I agree with Dr. Rodrigues. When you're doing population health research, you're getting a random sample of veterans, so you're going to get 13% females. A lot of the indicators of well-being—at Veterans Affairs, 21 indicators are used—are not gendered indicators. Across the seven domains of well-being, all of those should be disaggregated by sex. That's the ideal way to analyze the data. However, as she mentioned, they're small sample sizes, so the confidence intervals are wide. It's hard to tell if there is in fact a difference between women and men.
As Dr. Courchesne mentioned, there is some hope there in terms of getting larger sample sizes of veterans with the 2021 census data. There have been releases of some of that data already, but there haven't been enough releases yet to look specifically at women. Yes, data does need to be disaggregated.