Table 2.
Study characteristics: observational studies
| Reference | Country | Population, sample size | Mean age (y) | Female sex (%) | Exposure | Exposure assessment method | Outcome | Outcome assessment method | Adjustments | Model |
|---|---|---|---|---|---|---|---|---|---|---|
| Cross-sectional studies | ||||||||||
| Amadieu et al (2021)22 | Belgium | Alcohol use disorder individuals (n = 48) | 47 | 37.5 | Total dietary fiber | Three 24-h recalls | Depression and anxiety |
|
Age, sex, educational level, energy intake, BMI, tobacco, and alcohol consumption | Linear regression |
| Chrzastek et al (2020)43 | Poland | Older adults (n = 813) | 75 | 73 | Total dietary fiber | 24-h recall | Symptoms of depression | 15-item GDS (GDS score > 5)a | Age, education years, waist circumference, Cognitive function, and chronic obstructive pulmonary disease | Logistic regression (analyses for women and men performed separately) |
| Eissenstat et al (2020)23 | United States | Healthy adults (n = 4747) | ≥18 | 53 | Total dietary fiber | Two 24-h recalls | Symptoms of depression | PHQ-9 | Sex, age, education, income, birthplace, and race | Linear regression |
| Fang et al (2013)26 | United States | Healthy adults (n = 225) | 27 | 100 | Total dietary fiber | Three 24-h recalls | Symptoms of depression | 20-item CESD | Prior condition, age, race, education, marital status, hormonal contraceptive use, income, and history of heart disease | Linear regression |
| Gopinath et al (2016)35 | Australia | Older adults (n = 1504) | 73 | 62 | Total dietary fiber | 145-item, self-administered FFQ | Symptoms of depression | 10-scale CESD or use of antidepressants (score ≥ 10)a | Age, sex, energy, cognitive impairment, walking disability, receiving pension, and antidepressant use, previous history of stroke and arthritis | Logistic regression |
| Kim et al (2020)36 | Korea | Healthy adults (n = 546) | 42 | 65 | Total dietary fiber | FFQ | Clinical depression | Diagnosis of depression by a physician | Age, sex, economic status, education, smoking status, alcohol consumption, physical activity, subjective health status, BMI, and total energy intake | Logistic regression |
| Kim et al (2021)45 | Korea | Healthy adults/premenopausal women (n= 5807) | 47 | 100 | Total dietary fiber | 24-h recall | Depression | PHQ-9 (score ≥ 10)a | Age, BMI, education level, marital status, household income, smoking status, alcohol use, adequate physical activity, and chronic disease status | Logistic regression |
| Li et al (2020)33 | United States | Healthy adults/premenopausal women (n = 3054) | 46 | 100 | Total dietary fiber | FFQ | Symptoms of depression | 20-item CESD | Age, race/ethnicity, total family income, education, sport, BMI, dietary total caloric intake, use of antidepressant, SHBG, and FSH | Linear regression (analyses for premenopausal and early perimenopausal women performed separately) |
| Miki et al (2016)37 | Japan | Healthy adults (n = 1977) | 42 | 11 | Total dietary fiber | Brief self-administered diet history questionnaire | Symptoms of depression | Japanese version of CESD (score ≥16)a | Age, sex, site, marital status, job grade, and other factorsb | Logistic regression |
| Purnomo et al (2021)56 | Australia | HIV individuals (depressed vs nondepressed) (n = 58) | 43 | 3.4 | Total dietary fiber | FFQ | Symptoms of depression | 10-item CESD (score ≥ 10)a | – | Mann-Whitney U test |
| Rintamaki et al (2014)44 | Finland | Healthy vs individuals with depressive and anxiety disorder (n = 5351) | 48 | 59 | Total dietary fiber | 128- item FFQ | Depression and anxiety | A structured, computer-aided mental health interview | Age, education, social support smoking, and energy intake | General linear model (analyses for men and women performed separately) |
| Saghafian et al (2021)38 | Iran | Healthy adults (n = 3363) | 36 | 58 | Total dietary fiber | 106-item FFQ | Symptoms of depression and anxiety | Iranian version of HADS (score > 21)a | Age, sex, energy intake, and other factorsc | Logistic regression |
| Woo et al (2006)39 | Hong Kong | Older adults (n = 3395) | 72 | 44 | Total dietary fiber | 7-day FFQ | Depression | Face-to-face interviews, using a validated Chinese version of GDS (score ≥ 8)a | CSID score, age, sex, education level, socioeconomic status, and number of medical diseases | Logistic regression |
| Xia et al (2021)40 | China | General adults (n = 24 306) | 41 | 46 | Total dietary fiber | 100-food item modified FFQ | Symptoms of depression | Chinese version of the Zung SDS (score >45) | Age, BMI, type 2 diabetes, hypertension, hyperlipidemia, physical activity, and other factorsd | Logistic regression (analyses for men and women performed separately) |
| Xu et al (2018)25 | United States | Civilian noninstitutionalized US population (16 807) | 50 | 57 | Total dietary fiber | 24-h recall | Symptoms of depression | PHQ-9 (score ≥ 10)a | Age, sex, race, marital status and other factorse | Logistic regression |
| Yun et al (2021)57 | Korea | General adults (n = 10 106) | 65 | 58 | Total dietary fiber | Diet questionnaire (un-specified) | Symptoms of depression | PHQ-9 | Total food intake, sex, income, education, and marital status | Linear regression |
| Longitudinal studies | ||||||||||
| Gangwisch et al (2015)41 | Columbia | Postmenopausal women (n= 69 954) | 64 | 100 | Total dietary fiber | 145-item FFQ | Depressive symptoms | Burnam 8-item scale (standard threshold of 0.06)a | Nutrient density, race/ethnicity, education, income, BMI, and other factorsf | Logistic regression (3-y follow-up) |
| Perez-Cornago et al (2016)24 | Spain | Healthy adults (n = 14 539) | 38 | 59 | Total prebiotic | 136-item semi-quantitative FFQ | Depression | Self-reported, physician-diagnosed | Age, sex, BMI, smoking, marital status, personality traits, unemployment, and living alone | Multivariable Cox proportional hazard (9.3-y follow-up) |
| Ramin et al (2020)42 | United States | Postmenopausal women (n= 14 129) | 60 | 100 | Total dietary fiber | 127-food-item Harvard FFQ | Depressive symptoms | Mental Health Score | Total calorie, age, education, alcohol, physical activity, antidepressant use, and WHR | Linear regression (17-y follow-up) |
| Case-control studies | ||||||||||
| Gougeon et al (2017)58 | Canada | Older adults (n = 316; depressed vs nondepressed) | 75 | 61 | Total dietary fiber | 24-h recall | Depression | GDS or use of anti-depressants (score ≥ 11)a | Physical activity, functional autonomy, and stressful life events | General linear model |
| Guligowska et al (2016)59 | Poland | Older adults (n = 130; depressed vs nondepressed) | 71 | 76 | Total dietary fiber | 24-h recall | Depression | 15-item GDS (cutoff for depression not reported)a | NA | Mann-Whitney U test |
| Othman et al (2018)55 | Tunisia | Adults (n = 100; depressed vs nondepressed) | 44.5 | 67 | Total dietary fiber | Feeding (diet) history | Depression | HADS and PHQ-9 (cutoff for depression not reported)a | NA | Student’s t test |
| Park et al (2010)60 | South Korea | College students (n = 130; depressed vs nondepressed) | 20 | 100 | Total dietary fiber | 3-d recall | Depression | Korean version of CESD (score >16)a | NA | Student’s t test |
a
The cutoff scores for establishing either depressive symptoms or depression.
b,c,d,e,fdenotes the covariates used in statistical model adjustments in addition the ones listed in the table (further elaborated in Appendix S1).
Abbreviations: BMI, body mass index; CESD, Centre for Epidemiologic Studies Depression; CSID, Community Screening Instrument for Dementia; FFQ, food frequency questionnaire; FSH, follicle-stimulating hormone; GDS, Geriatric Depression Scale; HADS, Hospital Anxiety and Depression Scale; NA, not applicable; PHQ, Patient Health Questionnaire; SDS, Self-Rating Depression Scale; SHBG, sex hormone binding globulin; WHR, waist to hip ratio.