The Relationship of Aging and Smoking With Rotator Cuff Disease: A Systematic Review and Meta-analysis.

Rotator cuff disease was analysed as one broad group. Because the included studies defined tears, tendinopathy, tendinitis and syndrome differently, the review cannot tell you anything about any one of those diagnoses on its own. Five of the seven age studies and five of the ten smoking studies were cross-sectional, so age or smoking and the diagnosis were measured at the same time. That supports association, not cause. The age result is driven by one very large study, which the authors point out themselves. Smoking was usually recorded only as never, former or current. Most studies did not report pack-years, so a dose-response could not be examined. The authors could not formally check for small-study bias, and they say the pattern of stronger results in smaller studies that dichotomised smoking makes that bias likely. The findings cover traditional tobacco smoking only. Vaping and e-cigarettes were not studied. Diagnostic criteria and case definitions varied across studies, with little standardisation between them.

The article text supplied contains no funding statement and no conflict-of-interest declaration. The publication record lists NIH extramural research support.

Do not repeat the two-fold rise in risk among current smokers as a settled fact about smoking. That figure comes from five smaller studies that recorded smoking only as a simple yes or no. The five studies that separated never, former and current smokers found no association at all, and the authors' largest studies found nothing either. The age findings describe how common the diagnosis is in older people, not why it happens or how they respond to treatment.