Background: Perfluroakly acids (PFAAs) are persistent environmental contaminants linked to certain adverse health outcomes; however, their association with macrovascular diseases has been equivocal. PFAAs are also potent oxygen transporters used in organ preservation and synthetic blood development. Little information is available on the relationship of PFAAs with risk of stroke, for which ischemia in the brain blood vessels is a major cause. Persons with diabetes are at 2-to-9-fold increased risk of stroke, largely due to an increased risk of ischemic stroke. We thus tested the association of four common PFAAs with stroke, and whether this varied by diabetes status.
Methods: Participants were drawn from the C8 Health Project (n=5,270 and 48,832 w/ and w/out diabetes, respectively, age≥20 ys), which collected blood samples, self-reported demographics, medical diagnoses, height, and weight in 2005-2006 as part of a legal settlement following perfluorooctanoate (PFOA) contamination of drinking water in West Virginia and Ohio. Four PFAAs were assessed separately: PFOA, perfluorononaoic acid (PFNA), perfluorohexane sulfonate (PFHxS), and perfluoroctane sulfonate (PFOS). Estimated glomerular filtration rate (eGFR) was assessed via the MDRD formula.
Results: There were 344 (6.5%) stroke cases among those with diabetes and 731 (1.5%) among those without diabetes. Adjusting for age, sex, and eGFR, higher concentrations of each of the PFAAs were associated with a lower likelihood of stroke (OR for PFOA=0.93, 0.88-0.97; OR for PFNA=0.83, 0.74-0.94; OR for PFOS=0.83, 0.77-0.90; OR for PFHxS=0.82, 0.76-0.89). Effect modification by diabetes status was observed for the sulfonate PFAAs (p-interaction <0.001 each). In those with diabetes, higher concentrations of PFOS were associated with a significantly greater reduction in the likelihood of stroke (OR=0.76, 0.69-0.87) than the risk reduction in those without diabetes (OR=0.93, 0.84-1.04). For PFHxS, higher serum concentrations were associated with 31% lower likelihood of stroke in persons with diabetes (OR=0.69, 0.61-0.79) but only a 3% lower likelihood in persons without diabetes (OR=9=0.97, 0.88-1.07). No such effect modification by diabetes status was observed for PFOA or PFNA (p-interaction>0.30 each).
Conclusion: PFAAs are inversely related to stroke risk. This relationship appears to be stronger in persons with diabetes, perhaps due to the salutary effect of the high oxygen carrying capacity of these substances and the longer half- life of PFAA sulfonates in human proteins than the non-sufonate containing PFAAs.