Skip to main content
. Author manuscript; available in PMC: 2014 Sep 1.
Published in final edited form as: Arterioscler Thromb Vasc Biol. 2014 Jan 23;34(3):654–660. doi: 10.1161/ATVBAHA.113.302572

Table 3. Odds ratios of BMI or height change for cIMTa at 60-64 years in the upper quartile versus lower three quartiles: estimates from logistic regression models stratified by anthropometric exposure and sex.

Male Female Male and female
Nb ~1 SD OR (95% CI) P-value Nb ~1 SD OR (95% CI) P-value P-value for Z-score-by-sexc
Weight Z-score changed
 ZCHANGEwt0-2yr 129/377 1.3 kg/m2 1.035 (0.825, 1.300) 0.8 134/426 1.4 kg/m2 1.083 (0.878, 1.336) 0.5 0.8
BMI Z-score changed
 ZCHANGEbmi2-4yr 111/339 1.6 kg/m2 1.173 (0.933, 1.476) 0.2 125/358 1.5 kg/m2 1.017 (0.810, 1.275) 0.9 0.4
 ZCHANGEbmi4-7yr 105/347 1.1 kg/m2 0.856 (0.660, 1.109) 0.2 121/384 1.3 kg/m2 0.842 (0.661, 1.074) 0.2 0.9
 ZCHANGEbmi7-15yr 94/323 1.9 kg/m2 1.267 (0.950, 1.689) 0.1 104/351 2.1 kg/m2 0.972 (0.732, 1.290) 0.8 0.2
 ZCHANGEbmi15-20yr 92/305 1.7 kg/m2 1.283 (0.910, 1.808) 0.2 113/348 1.9 kg/m2 1.174 (0.854, 1.615) 0.3 0.7
Height Z-score changed
 ZCHANGEht2-4yr 114/353 4.1 cm 0.698 (0.534, 0.913) 0.009 136/374 4.1 cm 1.036 (0.813, 1.322) 0.8 0.03
 ZCHANGEht4-7yr 108/363 3.8 cm 1.280 (0.934, 1.752) 0.1 129/404 3.8 cm 1.099 (0.810, 1.490) 0.5 0.5
 ZCHANGEht7-15yr 97/339 5.7 cm 1.192 (0.842, 1.690) 0.3 111/370 3.8 cm 0.851 (0.603, 1.201) 0.4 0.2
 ZCHANGEht15-36yr 104/333 4.8 cm 0.798 (0.588, 1.082) 0.1 117/358 2.8 cm 1.341 (0.854, 2.107) 0.2 0.06

cIMT carotid intima-media thickness; BMI body mass index; OR odds ratio; CI confidence interval; SD standard deviation

a

cIMT was taken as the average of six repeat measurements (three on the left hand side of the body and three on the right hand side of the body) at the lateral view of the common carotid artery. Sex-stratified centiles were used to create the two groups (i.e., upper quartile and lower three quartiles).

b

N in upper quartile / N in lower three quartiles (of carotid intima-media thickness).

c

From logistic regression models including Z-score, sex (referent: male), and Z-score-by-sex (and baseline Z-score and baseline Z-score-by-sex).

d

Z-scores were created internally using the LMS method, stratified by sex and age (0-5, 6-15, 20-64 years). Z-score change variables were created as the residuals from sex-stratified general linear regressions of Z-score at time T+1 on Z-score at time T (and a Z-score at time T squared term); when incorporated into a model (with the variable they are conditional on) they can be interpreted as change over the age period, accounting for regression to the mean.