Fruit and vegetable consumption and the risk of depression: A meta-analysis

Introduction

- Some lifestyle behaviors have been reported to increase the risk of depression, such as sedentary behavior, long-term stress, and short or long hours of sleep.

- In recent years, the association between diet and depression has also drawn much attention. 

- Among the dietary factors, consumption of fruits and vegetables, which are rich in antioxidants and anti-inflammatory components, was hypothesized to play an important role in depression development

Method

-Search Period: until June 2015

-Inclusion criteria: observational study, Exposure: fruits / vegetables, Outcome: depreesion (determined by self-report based on doctor’s diagnosis, regular use of antidepressant drugs, or depression rating scales)

-Exclusion criteria: postpartum depression

Results

- Number of studies included: 10 (4 cohort, 7 cross-sectional)

1) Fruit intake & risk of depression

- Included studies: 4 cohort & 6 cross-sectional (n=227, 852 participants)

- Study location: Asia (n=3), Europe (n=4), Oceania (n=1), North America (n=1), South America (n=1)

Highest vs lowest category of fruit intake:

- Overall analysis: RR 0.86 (95% CI, 0.81-0.91; p<0.01), I²=48.2%; p=0.043

- Subgroup analysis (cross sectional studies): RR 0.84 (95% CI, 0.75-0.95; p=0.006), I²=55.7%; p=0.046

- Subgroup analysis (cohort studies): RR 0.83 (95% CI, 0.77-0.91; p=0.001), I²=41.2%; p=0.164

- Meta-regression: did not find any a priori determined covariates as the important contributors to the between-study heterogeneity 

- Leave-one-out sensitivity analysis: after excluding one study, heterogeneity of fruits studies decreased to 27.5% and the results still remained significant – therefore, the result is stable.

2) Vegetable intake & risk of depression

- Included studies: 4 cohort & 4 cross-sectional (n=218, 699 participants)

- Study location: Asia (n=3), Europe (n=2), Oceania (n=1), North America (n=1), South America (n=1)

Highest vs lowest category of vegetables intake:

- Overall analysis: RR 0.89 (95% CI, 0.83-0.94; p<0.01) I²=14.1%, p=0.319

- Subgroup analysis (cross sectional studies): RR 0.83 (95% CI, 0.68-1.02; p=0.078) I²=60.4%; p=0.056

- Subgroup analysis (cohort studies): RR 0.88 (95% CI, 0.79-0.96; p=0.007) I²=0%, p=0.923

Publication bias for vegetable intake: Absent

Publication bias for fruit intake: Present (Absent after one study was excluded—result still remained significant)

Limitations

- Diagnostic criteria of depression were inconsistent

- Diet assessment methods varied among the studies and dose-response relationship could not be accessed

- No causal relationship can be established as included studies are observational in design

Conclusion

Fruit and vegetables consumption was inversely associated with the risk of depression

Liu et al. Nutrition 2016; 32: 296-302. http://dx.doi.org/10.1016/j.nut.2015.09.009

Peers' comment

- Some studies are missed and not included in the meta-analysis

- Quality of studies was not accessed

- Some methodological insufficiency

[Saneei et al. Nutrition 2016 (in press) doi:10.1016/j.nut.2016.01.001]

Note:

RR: relative risk, CI: confidence interval
In this meta-analysis, RR<1 indicates lower risk of depression

Information on Statistical analysis

- Pooled measured: inverse-variance weighted mean of the logarithm of RR (95% CI) of depression for the highest versus lowest category of fruit and vegetable intake, respectively.

-Heterogeneity: I² test (If I² ≥ 50%, use random effect model pooling method; If I² < 50%, fixed effect model)

- Publication bias: funnel plot & Egger regression asymmetry test.

- Subgroup analysis: cross-sectional or cohort study

- Meta-regression: to access the potentially important covariates that might exert substantial impacts on between-study heterogeneity.

- Leave-one-out sensitive analysis: to evaluate the key studies that have remarkable impact on the between-study heterogeneity