A handful of general population epidemiological surveys retrospectively assessed lifetime exposure to disasters and prevalence of post-disaster PTSD. The first such study, the National Comorbidity Survey (NCS; Kessler et al. 1995), found that much lower proportions of disaster survivors developed post-disaster PTSD (3.7% of men, 5.4% of women) than in disaster-focused studies. More recent community epidemiological surveys in Europe (Ferry et al. 2014; Olaya et al. 2015) and the United States (Breslau et al. 1998, 2013) found similar results. Importantly, PTSD prevalence estimates in these surveys were considerably higher for some other lifetime traumatic experiences (Molnar et al. 2001; Darves-Bornoz et al. 2008; Olaya et al. 2015), suggesting that the low post-disaster PTSD prevalence estimates were not due to recall bias. The discrepancy between these low prevalence estimates in representative community samples and much higher estimates in post-disaster surveys raises the question whether demand characteristics and unrepresentative samples led to upwardly biased estimates in post-disaster surveys (Bonanno et al. 2010).
