A substantial fall in high-density lipoprotein cholesterol (HDL-C) during puberty in boys, but not girls, has been reported in Western populations. The fall in boys is believed to be due to hormonal changes – androgens have been shown to be associated with lower HDL-C, whereas oestrogens are associated with higher HDL-C. The fall in HDL-C during puberty was not observed, however, in a study of Moslem boys in Israel, nor in a group of Japanese boys. A diet high in phyto-oestrogens may account for the lack of a fall in HDL-C in these populations.

To examine the effect of dietary supplementation with phyto-oestrogens on the HDL-C concentration of adolescent boys from a Western population. We hypothesised that dietary supplementation of 50 mg of the isoflavones daidzein and genistein would produce a 12% higher HDL-C concentration than in controls at the end of a 6-week intervention period.

A randomised controlled trial.

Hellyer College in Burnie (Tasmania, Australia).

Adolescent boys (aged 16–18 years) were recruited through a letter sent to parents. A total of 132 eligible participants enrolled and five subjects withdrew from the trial.

No significant increase in HDL-C was observed in the treatment group (–0.02 mmol l–1, standard error (SE)=0.03, P = 0.53) or the placebo group (0.05 mmol l–1, SE = 0.03, P = 0.11).

Factors other than isolated dietary isoflavones may be responsible for the lack of fall in HDL-C during puberty in Japanese and Moslem boys.

To investigate the prevalence of anaemia and subclinical vitamin A deficiency among adolescent schoolboys in Dhaka City, Bangladesh, and to identify factors related to anaemia and vitamin A status.

A cross-sectional study.

Government high schools in Dhaka City, Bangladesh.

A total of 381 boys, aged 11–16 years, from 10 schools in Dhaka City participated in the study. Socio-economic, anthropometric and dietary data were collected. Haemoglobin and serum retinol (vitamin A) concentrations were determined.

Seven per cent of the boys were anaemic and 22% had serum vitamin A levels below the adequate level of 1.05 μmol l−1, with only 1.5% having subclinical vitamin A deficiency (<0.70 μmol l−1). Food frequency data revealed poor dietary habits. Multiple regression analysis showed that age, body mass index (BMI), parents' occupation, serum vitamin A level and frequency of intakes of meat and fruit were significantly independently related to haemoglobin level. The overall F-ratio (13.1) was highly significant (P< 0.000) and the adjusted R2 was 0.192. For serum vitamin A, BMI, father's education, per capita expenditure on food, haemoglobin concentration and frequency of intake of vitamin A-rich fruit were found to be significantly independently related. The overall F-ratio (14.5) was highly significant (P< 0.000) and the adjusted R2 was 0.186.

The data show that adolescent schoolboys in Dhaka City have anaemia and inadequate vitamin A status, although the extent of the problems is lower than in other population groups in the country. Sociodemographic and dietary factors appear to have important relationships with anaemia and vitamin A status of these boys.