Table 3 Longitudinal association of subjective mental health and its epigenetic signature with depressive symptoms (ESTHER and KORA study).

From: Subjective mental health, incidence of depressive symptoms in later life, and the role of epigenetics: results from two longitudinal cohort studies

 

Logistic regressions: odds ratios (OR) with 95% confidence interval (CI) for incidence of depressive symptoms*

Model 1a OR (CI)

Model 2b OR (CI)

ESTHER cohort (n = 1492)

MCS-12 (per 10 units)

0.46 (0.39–0.53)

0.46 (0.40–0.54)

Cg27115863 (per SD)

0.72 (0.61–0.86)

0.76 (0.63–0.91)

Cg27115863 (quartiles)

Q1 vs. Q4

2.20 (1.37–3.54)

1.95 (1.20–3.19)

Q2 vs. Q4

1.30 (0.80–2.12)

1.29 (0.78–2.14)

Q3 vs. Q4

1.53 (0.96–2.44)

1.48 (0.91–2.39)

KORA cohort (n = 890)

Cg27115863 (per SD)

0.84 (0.69–1.02)

0.84 (0.69–1.03)

Cg27115863 (quartiles)

Q1 vs. Q4

1.66 (0.95–2.89)

1.62 (0.92–2.87)

Q2 vs. Q4

1.58 (0.95–2.64)

1.53 (0.91–2.57)

Q3 vs. Q4

1.30 (0.80–2.14)

1.28 (0.78–2.12)

Meta-analysis of ESTHER and KORA

Cg27115863 (per SD)

0.77 (0.66, 0.89)

0.80 (0.70, 0.91)

Cg27115863 (quartiles)

Q1 vs. Q4

1.96 (1.36, 2.81)

1.80 (1.25, 2.61)

Q2 vs. Q4

1.43 (1.00, 2.03)

1.40 (0.98, 2.01)

Q3 vs. Q4

1.42 (1.01, 1.98)

1.38 (0.97, 1.95)

  1. MCS-12, Mental Component Summary of the Short-Form 12 Health Survey; SD standard deviation.
  2. *Depressive symptoms were measured with the Geriatric Depression Scale (GDS) in the ESTHER and with the Depression and Exhaustion Scale (DEEX) in the Kora cohort. A GDS score > 5 and a DEEX score ≥9 in men and ≥11 in women were taken as cut-off for depressive symptoms.
  3. aModel 1 adjusted for age, sex, educational level. Model 1 including cg27115863 also adjusted for batch effects and leukocyte distribution in both cohorts and for the DEEX as measured at baseline in the KORA cohort.
  4. bModel 2 additionally adjusted in both cohorts for smoking, physical activity, body mass index, and lifetime history of cardiovascular disease (KORA: baseline prevalence).