Three thousand, Eight hundred and eighty-two (3,882) children in grades 2-5
, attending 16 rural primary and all-age schools in central Jamaica were we
ighed and their weight-for-age standard deviation scores calculated using t
he World Health Organization/National Center for Health Statistics (WHO/NCH
S) references. Heights were also measured in a random sample of the grade 5
children (n= 793) and height-for-age and body mass index (BMI - kg/m(2)) c
alculated. Sixty-nine per cent of the total sample were of normal weight-fo
r-age, 2% were moderately undernourished (weight-for-age >-3 Z-score, less
than or equal to-2 Z-score), and a further 24% mildly undernourished (weigh
t-for-age >-2 Z-score, less than or equal to-1 Z-score). Few children were
overweight. The frequency distribution of weight-for-age was similar in gir
ls and boys. In the subsample of children in whom heights were measured, 25
.8% were less than or equal to-1 Z-score height-for-age, and of these 4.9%
were <-2 Z-score. Compared with a survey conducted in a similar rural area
in the 1960s, the children's mean weights for age group categories were 1.1
to 3.7 kg heavier. Children who were older than appropriate for their grad
e were more likely to be undernourished (Odds ratio 3.94, 95% CI 3.21, 4.83
), which suggests that undernourished children may be more likely to repeat
a grade or start school later.