A test of recently proposed BMI standards with respect to old age

The purpose of this study teas to investigate the effects of age on the rel ationship between BMI and multiple coronary risk factors, and to determine whether the BMI classification by NHLBI and WHO is applicable as a predicto r of coronary risk factors in older (265 years) as well as in younger (< 65 years) men and women. Effects of age on ten coronary risk factors were exa mined. Sex differences in the slopes of BMI on risk factors were compared b etween younger and older subjects in order to examine the effects of age on these relationships. The frequency of risk factor abnormality in individua l BMI groups (18.5-24.9 25.029. 9, 30.0+) was examined for four age-sex gro ups. The significance of an age group-BMI interaction term was tested by th e logistic regression model to see whether there is a significant differenc e in the relationship between BMI and the individual risk factor abnormalit ies between younger and older subjects. Older subjects had significantly hi gher values for most risk factors than younger subjects. The slopes of BMI on risk factors were different between younger and older subjects for fasti ng glucose, total, HDL- and LDL-cholesterol in men, and for diastolic blood pressure, total and LDL-cholesterol in women. The proportion of subjects w ith abnormal risk factor levels in each of the three BMI groups was higher in older than in younger subjects for most risk factors. There was generall y a progressive worsening of the risk factor levels with increasing BMI in both age groups. There was no consistent age difference in the relationship between BMI groups and the frequency of risk factor abnormality. We conclu de that, although age increases the frequency of most cardiovascular risk f actor abnormalities, in general, itdoes not affect the trend of the relatio nship between the risk factors and the normal, overweight and obese BMI gro ups defined by NHLBI and WHO. Therefore, these BMI categories are applicabl e as predictors of risk factor levels in older as well as in younger men an d women.