Skip to main content
. 2016 Nov 23;7:1828. doi: 10.3389/fpsyg.2016.01828

Table 7.

Role of differential recall bias in estimating total and direct effects.

% difference in total effectsa % difference in direct effectsb
% (99% CI)c % (99% CI)c
MENTAL HEALTH (SCL-10)
Men (n = 4754) 32.14 (23.38–43.66)* 35.75 (26.55–48.04)*
Women (n = 5571) 26.85 (18.31–37.55)* 28.31 (19.37–40.52)*
GENERAL HEALTH (EQ-5D)
CSES Men (n = 4754) 30.66 (25.95–40.75)* 32.21 (28.12–43.24)*
Women (n = 5571) 44.42 (36.52–52.49)* 47.66 (38.50–57.30)*
WELL-BEING (SWLS)
Men (n = 4754) 27.89 (21.94–41.28)* 29.20 (22.55–42.83)*
Women (n = 5571) 18.99 (11.67–29.53)* 19.48 (11.20–31.02)*
*

p < 0.001.

a

Adjusted for age, parental history of psychological problems, exposure to passive smoke in childhood, and living in Norway at age 1.

b

Adjusted for age, parental history of psychological problems, exposure to passive smoke in childhood, living in Norway at age 1, and psychological abuse and physical abuse in childhood.

c

The percentages show the proportion of difference in total and direct effects estimates using CSES in Tromsø IV and Tromsø VI as exposure.

SCL-10, Hopkins Symptoms Check List-10; scale (0–10), where 0 represents perfect mental health, and 10 represents worst mental health.

EQ-5D, Euroqol-5Dimension generic measure of health-related quality of life; scale (0–10), where 0 represents perfect health, and 10 represents worst health.

SWLS, satisfaction with life scale; scale (0–10), where 0 represents highest well-being, and 10 represents lowest well-being.

Childhood socioeconomic status was measured in Tromsø IV study (1994–1995), and Tromsø VI study (2007–2008), while mental health, general health, and well-being was measured in Tromsø VI study (2007–2008).