Skip to main content
. 2008 Nov 20;169(3):347–354. doi: 10.1093/aje/kwn333

Table 3.

Associations (Mixed-Models Analyses) Between Hostility Levels and Body Mass Index Trajectories From Phase 1 (1985–1988) to Phase 7 (2002–2004) of the Whitehall II Study, United Kingdom

Men
Women
β SE P Value β SE P Value
Model 1a (n = 4,494) (n = 1,990)
    Time, per year 0.113 0.006 <0.001 0.151 0.010 <0.001
    Quartile of hostility level
        Highest 0.731 0.146 <0.001 1.326 0.306 <0.001
        Middle highest 0.220 0.139 0.115 0.552 0.290 0.057
        Middle lowest 0.132 0.141 0.352 −0.100 0.293 0.732
        Lowestb 0 0
            P for linear contrast <0.001 <0.001
    Time × quartile of hostility level
        Highest 0.016 0.007 0.023 0.006 0.014 0.709
        Middle highest 0.012 0.006 0.052 0.016 0.014 0.238
        Middle lowest 0.005 0.007 0.424 0.008 0.014 0.547
        Lowestb 0 0
            P for linear contrast 0.011 0.562
Model 2c (n = 4,374)d (n = 1,924)d
    Time, per year 0.174 0.391 <0.001 0.133 0.028 <0.001
    Quartile of hostility level
        Highest 0.555 0.124 <0.001 0.333 0.255 0.191
        Middle highest 0.212 0.116 0.068 0.240 0.232 0.302
        Middle lowest 0.139 0.117 0.233 −0.137 0.235 0.559
        Lowestb 0 0
            P for linear contrast <0.001 0.055
    Time × quartile of hostility level
        Highest 0.014 0.007 0.043 0.010 0.016 0.525
        Middle highest 0.010 0.006 0.125 0.011 0.013 0.478
        Middle lowest 0.006 0.006 0.309 0.009 0.013 0.491
        Lowestb 0 0
            P for linear contrast 0.041 0.696

Abbreviation: SE, standard error.

a

Results were adjusted for time, quartile of hostility level, and time × quartile of hostility level.

b

Reference category.

c

In addition to the factors listed above for model 1, results were adjusted for age, time × age, socioeconomic status, time × socioeconomic status (not in men), ethnicity, time × ethnicity, body mass index (weight (kg)/height (m)2) at age 25 years, and time × body mass index at age 25 years.

d

Numbers should be similar to those of Table 2. They are slightly higher because marital status was not included as a covariate (P > 0.05 in Table 2).