In an important and interesting study, Laura Targownik and colleagues1 state that the effect of exposure to proton pump inhibitors on osteoporosis-related fractures might be “similar in size to [the effect of] other established osteoporotic-fracture risk factors, such as smoking, low body mass index and excessive alcohol intake.” However, the authors did not take any of these factors into account in their analysis. Were the case and control groups comparable in terms of these 3 variables? If not, what is the role of these variables in the observed associations? For example, people who use proton pump inhibitors might be less likely to be active smokers because of gastresophageal reflux disease but, at the same time, they could be more likely to have a history of smoking than people who do not use this class of drugs.
This may have an impact (of unknown magnitude) on the estimation of the effect of proton pump inhibitors.
This study relies on the analysis of administrative data, which were probably collected for purposes quite distinct from research into the causes of osteoporosis. This does not mean that such data are inadequate for health research, but that we should be more critical about the inferences drawn from them.
Footnotes
-
Competing interests: None declared.
REFERENCE
- 1.