of HOMA-IR had 11% and 16% increased risk of incident depression in male and female young adults, respectively. The association between HOMA-IR and the risk of incident depression was attenuated in middle-aged adults (Table 2). Results from sensitivity analyses using multiple imputation were consistent with the primary findings (eTable 4).
Table 2.
Association of insulin resistance with the risk of incident depression by age and sex.
| Person-years | n. of cases | HR (95% CI) | p-value | |
|---|---|---|---|---|
| Total (n = 233,452) | ||||
| HOMA-IR 1st quartile | 305,698 | 10,219 | 1 [reference] | – |
| HOMA-IR 2nd quartile | 286,820 | 9838 | 1.05 (1.02–1.08) | <0.001 |
| HOMA-IR 3rd quartile | 276,171 | 9195 | 1.06 (1.03–1.09) | <0.001 |
| HOMA-IR 4th quartile | 255,406 | 8829 | 1.15 (1.11–1.19) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.08 (1.06–1.10) | <0.001 |
| 18–39 years old (n = 171,227) | ||||
| HOMA-IR 1st quartile | 240,468 | 8295 | 1 [reference] | |
| HOMA-IR 2nd quartile | 227,552 | 8008 | 1.05 (1.02–1.08) | 0.002 |
| HOMA-IR 3rd quartile | 216,995 | 7492 | 1.06 (1.02–1.09) | <0.001 |
| HOMA-IR 4th quartile | 191,379 | 6888 | 1.14 (1.10–1.19) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.08 (1.06–1.10) | <0.001 |
| 18–39 years old, male (n = 90,301) | ||||
| HOMA-IR 1st quartile | 109,433 | 2875 | 1 [reference] | |
| HOMA-IR 2nd quartile | 114,734 | 3265 | 1.08 (1.03–1.14) | 0.003 |
| HOMA-IR 3rd quartile | 121,307 | 3355 | 1.04 (0.99–1.09) | 0.186 |
| HOMA-IR 4th quartile | 125,397 | 3761 | 1.11 (1.04–1.18) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.08 (1.04–1.12) | <0.001 |
| 18–39 years old, female (n = 80,926) | ||||
| HOMA-IR 1st quartile | 131,035 | 5420 | 1 [reference] | |
| HOMA-IR 2nd quartile | 112,817 | 4743 | 1.03 (0.99–1.07) | 0.208 |
| HOMA-IR 3rd quartile | 95,689 | 4137 | 1.06 (1.02–1.11) | 0.003 |
| HOMA-IR 4th quartile | 65,983 | 3127 | 1.16 (1.11–1.23) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.08 (1.05–1.11) | <0.001 |
| ≥40 years old (n = 62,225) | ||||
| HOMA-IR 1st quartile | 65,230 | 1924 | 1 [reference] | |
| HOMA-IR 2nd quartile | 59,268 | 1830 | 1.06 (1.00–1.14) | 0.063 |
| HOMA-IR 3rd quartile | 59,176 | 1703 | 1.02 (0.96–1.10) | 0.500 |
| HOMA-IR 4th quartile | 64,027 | 1941 | 1.12 (1.03–1.21) | 0.015 |
| HOMA-IR per 1SD increase | – | – | 1.07 (1.02–1.12) | 0.003 |
| ≥40 years old, male (n = 36,086) | ||||
| HOMA-IR 1st quartile | 32,661 | 788 | 1 [reference] | |
| HOMA-IR 2nd quartile | 32,875 | 800 | 0.99 (0.89–1.10) | 0.854 |
| HOMA-IR 3rd quartile | 36,189 | 853 | 0.96 (0.86–1.06) | 0.411 |
| HOMA-IR 4th quartile | 44,479 | 1186 | 1.08 (0.96–1.21) | 0.191 |
| HOMA-IR per 1SD increase | – | – | 1.07 (1.00–1.15) | 0.044 |
| ≥40 years old, female (n = 26,139) | ||||
| HOMA-IR 1st quartile | 32,570 | 1136 | 1 [reference] | |
| HOMA-IR 2nd quartile | 26,484 | 1030 | 1.12 (1.03–1.22) | 0.012 |
| HOMA-IR 3rd quartile | 22,986 | 850 | 1.07 (0.98–1.18) | 0.138 |
| HOMA-IR 4th quartile | 19,549 | 755 | 1.13 (1.01–1.26) | 0.030 |
| HOMA-IR per 1SD increase | – | – | 1.08 (1.01–1.14) | 0.020 |
HR, hazard ratio; CI, confidence interval; HOMA-IR, homeostasis model assessment of insulin resistance.
Multivariate Cox proportional hazard analyses were performed, adjusted for age (years), sex (male or female), alcohol consumption (g/day), smoking (never, former, or current smoker), physical activity (MET, minutes/week), triglycerides (mg/dL), HDL cholesterol (mg/dL), hypertension (yes or no), history of coronary artery disease (yes or no), history of stroke (yes or no), previous history of depression (yes or no), glycemic status (euglycemia, prediabetes, or diabetes), body mass index (kg/m2), and muscle-to-fat ratio (normal or low).
p-values < 0.0038 (Bonferroni-adjusted threshold for multiple comparisons) were considered statistically significant and are shown in bold.
The association of HOMA-IR levels with the risk of incident depression was moderated by the glycemic status (HR = 1.04, 95% CI = 1.00–1.07, p < 0.001 for the glycemic status ∗ HOMA-IR interaction term). In euglycemic participants, the 3rd and 4th HOMA-IR quartile groups exhibited 6% and 16% increased risk of incident depression, respectively, compared with the lowest quartile group. A 1 SD-increase of HOMA-IR was associated with a 9% increased risk of incident depression in euglycemia. The association between HOMA-IR and the risk of incident depression was attenuated in prediabetes, whereas the association was not significant in T2DM (Table 3). The findings remained robust in sensitivity analyses using multiple imputation (eTable 5).
Table 3.
Association of insulin resistance with the risk of incident depression by glycemic status.
| Person-years | n. of cases | HR (95% CI) | p-value | |
|---|---|---|---|---|
| Euglycemia (n = 153,184) | ||||
| HOMA-IR 1st quartile | 238,891 | 8153 | 1 [reference] | – |
| HOMA-IR 2nd quartile | 207,981 | 7362 | 1.05 (1.01–1.09) | 0.006 |
| HOMA-IR 3rd quartile | 175,154 | 6068 | 1.06 (1.03–1.10) | <0.001 |
| HOMA-IR 4th quartile | 109,680 | 4105 | 1.16 (1.12–1.22) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.09 (1.06–1.11) | <0.001 |
| Prediabetes (n = 73,558) | ||||
| HOMA-IR 1st quartile | 65,060 | 2025 | 1 [reference] | – |
| HOMA-IR 2nd quartile | 76,029 | 2397 | 1.03 (0.97–1.10) | 0.295 |
| HOMA-IR 3rd quartile | 96,647 | 3012 | 1.05 (0.99–1.12) | 0.101 |
| HOMA-IR 4th quartile | 127,783 | 4137 | 1.09 (1.03–1.16) | 0.006 |
| HOMA-IR per 1SD increase | – | – | 1.06 (1.02–1.10) | 0.002 |
| Diabetes (n = 6710) | ||||
| HOMA-IR 1st quartile | 1748 | 10,219 | 1 [reference] | – |
| HOMA-IR 2nd quartile | 2810 | 9838 | 1.14 (0.78–1.67) | 0.503 |
| HOMA-IR 3rd quartile | 4370 | 9195 | 1.09 (0.75–1.57) | 0.658 |
| HOMA-IR 4th quartile | 17,943 | 8829 | 1.27 (0.90–1.80) | 0.178 |
| HOMA-IR per 1SD increase | – | – | 1.12 (0.99–1.26) | 0.066 |
HR, hazard ratio; CI, confidence interval; HOMA-IR, homeostasis model assessment of insulin resistance.
Multivariate Cox proportional hazard analyses were performed, adjusted for age (years), sex (male or female), alcohol consumption (g/day), smoking (never, former, or current smoker), physical activity (MET, minutes/week), triglycerides (mg/dL), HDL cholesterol (mg/dL), hypertension (yes or no), history of coronary artery disease (yes or no), history of stroke (yes or no), previous history of depression (yes or no), body mass index (kg/m2), and muscle-to-fat ratio (normal or low).
p-values < 0.0038 (Bonferroni-adjusted threshold for multiple comparisons) were considered statistically significant and are shown in bold.
The association of HOMA-IR levels with the risk of incident depression was moderated by BMI (HR = 1.004, 95% CI = 1.002–1.005, p < 0.001 for the BMI ∗ HOMA-IR interaction term). As shown in Table 4, HOMA-IR had a dose-dependent association with the risk of incident depression regardless of BMI level. However, the association was prominent in overweight participants; the 2nd and 4th HOMA-IR quartile groups had 15% and 21% increased risk of incident depression, respectively, compared with the lowest quartile group. A 1 SD-increase of HOMA-IR was associated with a 12% increased risk of incident depression in the overweight group (Table 4). The findings remained robust in sensitivity analyses using multiple imputation (eTable 6).
Table 4.
Association of insulin resistance with the risk of incident depression by body mass index.a
| Person-years | n. of cases | HR (95% CI) | p-value | |
|---|---|---|---|---|
| Normal (n = 103,260) | ||||
| HOMA-IR 1st quartile | 190,098 | 6589 | 1 [reference] | |
| HOMA-IR 2nd quartile | 150,918 | 5419 | 1.04 (1.00–1.08) | 0.047 |
| HOMA-IR 3rd quartile | 112,382 | 4082 | 1.05 (1.01–1.10) | 0.013 |
| HOMA-IR 4th quartile | 52,897 | 2021 | 1.13 (1.07–1.19) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.06 (1.03–1.09) | <0.001 |
| Underweight (n = 14,090) | ||||
| HOMA-IR 1st quartile | 35,086 | 1386 | 1 [reference] | |
| HOMA-IR 2nd quartile | 19,954 | 818 | 1.03 (0.95–1.13) | 0.486 |
| HOMA-IR 3rd quartile | 12,228 | 506 | 1.03 (0.92–1.14) | 0.627 |
| HOMA-IR 4th quartile | 3647 | 166 | 1.16 (0.98–1.37) | 0.085 |
| HOMA-IR per 1SD increase | – | – | 1.05 (0.98–1.12) | 0.139 |
| Overweight (n = 49,719) | ||||
| HOMA-IR 1st quartile | 53,045 | 1470 | 1 [reference] | |
| HOMA-IR 2nd quartile | 64,452 | 2057 | 1.15 (1.08–1.23) | <0.001 |
| HOMA-IR 3rd quartile | 68,363 | 2031 | 1.08 (1.01–1.16) | 0.029 |
| HOMA-IR 4th quartile | 51,956 | 1738 | 1.21 (1.12–1.31) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.12 (1.07–1.17) | <0.001 |
| Obesity (n = 66,383) | ||||
| HOMA-IR 1st quartile | 27,470 | 774 | 1 [reference] | |
| HOMA-IR 2nd quartile | 51,496 | 1544 | 1.07 (0.98–1.16) | 0.158 |
| HOMA-IR 3rd quartile | 83,198 | 2576 | 1.09 (1.00–1.18) | 0.046 |
| HOMA-IR 4th quartile | 146,906 | 4904 | 1.13 (1.04–1.23) | 0.003 |
| HOMA-IR per 1SD increase | – | – | 1.08 (1.04–1.13) | <0.001 |
HR, hazard ratio; CI, confidence interval; HOMA-IR, homeostasis model assessment of insulin resistance.
Multivariate Cox proportional hazard analyses were performed, adjusted for age (years), sex (male or female), alcohol consumption (g/day), smoking (never, former, or current smoker), physical activity (MET, minutes/week), triglycerides (mg/dL), HDL cholesterol (mg/dL), hypertension (yes or no), history of coronary artery disease (yes or no), history of stroke (yes or no), previous history of depression (yes or no), glycemic status (euglycemia, prediabetes, or diabetes), and muscle-to-fat ratio (normal or low).
p-values < 0.0038 (Bonferroni-adjusted threshold for multiple comparisons) were considered statistically significant and are shown in bold.
Normal, body mass index (BMI) 18.5–22.9 kg/m2; underweight, BMI < 18.5 kg/m2; overweight, BMI 23.0–24.9 kg/m2; obesity, BMI ≥25.0 kg/m2.
The association of HOMA-IR levels with the risk of incident depression was moderated by MFR (HR = 1.12, 95% CI = 1.08–1.15, p < 0.001 for the MFR ∗ HOMA-IR interaction term <0.001). As shown in Table 5, HOMA-IR had a dose-dependent association with the risk of incident depression regardless of MFR level. However, the association was prominent in participants with low MFR; the 2nd, 3rd, and 4th HOMA-IR quartile groups had 15%, 16%, and 22% increased risk of incident depression, respectively, compared with the lowest quartile group. A 1 SD-increase of HOMA-IR was associated with a 13% increased risk of incident depression in the low MFR group (Table 5). The findings remained robust in sensitivity analyses using multiple imputation (eTable 7).
Table 5.
Association of insulin resistance with the risk of incident depression by muscle-to-fat ratio.a
| Person-years | n. of cases | HR (95% CI) | p-value | |
|---|---|---|---|---|
| Normal MFR (n = 174,548) | ||||
| HOMA-IR 1st quartile | 277,444 | 9237 | 1 [reference] | – |
| HOMA-IR 2nd quartile | 241,618 | 8142 | 1.05 (1.02–1.08) | 0.003 |
| HOMA-IR 3rd quartile | 206,782 | 6688 | 1.04 (1.01–1.08) | 0.017 |
| HOMA-IR 4th quartile | 132,125 | 4311 | 1.13 (1.09–1.18) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.06 (1.04–1.09) | <0.001 |
| Low MFR (n = 58,904) | ||||
| HOMA-IR 1st quartile | 27,465 | 945 | 1 [reference] | – |
| HOMA-IR 2nd quartile | 44,239 | 1656 | 1.15 (1.06–1.25) | <0.001 |
| HOMA-IR 3rd quartile | 68,565 | 2478 | 1.16 (1.07–1.25) | <0.001 |
| HOMA-IR 4th quartile | 122,513 | 4472 | 1.22 (1.12–1.31) | <0.001 |
| HOMA-IR per 1SD increase | – | – | 1.13 (1.08–1.18) | <0.001 |
HR, hazard ratio; CI, confidence interval; MFR, muscle-to-fat ratio; HOMA-IR, homeostasis model assessment of insulin resistance.
Multivariate Cox proportional hazard analyses were performed, adjusted for age (years), sex (male or female), alcohol consumption (g/day), smoking (never, former, or current smoker), physical activity (MET, minutes/week), triglycerides (mg/dL), HDL cholesterol (mg/dL), hypertension (yes or no), history of coronary artery disease (yes or no), history of stroke (yes or no), previous history of depression (yes or no), glycemic status (euglycemia, prediabetes, or diabetes), and body mass index (kg/m2).
p-values < 0.0038 (Bonferroni-adjusted threshold for multiple comparisons) were considered statistically significant and are shown in bold.
Low MFR, male < 2.57, female < 1.85.
Discussion
This study demonstrated that insulin resistance is associated with an increased risk of incident depression in a dose-dependent manner, even after adjusting for various confounders. Additionally, this association was moderated by factors such as age, glycemic status, and anthropometric profiles.
The strong association between insulin resistance and depression risk observed in this study may be explained by several mechanisms. Insulin resistance can activate the hypothalamic-pituitary-adrenal axis, leading to elevated cortisol levels, which may increase depression risk.6,30 Additionally, insulin resistance can promote neuroinflammation causing depression by dysregulation of neurotransmitters and decreased neurogenesis.31 Our findings also suggest a potential role of insulin action in the brain in the development of depression. Insulin receptors are present in neurons and glial cells in the brain, particularly in mood-regulating regions such as the prefrontal cortex, hippocampus, amygdala, and dorsal striatum.32,33 Insulin signaling in these regions influences mood regulation; animal studies have shown that blocking insulin receptors can induce depression-like behaviors in mice.34 However, the role of brain insulin resistance in human depression remains vague. Further studies, including neuroimaging and experimental research, are required to better understand this mechanism.
This study identified that insulin resistance was associated with an increased risk of developing depression, particularly in younger adults. Most previous longitudinal studies with conflicting results have solely focused on middle-aged or older adults,10,13, 14, 15 which, in this study, showed a weaker association between insulin resistance and depression. The Netherlands Study of Depression and Anxiety,12 which analyzed 601 adults aged 18–65 years (mean 40.9 ± 14.5 years), also reported a positive association between insulin resistance and the risk of incident depression. Together with our findings, this suggests that the impact of insulin resistance on depression risk may be stronger in younger adults, possibly because other risk factors for depression—such as a range of physical illnesses outside of metabolic conditions—are less prevalent in this age group.35 In particular, neuroinflammation driven primarily by insulin resistance may play a relatively greater role in younger individuals, whereas in middle-aged and older adults, systemic inflammation may arise from multiple sources (e.g., multimorbidity, immunosenescence, and age-related oxidative stress) thereby diluting the relative contribution of insulin resistance.36 Given the rising global prevalence of insulin resistance in individuals aged 20–30 years,37,38 our findings underscore the importance of timely screening and management of insulin resistance to improve mental health outcomes in this population.
As shown in Table 3, the association between insulin resistance and incident depression was significant in individuals with euglycemia, but weaker in those with dysglycemia. This differential association may reflect underlying differences in the pathophysiological role of insulin resistance across stages of glycemic dysregulation. In the non-diabetic stage, insulin resistance is primarily driven by hyperinsulinemia and increases with disease progression. However, as individuals advance to diabetes, insulin resistance plateaus, with insulin levels decreasing and glucose levels rising progressively.19 Thus, as diabetes progresses, the neurotoxic effect of chronic hyperglycemia39,40 may play a more dominant role in depression risk than insulin resistance itself. The mixed findings in previous longitudinal studies may partly reflect the varied proportions of participants with diabetes, ranging from unknown to 0.0%–41.5%.10, 11, 12, 13, 14, 15 While most studies have statistically adjusted for diabetes, no study has analyzed the association separately by the glycemic status. By addressing that limitation, the current study was the first to reveal that the insulin resistance–depression link varies by the glycemic status.
Notably, we identified that higher insulin resistance was associated with an increased risk of incident depression even among individuals with normal BMI and MFR. This finding challenges the “common soil hypothesis,” which suggests that the link between insulin resistance and depression is primarily due to the shared influence of obesity on both conditions.20 Our study demonstrates a significant relationship between insulin resistance and depression, independent of obesity. However, we also observed that this association was more pronounced in individuals who were overweight and those with high adiposity. In conditions of high adiposity and insulin resistance, neuronal and glial cells may increasingly rely on fatty acid oxidation instead of glucose for energy production.41, 42, 43 The toxic by-products of fatty acid metabolism can damage brain cells through oxidative stress, impair mitochondrial function, and promote neuroinflammation, all of which are linked to the development of depression.42,44, 45, 46 Further research is required to clarify the biological mechanisms underlying the interaction between obesity, insulin resistance, and depression.
Despite its valuable findings, this study has some limitations. First, the cohort comprised relatively healthy Korean adults who had good access to health care services for health examinations; this may limit the generalizability of our findings to other populations. Second, incident depression was assessed using a self-reported questionnaire instead of a structured clinical interview for diagnosing major depressive disorder. Moreover, self-reported depression, along with other self-reported variables (e.g., alcohol consumption), was subject to potential recall bias. Third, we used the HOMA-IR as a measure of insulin resistance, which may be less accurate than gold standard methods such as the euglycemic glucose clamp technique. It is a known fact that the validity of HOMA-IR is reduced, particularly in individuals with advanced beta-cell dysfunction, diabetes medication, and low BMI.47,48 To raise the validity of HOMA-IR, we excluded those with diabetes medication and adjusted for the glycemic status and BMI in our analyses. Fourth, this study did not include sociodemographic variables such as educational attainment or socioeconomic status. While our primary aim was to examine the role of insulin resistance as a modifiable biological risk factor, these sociodemographic factors could act as important confounding variables. Fifth, there was a modest amount of missing data for MFR (0.6%) and BMI (0.5%), which differed significantly across HOMA-IR quartile groups. Multiple imputation was applied to address this issue, and the consistency of results across sensitivity analyses suggests that the influence of missing data on our findings was limited. Finally, due to the very large sample size of this study, some associations may have reached statistical significance despite having small effect sizes that may not necessarily indicate clinical relevance. While such findings should be interpreted with caution, even modest associations may still have public health implications at the population level, particularly for common conditions such as insulin resistance and depression. For example, although the hazard ratio for incident depression in the second HOMA-IR quartile compared with the lowest quartile was relatively modest (HR = 1.05), this corresponds to an excess incidence of approximately 872 additional cases per 100,000 person-years. Considering the substantial and rapidly growing burden of insulin resistance across the Western Pacific region, which includes an estimated 2.3 billion people, such incremental risk may translate into a considerable public health impact.49
Despite these limitations, this prospective cohort study, utilizing the largest longitudinal dataset to date, demonstrates that insulin resistance is a significant risk factor for developing depression. Importantly, this study is the first to show that the association between insulin resistance and depression risk is moderated by age, glycemic status, and adiposity. These findings may highlight the inconsistencies in previous studies, thus underscoring the necessity of further research into the neuroendocrine and metabolic pathways underlying depression. For mental health promotion, monitoring and managing insulin resistance may be beneficial, especially in younger adults with high adiposity, even if their blood glucose level is normal.
Contributors
D.J.O., S-W.J., and S.J.C. conceptualised and designed the study. All authors had full access to the data, verified the data analysis, and contributed to the interpretation of the data. D.J.O. and S-W.J. drafted the manuscript. All authors critically reviewed the manuscript. All authors approved the final version of the manuscript and had final responsibility for the decision to submit for publication.
Data sharing statement
The data will not be made publicly available owing to our institutional review board’s regulations. However, the analytical methods are available from the corresponding author upon reasonable request.
Declaration of interests
None.
Acknowledgements
None.
Footnotes
Supplementary data related to this article can be found at https://doi.org/10.1016/j.lanwpc.2025.101672.
Contributor Information
Sang-Won Jeon, Email: sangwonyda@hanmail.net.
Sung Joon Cho, Email: sjcho0812@hanmail.net.
Appendix A. Supplementary data
References
- 1.Institute for Health Metrics and Evaluation (IHME) Global burden of disease (GBD) study 2021. IHME, University of Washington; Seattle, WA: 2023. GBD compare data visualization.http://vizhub.healthdata.org/gbd-compare Available from: [Google Scholar]
- 2.Milaneschi Y., Lamers F., Berk M., Penninx B. Depression heterogeneity and its biological underpinnings: toward immunometabolic depression. Biol Psychiatry. 2020;88(5):369–380. doi: 10.1016/j.biopsych.2020.01.014. [DOI] [PubMed] [Google Scholar]
- 3.Martin B.C., Warram J.H., Krolewski A.S., Bergman R.N., Soeldner J.S., Kahn C.R. Role of glucose and insulin resistance in development of type 2 diabetes mellitus: results of a 25-year follow-up study. Lancet. 1992;340(8825):925–929. doi: 10.1016/0140-6736(92)92814-v. [DOI] [PubMed] [Google Scholar]
- 4.Kullmann S., Kleinridders A., Small D.M., et al. Central nervous pathways of insulin action in the control of metabolism and food intake. Lancet Diabetes Endocrinol. 2020;8(6):524–534. doi: 10.1016/S2213-8587(20)30113-3. [DOI] [PubMed] [Google Scholar]
- 5.Milstein J.L., Ferris H.A. The brain as an insulin-sensitive metabolic organ. Mol Metab. 2021;52 doi: 10.1016/j.molmet.2021.101234. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Mansur R.B., Brietzke E., McIntyre R.S. Is there a “metabolic-mood syndrome”? A review of the relationship between obesity and mood disorders. Neurosci Biobehav Rev. 2015;52:89–104. doi: 10.1016/j.neubiorev.2014.12.017. [DOI] [PubMed] [Google Scholar]
- 7.Fernandes B.S., Salagre E., Enduru N., Grande I., Vieta E., Zhao Z. Insulin resistance in depression: a large meta-analysis of metabolic parameters and variation. Neurosci Biobehav Rev. 2022;139 doi: 10.1016/j.neubiorev.2022.104758. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kan C., Silva N., Golden S.H., et al. A systematic review and meta-analysis of the association between depression and insulin resistance. Diabetes Care. 2013;36(2):480–489. doi: 10.2337/dc12-1442. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Behnoush A.H., Mousavi A., Ghondaghsaz E., et al. The importance of assessing the triglyceride-glucose index (TyG) in patients with depression: a systematic review. Neurosci Biobehav Rev. 2024;159 doi: 10.1016/j.neubiorev.2024.105582. [DOI] [PubMed] [Google Scholar]
- 10.Ford A.H., Flicker L., Hankey G.J., et al. Insulin resistance and depressive symptoms in older men: the health in men study. Am J Geriatr Psychiatry. 2015;23(8):872–880. doi: 10.1016/j.jagp.2014.10.010. [DOI] [PubMed] [Google Scholar]
- 11.Perry B.I., Khandaker G.M., Marwaha S., et al. Insulin resistance and obesity, and their association with depression in relatively young people: findings from a large UK birth cohort. Psychol Med. 2020;50(4):556–565. doi: 10.1017/S0033291719000308. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Watson K.T., Simard J.F., Henderson V.W., et al. Incident major depressive disorder predicted by three measures of insulin resistance: a Dutch cohort study. Am J Psychiatry. 2021;178(10):914–920. doi: 10.1176/appi.ajp.2021.20101479. [DOI] [PubMed] [Google Scholar]
- 13.Akbaraly T.N., Kumari M., Head J., et al. Glycemia, insulin resistance, insulin secretion, and risk of depressive symptoms in middle age. Diabetes Care. 2013;36(4):928–934. doi: 10.2337/dc12-0239. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Geraets A.F.J., Kohler S., Muzambi R., et al. The association of hyperglycaemia and insulin resistance with incident depressive symptoms over 4 years of follow-up: the maastricht Study. Diabetologia. 2020;63(11):2315–2328. doi: 10.1007/s00125-020-05247-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lawlor D.A., Ben-Shlomo Y., Ebrahim S., et al. Insulin resistance and depressive symptoms in middle aged men: findings from the caerphilly prospective cohort study. BMJ. 2005;330(7493):705–706. doi: 10.1136/bmj.38377.616921.F7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Villarroel M.A., Terlizzi E.P. National Center for Health Statistics; Hyattsville, MD: 2020. Symptoms of depression among adults: united States, 2019. NCHS data brief, no 379. [PubMed] [Google Scholar]
- 17.Barzilai N., Ferrucci L. Insulin resistance and aging: a cause or a protective response? J Gerontol A Biol Sci Med Sci. 2012;67(12):1329–1331. doi: 10.1093/gerona/gls145. [DOI] [PubMed] [Google Scholar]
- 18.Tramunt B., Smati S., Grandgeorge N., et al. Sex differences in metabolic regulation and diabetes susceptibility. Diabetologia. 2020;63(3):453–461. doi: 10.1007/s00125-019-05040-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Tabak A.G., Jokela M., Akbaraly T.N., Brunner E.J., Kivimaki M., Witte D.R. Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the whitehall II study. Lancet. 2009;373(9682):2215–2221. doi: 10.1016/S0140-6736(09)60619-X. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Tabak A.G., Akbaraly T.N., Batty G.D., Kivimaki M. Depression and type 2 diabetes: a causal association? Lancet Diabetes Endocrinol. 2014;2(3):236–245. doi: 10.1016/S2213-8587(13)70139-6. [DOI] [PubMed] [Google Scholar]
- 21.Chang Y., Ryu S., Cho J., Pastor-Barriuso R., Guallar E. Metabolically healthy obesity and development of chronic kidney disease. Ann Intern Med. 2016;165(10):744–745. doi: 10.7326/L16-0405. [DOI] [PubMed] [Google Scholar]
- 22.Lee J.H., Park S.K., Ryoo J.H., et al. The association between insulin resistance and depression in the Korean general population. J Affect Disord. 2017;208:553–559. doi: 10.1016/j.jad.2016.10.027. [DOI] [PubMed] [Google Scholar]
- 23.Moon J.S., Kang S., Choi J.H., et al. 2023 Clinical practice guidelines for diabetes management in Korea: full version recommendation of the Korean diabetes association. Diabetes Metab J. 2024;48(4):546–708. doi: 10.4093/dmj.2024.0249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Matthews D.R., Hosker J.P., Rudenski A.S., Naylor B.A., Treacher D.F., Turner R.C. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–419. doi: 10.1007/BF00280883. [DOI] [PubMed] [Google Scholar]
- 25.Park S.H., Yu H.Y. How useful is the center for epidemiologic studies depression scale in screening for depression in adults? An updated systematic review and meta-analysis. Psychiatry Res. 2021;302 doi: 10.1016/j.psychres.2021.114037. [DOI] [PubMed] [Google Scholar]
- 26.Weissman M.M., Sholomskas D., Pottenger M., Prusoff B.A., Locke B.Z. Assessing depressive symptoms in five psychiatric populations: a validation study. Am J Epidemiol. 1977;106(3):203–214. doi: 10.1093/oxfordjournals.aje.a112455. [DOI] [PubMed] [Google Scholar]
- 27.Chang Y., Ryu S., Kim Y., et al. Low levels of alcohol consumption, obesity and development of fatty liver with and without evidence of advanced fibrosis. Hepatology. 2019;71(3):861–873. doi: 10.1002/hep.30867. [DOI] [PubMed] [Google Scholar]
- 28.Craig C.L., Marshall A.L., Sjostrom M., et al. International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc. 2003;35(8):1381–1395. doi: 10.1249/01.MSS.0000078924.61453.FB. [DOI] [PubMed] [Google Scholar]
- 29.Seo M.H., Lee W.Y., Kim S.S., et al. 2018 Korean society for the study of obesity guideline for the management of obesity in Korea. J Obes Metab Syndr. 2019;28(1):40–45. doi: 10.7570/jomes.2019.28.1.40. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Vreeburg S.A., Hoogendijk W.J., van Pelt J., et al. Major depressive disorder and hypothalamic-pituitary-adrenal axis activity: results from a large cohort study. Arch Gen Psychiatry. 2009;66(6):617–626. doi: 10.1001/archgenpsychiatry.2009.50. [DOI] [PubMed] [Google Scholar]
- 31.Krishnadas R., Cavanagh J. Depression: an inflammatory illness? J Neurol Neurosurg Psychiatry. 2012;83(5):495–502. doi: 10.1136/jnnp-2011-301779. [DOI] [PubMed] [Google Scholar]
- 32.Kullmann S., Heni M., Fritsche A., Preissl H. Insulin action in the human brain: evidence from neuroimaging studies. J Neuroendocrinol. 2015;27(6):419–423. doi: 10.1111/jne.12254. [DOI] [PubMed] [Google Scholar]
- 33.Nijssen K.M.R., Mensink R.P., Joris P.J. Effects of intranasal insulin administration on cerebral blood flow and cognitive performance in adults: a systematic review of randomized, placebo-controlled intervention studies. Neuroendocrinology. 2023;113(1):1–13. doi: 10.1159/000526717. [DOI] [PubMed] [Google Scholar]
- 34.Lyra E.S.N.M., Lam M.P., Soares C.N., Munoz D.P., Milev R., De Felice F.G. Insulin resistance as a shared pathogenic mechanism between depression and type 2 diabetes. Front Psychiatry. 2019;10:57. doi: 10.3389/fpsyt.2019.00057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Alexopoulos G.S. Mechanisms and treatment of late-life depression. Transl Psychiatry. 2019;9(1):188. doi: 10.1038/s41398-019-0514-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Franceschi C., Garagnani P., Parini P., Giuliani C., Santoro A. Inflammaging: a new immune–metabolic viewpoint for age-related diseases. Nat Rev Endocrinol. 2018;14(10):576–590. doi: 10.1038/s41574-018-0059-4. [DOI] [PubMed] [Google Scholar]
- 37.Magliano D.J., Islam R.M., Barr E.L.M., et al. Trends in incidence of total or type 2 diabetes: systematic review. BMJ. 2019;366 doi: 10.1136/bmj.l5003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Wagenknecht L.E., Lawrence J.M., Isom S., et al. Trends in incidence of youth-onset type 1 and type 2 diabetes in the USA, 2002-18: results from the population-based SEARCH for diabetes in youth study. Lancet Diabetes Endocrinol. 2023;11(4):242–250. doi: 10.1016/S2213-8587(23)00025-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Lee K.S., Yoon S.H., Hwang I., et al. Hyperglycemia enhances brain susceptibility to lipopolysaccharide-induced neuroinflammation via astrocyte reprogramming. J Neuroinflammation. 2024;21(1):137. doi: 10.1186/s12974-024-03136-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Tomlinson D.R., Gardiner N.J. Glucose neurotoxicity. Nat Rev Neurosci. 2008;9(1):36–45. doi: 10.1038/nrn2294. [DOI] [PubMed] [Google Scholar]
- 41.Asadollahi E., Trevisiol A., Saab A.S., et al. Oligodendroglial fatty acid metabolism as a central nervous system energy reserve. Nat Neurosci. 2024;27(10):1934–1944. doi: 10.1038/s41593-024-01749-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Heni M. The insulin resistant brain: impact on whole-body metabolism and body fat distribution. Diabetologia. 2024;67(7):1181–1191. doi: 10.1007/s00125-024-06104-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Sanchez-Alegria K., Arias C. Functional consequences of brain exposure to saturated fatty acids: from energy metabolism and insulin resistance to neuronal damage. Endocrinol Diabetes Metab. 2023;6(1) doi: 10.1002/edm2.386. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Assies J., Mocking R.J., Lok A., Ruhe H.G., Pouwer F., Schene A.H. Effects of oxidative stress on fatty acid- and one-carbon-metabolism in psychiatric and cardiovascular disease comorbidity. Acta Psychiatr Scand. 2014;130(3):163–180. doi: 10.1111/acps.12265. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Black C.N., Bot M., Scheffer P.G., Cuijpers P., Penninx B.W. Is depression associated with increased oxidative stress? A systematic review and meta-analysis. Psychoneuroendocrinology. 2015;51:164–175. doi: 10.1016/j.psyneuen.2014.09.025. [DOI] [PubMed] [Google Scholar]
- 46.Palta P., Samuel L.J., Miller E.R., 3rd, Szanton S.L. Depression and oxidative stress: results from a meta-analysis of observational studies. Psychosom Med. 2014;76(1):12–19. doi: 10.1097/PSY.0000000000000009. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Kang E.S., Yun Y.S., Park S.W., et al. Limitation of the validity of the homeostasis model assessment as an index of insulin resistance in Korea. Metabolism. 2005;54(2):206–211. doi: 10.1016/j.metabol.2004.08.014. [DOI] [PubMed] [Google Scholar]
- 48.Singh B., Saxena A. Surrogate markers of insulin resistance: a review. World J Diabetes. 2010;1(2):36–47. doi: 10.4239/wjd.v1.i2.36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.International Diabetes Federation . Diabetes Atlas; 2023. Western Pacific region: prevalence and trends of diabetes and related conditions.https://diabetesatlas.org/data-by-location/region/western-pacific Available from: [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.