Interactions between the apolipoprotein E ɛ4 allele status and adverse childhood experiences on depressive symptoms in older adults

Figures & data

Table 1 Characteristics of the study participants according to their apolipoprotein ɛ-4 (APOE-ɛ4) carrier status

	Non-carriers (N=116)	Carriers (N=21)	Total	Test statistics^a	Effect size^a	p ^a
Gender, female, n (%)	69 (59.5)	13 (61.9)	82 (59.9)	0.04^b	0.02^c	0.835
Age (years), mean (SD)	59.01 (6.00)	60.38 (6.73)	59.22 (6.11)	−0.95^d	−0.21^e	0.346
Educational years, mean (SD)	12.30 (4.10)	11.33 (4.04)	12.15 (4.10)	0.99^c	0.24^e	0.323
Marital status				0.46^b	0.06^d	0.550
Married	92 (79.3)	18 (85.7)	110 (80.3)
Widowed/divorced/separated	24 (20.7)	3 (14.3)	27 (19.7)
Employment				0.36^b	0.05^d	0.547
Employed	58 (50.0)	9 (42.9)	67 (48.9)
Unemployed	58 (50.0)	12 (57.1)	70 (51.1)
Apolipoprotein genotype, n				NA	NA	NA
ɛ2/ɛ3	16	0	16
ɛ3/ɛ3	100	0	100
ɛ3/ɛ4	0	21	21
ACE score (0–10)	1.16 (2.23)	1.86 (2.99)	1.26 (2.36)	0.99^c	−0.04^e	0.323
QIDS-SR score (0–27)	7.17 (5.07)	8.71 (5.43)	7.41 (5.15)	−1.27^c	−0.29^e	0.207

NA=not applicable; ACE=adverse childhood experiences; QIDS-SR=Quick Inventory of Depressive Symptomatology-Self Rating.

^aComparison between non-carriers and carriers.

^bChi-square test analysis.

^cΦ analysis.

^d t-statistics analysis.

^eCohen's d analysis.

Table 2 Results of multiple linear regression models exploring the effect of childhood adversity and an apolipoprotein-ɛ4 (APOE-ɛ4) carrier status on depressive symptoms

	B (95% CI)	SE	t	p
Model 1: Overall
Intercept	−0.04 (−11.01, 10.92)	5.54	−0.01	0.994
ACE scores	0.54 (0.20, 0.88)	0.17	3.14	0.002
APOE-ɛ4 carrier status (for ACE=0)	1.55 (−0.54, 3.64)	1.06	1.46	0.145
Model 2
Non-carrier
Intercept	1.37 (−9.50, 12.23)	5.49	0.25	0.804
ACE scores	0.32 (−0.07, 0.71)	0.20	1.63	0.107
Carrier
Intercept	1.56 (−9.42, 12.54)	5.55	0.28	0.779
ACE scores	1.17 (0.52, 1.82)	0.33	3.56	0.001
APOE-ɛ4 carrier status (for ACE=0)	0.19 (−2.19, 2.58)	1.21	0.16	0.874
ACE×APOE-ɛ4	0.85 (0.10, 1.61)	0.38	2.24	0.027

B=unstandardized regression coefficient; CI=confidence interval; SE=standard error; ACE=adverse childhood experiences.

Independent variables are displayed in the first column and the dependent variables are the Quick Inventory of Depressive Symptomatology-Self Rating (QIDS-SR) scores. Model 1 includes sex, age, educational years, marital and occupational status, ACE scores, and APOE-ɛ4 carrier status as independent variables; Model 2 includes an ACE×APOE-ɛ4 interaction term in addition to the variables in model 1. The intercept is the value when all the adjustment variables have a value of 0 (i.e., sex=male, age=0 year, educational years=0, marital status=married, occupational status=employed).

F=7.50, p<0.001, and R ²=0.28 in Model 1, and F=7.43, p<0.001, and R ²=0.30 in Model 2.

The proportions of floor and ceiling values were 58.4 and 0% for ACE and 3.6 and 0% for QIDS-SR.