A recent brief investigation of blood pressure-associated cardiovascular risk differences between men and women concluded that “outcomes-based results suggest the possible need for a lower sex-specific definition of optimal [systolic blood pressure] for women.” Before our proverbial blood pressure is elevated by jumping to this conclusion, let’s talk more about the evidence. 

The study looked at data from the Framingham Heart Study, Multi-Ethnic Study of Atherosclerosis, Atherosclerosis Risk in Communities Study, and Coronary Artery Risk Development in Young Adults Study and used sex-pooled propensity analyses to determine risk of certain cardiovascular outcomes over time (hazard ratios) for self-identified women compared to men. The authors found that the risk of cardiovascular morbidity started at a lower systolic blood pressure in women compared to men and went straight to suggesting that blood pressure should be acted upon at a lower threshold for women. 

The association of blood pressure with cardiovascular risk goes way back. Almost every major study, report, or guideline on blood pressure diagnostic thresholds or treatment targets cites some iteration of the same epidemiologic data. Graphical representations of these data depict a clear, climbing line between blood pressure and risk of cardiovascular outcomes, with a very close graphical resemblance to the data published in this recent brief investigation. Looking at those figures, it’s easy to see why so many people have arrived at the what-would-appear-to-be logical conclusion that there is a linear correlation between blood pressure and cardiovascular risk — the lower the blood pressure the better. But the interpretation of those seemingly simple graphs with a more critical lens can take us on a few turns. 

For starters, there is actually more data than is depicted in those oft-cited figures, representing a version of selective data reporting. If you follow that perfectly linear line in towards the y axis (i.e., including lower blood pressure measurements), it turns a corner and becomes a "J" or a "U" shaped curve, indicating that at some threshold, cardiovascular-specific risk increases with lower blood pressures. So does the risk of death.

Never mind actual trial-based data indicating that overtreating blood pressure can lead to electrolyte abnormalities, falls, kidney injury, among other adverse events. Oh, and also death. 

Most importantly, let’s not forget that this brief investigation is based on observational data. You know, the kind that is lower down on the evidence pyramid because it is less internally valid and potentially further from the truth than data collected prospectively through well-planned and well-executed randomized trials and then analyzed with a systematic review and meta-analysis. Lack of randomization leaves room for the very real possibility that there are known or unknown confounders that could play more of a causal role in cardiovascular risk than blood pressure itself. Perhaps people with lower blood pressure have something else in common that makes them also less likely to have a heart attack and die. That possibility may be at least as likely as the idea that high blood pressure is the direct link to cardiovascular morbidity and mortality. We cannot make causal assumptions from observational data.

So, jumping straight from observational data to the assertion that perhaps guidelines need to be changed to reflect these outcomes skips what is perhaps the most important step: the randomized controlled trial. These data about sex differences in blood pressure-associated cardiovascular risk do make a convincing argument — that a trial to assess any causal link is warranted.