StATS: Adjusting a variable for age and sex (October 26, 2006).

Someone asked me how to adjust bone mineral density (BMD) for age and sex. I presume that BMD changes as children grow (or as adults age) and that BMD is different for men and women. If you did not adjust for age and sex, then are statistical comparison that you make between a treatment group and control group could be biased by a differential in the sex ratio and/or average age between the two groups.

The best thing is to see if there are any published norms for bone mineral density. I know that there are norms for other measures that are age and sex adjusted. Typically these norms will produce z-scores and/or percentiles. You would then do the analysis on the z-score or percentile rather than the raw score.

If norms are not available, then you do an internal adjustment. If you have an obvious control group and an obvious treatment group then compare the age and gender distribution of the two groups. If the distributions are identical then there is no need to do any adjustment, except, perhaps, to add a bit of precision to your model or to produce a result that is more easily interpretable. If there is a small or moderate discrepancy, then you can use age and sex as covariates in the model that compares the control group to the treatment group. If there is a large discrepancy--large enough so there is little or no overlap in the demographics of the two groups, then no adjustment is possible.

There are other approaches--matching, stratification, and propensity scores, that also work well if you need to make an internal adjustment. Just keep in mind that the internal adjustment is never as good as using a published norm.

This page was written by Steve Simon while working at Children's Mercy Hospital. Although I do not hold the copyright for this material, I am reproducing it here as a service, as it is no longer available on the Children's Mercy Hospital website. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Covariate adjustment.