Author + information
- Jessica K. Paulus, ScD and
- David M. Kent, MD, MSc∗ ()
- ↵∗Predictive Analytics and Comparative Effectiveness Center, Institute for Clinical Research and Health Policy Studies, Tufts Medical Center, 800 Washington Street, Box 63, Boston, Massachusetts, 02111
We read with interest the report by Pelletier et al. (1). The investigators found that young adult ACS survivors with the strongest feminine gender score had 4.5-fold increased odds of recurrent events regardless of sex. We suggest caution is needed when interpreting these findings for several reasons but particularly given the potential influence of index event bias (IEB).
IEB arises in recurrence risk research when studies select patients based on the occurrence of an index event (2). Such selection induces dependence between risk factors (known and unknown), even when these risk factors are independent in the general, unselected population. This bias is thought to underlie some well-known paradoxes in recurrence risk research (e.g., the “obesity paradox”), whereby some factors (misleadingly) appear to have reversed effects on recurrent events as they do for the primary index event. Although it is not established that feminine characteristics are actually protective of ACS, we speculate that these findings may similarly reflect a spurious “femininity paradox” due to IEB.
Indeed, although the collinearity of sex and gender somewhat complicates the diagnosis, the distribution of risk factors (see Table 1 in the article by Pelletier et al. ) shows a pattern characteristic of IEB; the most feminine (versus masculine) patients have a higher burden of risk factors, with nearly twice the prevalence of diabetes and prior cardiovascular disease (CVD) and a higher likelihood of hypertension and smoking. This pattern (which can extend to unmeasured/unknown variables) might be expected if feminine characteristics were protective for ACS, as only those feminine patients with an extreme risk factor burden would be at risk for the index ACS event. Without accounting for all shared risk factors for the index and recurrent event, residual bias (i.e., positively associating femininity with the recurrent event) will be present. Although regression analyses can be adjusted for CVD risk factors, residual bias remains a concern because of omission of some risk factors (e.g., diabetes), model misspecification, and perhaps most insidiously, unknown risk factors that cause both ACS and its recurrence.
The hypothesis that gender effects might be separable from sex effects is an intriguing and creative one, but perhaps early explorations of this idea should focus on population samples since recurrence risk studies that condition on having a first event are notoriously unreliable with regards to causal inference. For this reason (and many others), we recommend caution before advising post-ACS patients to adopt more masculine attributes.
Please note: The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation