Skip to main content
. Author manuscript; available in PMC: 2010 Sep 1.
Published in final edited form as: JAMA. 2010 Mar 24;303(12):1159–1166. doi: 10.1001/jama.2010.297

Table 5.

The role of health behaviors in explaining the association between socioeconomic position and non-cancer/non-CVD mortality. The Whitehall II Study. N=9585, Deaths=150.

Baseline assessment of health behaviors Longitudinal assessment of health behaviors (Phases 1, 3, 5 and 7) Difference between longitudinal & baseline assessments

HR 95% CI % attenuation 95% CI HR 95% CI % attenuation 95%CI Δ% attenuation# 95%CI
Model 1* 1.67 (1.01, 2.75) 1.67 (1.01, 2.75)
Model 1 + Smoking 1.32 (0.80, 2.17) 46% (18%, 277%) 1.38 (0.84, 2.28) 37% (13%, 265%) 9% (−5%, 57%)
Model 1 + Alcohol 1.69 (1.01, 2.80) 2% (−47%, 42%) 1.58 (0.95, 2.64) 10% (−19%, 126%) 12% (−11%, 127%)
Model 1 + Diet 1.50 (0.91, 2.49) 20% (2%, 192%) 1.47 (0.89, 2.44) 24% (2%, 198%) 4% (−42%, 52%)
Model 1 + Physical activity 1.61 (0.97, 2.68) 7% (−19%, 80%) 1.36 (0.83, 2.23) 40% (13%, 255%) 33% (3%, 221%)
Fully adjusted Model 1.22 (0.72, 2.06) 61% (16%, 425%) 1.03 (0.61, 1.74) 94% (35%, 595%) 33% (−16%, 262%)
*

HR for lowest versus highest occupational position adjusted for sex and year of birth

HR for lowest versus highest occupational position adjusted for sex, year of birth and all health behaviors

Percent attenuation= 100 × (β Model 1+health behavior(s) − βModel1)/(β Model1)

#

Difference between the model with repeated assessment compared with the baseline assessment of health behaviors