Managing cardiovascular risk is a major concern for healthcare systems. Statins are lipid-lowering drugs used to prevent adverse cardiovascular events. Using an instrumental variables approach, this study provides district-level evidence consistent with a causal effect of statin prescribing on hospitalisation rates for cardiovascular disease. By exploiting variation in prescribing patterns and physicians’ preferences to construct an instrument for district-level statin consumption, we move beyond the correlational evidence from existing ecological studies. We focus on the Emilia-Romagna region (Italy) from 2010 to 2019, using district-level longitudinal data. We consider two study populations to measure hospitalisations among individuals aged 50 and over: firstly, the general population living in the region; secondly, those on statin therapy (statin users). The IV estimates show that increasing statin use at the district level is linked to reduced cardiovascular hospitalisation rates, with effects varying by diagnosis and type of admission. Comparing district-level measures of coverage and average treatment intensity, we find more consistent evidence for the coverage-based specification, while intensity estimates based on DDDs per treated patient are less stable and require cautious interpretation. Exploratory analyses by recent hospitalised vascular-event history suggest some differences across treated users, but formal tests provide limited evidence of statistically significant heterogeneity. Overall, the findings support population-level pharmaceutical prevention while cautioning against interpreting aggregate coverage–intensity margins as clean individual-level mechanisms.
This paper tests whether high ambient temperatures during pregnancy – a channel proposed by medical researchers and biologists – help explain the onset of congenital birth disorders (CBDs). Our analysis leverages administrative data on approximately 15 million births across 56,000 localities in Mexico between 2008 and 2021 which we combine with high-resolution temperature data. Leveraging a high-dimensional fixed-effects model at the administrative level-3 (locality), we isolate the causal effect of in-utero temperature exposure on CBD incidence.Overall, we find evidence in support of the view that high ambient temperatures induce CBDs. Our results suggest that an average increase of 1 °C during pregnancy leads to a rise in CBDs by 0.013ppt (1.3 percent). The effects are concentrated in early gestation (first pregnancy trimester), consistent with medical evidence on critical periods of fetal development.
In this paper, we investigate the impact of excess body weight on biological aging and health in old age. Drawing on a large sample of U.S. adults from the Health and Retirement Study, we construct a frailty index, defined as the proportion of age-related health deficits present in a person. Our findings reveal a positive association between BMI (and other indicators of excess weight) and the frailty index. To identify a causal relationship, we use a polygenic risk score (PRS) for BMI as an instrumental variable for excess weight. We further demonstrate that the effect of the PRS on BMI strengthens with later birth years, indicating that, for given age, younger cohorts have experienced longer exposure to obesogenic environments. This trend allows us to construct a gene-environment instrument for excess weight. Our results show a significant, substantial, and nearly log-linear effect of excess weight on health in old age: on average, each additional unit increase in BMI leads to a 4 to 5 percent rise in the frailty index.
