Table 2. Multivariable-adjusted odds ratios (95% CI) of depressive symptoms by tertile or quartile of serum folate, 25-hydroxyvitamin D, and dietary pattern.
Quartiles of serum folate concentrations | P for trenda | ||||
Q1 (low) | Q2 | Q3 | Q4 (high) | ||
Serum folate (cross-sectional association)22 | |||||
Menb | |||||
CES-D score ≥16 | 1.00 (reference) | 0.53 (0.27–1.03) | 0.33 (0.16–0.68) | 0.51 (0.25–1.03) | 0.03 |
CES-D score ≥19 | 1.00 (reference) | 0.61 (0.31–1.23) | 0.37 (0.17–0.82) | 0.35 (0.16–0.75) | 0.003 |
Womenb | |||||
CES-D score ≥16 | 1.00 (reference) | 0.93 (0.40–2.18) | 0.92 (0.41–2.08) | 0.91 (0.38–2.16) | 0.83 |
CES-D score ≥19 | 1.00 (reference) | 0.67 (0.26–1.71) | 1.03 (0.44–2.42) | 0.80 (0.31–2.03) | 0.89 |
Tertiles of serum folate concentrations | P for trenda | ||||
T1 (low) | T2 | T3 (high) | |||
Serum folate (prospective association)19,c | |||||
CES-D score ≥16 | 1.00 (reference) | 0.66 (0.29–1.52) | 0.40 (0.16–0.99) | 0.047 | |
CES-D score ≥19 | 1.00 (reference) | 0.26 (0.08–0.84) | 0.27 (0.08–0.86) | 0.02 | |
Quartiles of serum 25-hydroxyvitamin D concentrations | P for trenda | ||||
Q1 (low) | Q2 | Q3 | Q4 (high) | ||
Serum 25-hydroxyvitamin D21,d,e | |||||
Workplace A (survey in July) | 1.00 (reference) | 0.60 (0.20–1.76) | 0.75 (0.27–2.08) | 0.70 (0.24–2.05) | 0.62 |
Workplace B (survey in November) | 1.00 (reference) | 0.84 (0.45–1.58) | 0.83 (0.44–1.58) | 0.59 (0.30–1.15) | 0.14 |
Tertiles of dietary pattern score | P for trenda | ||||
T1 (low) | T2 | T3 (high) | |||
Dietary patterns20,e,f | |||||
Healthy Japanese pattern | 1.00 (reference) | 0.99 (0.62–1.59) | 0.44 (0.25–0.78) | 0.006 | |
Animal food pattern | 1.00 (reference) | 1.47 (0.93–2.32) | 0.97 (0.61–1.55) | 0.91 | |
Westernized breakfast pattern | 1.00 (reference) | 1.02 (0.64–1.64) | 1.27 (0.77–2.10) | 0.34 |
Abbreviations: CES-D, Center for Epidemiologic Studies Depression; Q, quartile; T, tertile.
aBased on multiple logistic regression analysis with assignment of ordinal numbers to categories of each food or nutrient.
bAdjusted for age, workplace, job position, marital status, job physical activity, non-job physical activity, smoking status, and alcohol consumption.
cAssociation of serum folate at baseline (in 2006) with prevalence of depressive symptoms at follow-up survey (in 2009); adjusted for all variables in footnote b, plus sex and CES-D scale score at baseline.
dAdjusted for all variables in footnote b except workplace, but including sex, BMI, and dietary folate intake.
eDepressive symptoms was defined as a CES-D scale score ≥16.
fHealthy Japanese pattern: high intakes of vegetables, fruit, soy products, mushrooms, and green tea. Animal food pattern: high intakes of fish/shellfish, meat, processed meat, mayonnaise, and egg. Westernized breakfast pattern: high intakes of bread, confectioneries, milk/yogurt, mayonnaise, and egg and low intakes of rice, alcohol, and fish. Adjusted for all variables in footnote b, plus sex, BMI, history of hypertension, history of diabetes mellitus, and total dietary intake.