Does drinking a lot of coffee increase the risk of bone fractures?
For most people, coffee does not appear to be a proven cause of bone fractures, but for women who drink a lot of coffee some caution is reasonable; in that case, make sure to get enough calcium and vitamin D in your diet.
A large meta-analysis of more than 500,000 participants found that the risk of hip fracture in coffee drinkers was barely different from that in people who drink little coffee. Bone density also did not differ in any measurable way. Two separate studies that used genetic variants to investigate causality -- a method that is less susceptible to confounding factors such as lifestyle -- reached the same conclusion: drinking coffee probably has no causal effect on bone density or fracture risk, in either men or women.
There is one caveat for women. One large umbrella review, a summary of multiple meta-analyses, found an association in women between high coffee consumption and an increased fracture risk. In men, that association was absent. This is a statistical association, not a proven cause-and-effect relationship. Moreover, other large analyses contradict this association in women, so the picture on this point is divided.
At the molecular level there are indications that caffeine can promote bone breakdown by influencing calcium metabolism. But in studies in humans, this mechanism does not translate in a consistent way into measurable bone loss or more fractures. Laboratory findings and what actually happens in people therefore diverge here.
Excessive coffee consumption is mentioned in several review articles as something to be mindful of when preventing osteoporosis, alongside proven risk factors such as smoking, vitamin D deficiency, and low calcium intake. Smoking was, in the genetic research, causally linked to a higher fracture risk; for coffee that was not found. No specific threshold is known above which coffee definitively becomes harmful to your bones.
Evidence based on one umbrella review (PMID 29167102), one meta-analysis of 20 studies with >500,000 participants (PMID 37374383), two Mendelian randomisation studies (PMID 31482193 and 40692555), and three narrative/mechanistic reviews (PMID 34981132, 34002830, 23871889). The causality question is most robustly answered by the genetic studies.