Abstract
The few prospective studies examining the relation between proton pump inhibitor (PPI) use and risk of vertebral fracture (VF) suggest a higher risk, but the magnitude of the association has been inconsistent. Moreover, no prospective studies have examined the association between substantially longer duration of PPI use and VF risk. Our objective was to determine the association between PPI use, H(2)RA use, and incident clinical VF in women. We conducted a prospective study in 55,545 women participating in the Nurses' Health Study. PPI and H(2)RA use was assessed by questionnaire every 4 years. Self-reports of VF were confirmed by medical record. Our analysis included 547 incident VF cases (2002-2014). The multivariate adjusted relative risk (MVRR) of VF for women taking PPIs was 1.29 (95% CI 1.04-1.59) compared with non-users. Longer duration of PPI use was associated with higher VF risk (MVRR 1.16 [0.90-1.49] for < 4 years; 1.27 [0.93-1.73] for 4-7.9 years; 1.64 [1.02-2.64] for ≥ 8 years; p(trend) = 0.01). The MVRR of VF for women taking H(2)RAs was 1.22 (0.90-1.67) compared with non-users. Longer duration of H(2)RA use was not associated with VF risk (MVRR 1.16 [0.88-1.53] for < 4 years; 0.98 [0.60-1.59] for ≥ 4 years; p(trend) = 0.72). PPI use is independently associated with a modestly higher risk of VF and the risk increases with longer duration of use. There was no statistically significant association between H(2)RA use and VF risk. Our findings add to the growing evidence suggesting caution with PPI use, particularly with longer duration of use.