Skip to main content
. 2018 May 8;21(14):2630–2641. doi: 10.1017/S1368980018001118

Table 1.

Baseline characteristics in 2005 of participants followed up from 2005 to 2010 (the cohort), compared with subjects lost to follow-up and compared with 2010; PURE-NWP-SA study

Baseline data in 2005 of participants followed up in 2010 Baseline data in 2005 of subjects lost to follow-up in 2010 Cohort data of participants in 2010
Variable n Mean sd n Mean sd P value* n Mean sd P value
Age (years) 1154 50·8 10·3 856 48·6 10·4 <0·001 1154
Sex (male; %) 1154 33·6 856 41·8 <0·001 1154
Locality (rural; %) 1154 51·7 856 47·8 0·080 1154
Education attainment (no formal education; %) 1154 36·0 856 32·1 0·073 1154
HIV positive (%) 1154 17·1 856 22·2 0·004 1154 4·9 % new cases
BMI (kg/m2)‡ 1154 25·1 7·0 853 24·1 7·0 <0·001 1154 25·5 7·3 <0·001
Waist circumference (cm)‡ 1147 80·3 12·8 843 79·0 13·3 0·007 1147 81·9 13·0 <0·001
HDL-cholesterol (mmol/l)‡ 1087 1·54 0·61 805 1·48 0·66 0·006 1087 1·42 0·59 <0·001
n Median P25–P75 n Median P25–P75 P value* n Median P25–P75 P value
Energy intake (MJ/d) 1154 7·4 5·6–10·1 796 6·9 4·7–9·8 0·003 1154 10·5 7·5–13·9 <0·0001

PURE-NWP-SA, Prospective Urban and Rural Epidemiology in the North West Province of South Africa; P25, 25th percentile; P75, 75th percentile.

*

P value for significance of differences between subjects followed up in 2010 and those lost to follow-up (categorical variables: χ 2 test; continuous variables: two-sample Wilcoxon test).

P value for significance of differences between the cohort in 2005 and 2010 (categorical variables: χ 2 test; continuous variables: two-sample Wilcoxon test).