Skip to main content
. Author manuscript; available in PMC: 2011 Jul 1.
Published in final edited form as: Gastroenterology. 2010 Mar 27;139(1):93–101. doi: 10.1053/j.gastro.2010.03.055

Table 5.

≥2 years of proton pump inhibitor use and the risk of hip fracture, stratified by presence or absence of other specific risk factors for hip fracture.

Risk Factor Risk Factor Absent Risk Factor Present
Odds Ratios (95% Confidence Intervals1) Odds Ratios (95% Confidence Intervals1)
Fracture Risk Increased Among PPI users with Risk Factor
Alcohol abuse 1.29 (1.20 –1.39) 1.45 (0.71 – 2.96)
Arthritis 1.26 (1.10 – 1.45) 1.37 (1.22 – 1.54)
Diabetes 1.22 (1.12 – 1.33) 1.43 (1.12 – 1.82)
Kidney Disease 1.26 (1.17 – 1.36) 2.02 (0.80 – 5.06)
Glucocorticoids 1.11 (1.00 – 1.24) 1.51 (1.28 – 1.78)
Fracture Risk Similar or Lower Among PPI users with Risk Factor
Cerebrovascular Disease 1.32 (1.20 – 1.44) 1.06 (0.85 – 1.33)
Dementia 1.36 (1.26 – 1.48) 0.81 (0.58 – 1.14)
Epilepsy 1.30 (1.21 – 1.40) Not Available2
Gait Disorder 1.18 (1.08 – 1.30) 0.90 (0.32 – 2.49)
Hemiplegia 1.30 (1.21 – 1.40) 1.04 (0.33 – 3.27)
Psychoses 1.30 (1.18 – 1.42) 1.06 (0.86 – 1.31)
Smoking 1.32 (1.19 – 1.47) 1.16 (1.00 – 1.35)
Visual Impairment 1.29 (1.20 – 1.39) Not Available2
Anxiolytics 1.29 (1.14 – 1.44) 0.79 (0.39 – 1.60)
≥1 Risk Factor3 0.66 (0.38–1.12) 1.25 (1.16–1.35)
*

P-value interaction=0.02

1

Odds ratios contrast risk of hip fracture among persons with ≥2 years of PPI use vs. reference group. Reference group for all comparisons consists of members with no PPI or H2RA use. For example, for the diabetes strata, the odds ratio evaluates the association between PPI use and fracture risk among persons with a diabetes diagnosis (risk factor present) and the same association among persons without a diabetes diagnosis (risk factor absent). The fracture risk associated with PPI use was 21% greater among diabetics than among non-diabetics.

2

Not available; too few matched cells available for calculation

3

P-value interaction=0.02