Smoking

Cartilage has no blood supply of its own and already lives at low oxygen tension. Smoking narrows vessels through nicotine, displaces oxygen with carbon monoxide, raises systemic inflammatory tone, and impairs the matrix synthesis of fibroblasts and chondrocytes — the same wound-healing impairment documented across orthopaedic surgery. No study has isolated which pathway matters for cartilage; the mechanism here is inferred from the broader biology. The human data, though, point one way.

Cartilage loss on serial MRI

Two independent cohorts watched knees over time with MRI. In 159 men with symptomatic knee osteoarthritis followed for 30 months, current smokers had 2.3-fold odds of medial tibiofemoral cartilage loss and 2.5-fold odds of patellofemoral loss, plus higher pain scores at both timepoints. In 271 community adults without diagnosed OA, smoking was associated with 13.4 microlitres per year of additional medial cartilage volume loss and 11.4-fold odds of bone marrow lesions persisting over two years.

Both studies are observational and both counted few smokers — 19 of 159 in the first, with wide confidence intervals in both — so precision is poor even if direction is consistent.

Repair outcomes after surgery

For anyone considering a cartilage procedure, the surgical data are the most pointed. In a case-control series of autologous chondrocyte implantation — 48 smokers against 66 non-smokers — smokers scored worse before and after surgery, improved less at two years, and every graft failure occurred in a smoker. Cigarettes per day correlated at r = −0.65 with outcome: heavier smoking, worse result. A systematic review of knee ligament and cartilage surgery found the same direction nearly everywhere it looked: seven of eight ligament studies and three of five clinical cartilage series had smokers doing worse, the exceptions being a radiographic-OA count after ACL reconstruction and a microfracture and a mosaicplasty series that found no smoking effect.

The ACI series was non-randomized and single-centre, with baseline differences between groups. It cannot prove causation. But a dose-response gradient inside an already-consistent literature is hard to dismiss.

The inverse association in the pooled data

A meta-analysis of 48 studies covering 537,730 people found smoking inversely associated with osteoarthritis — OR 0.87, apparently protective. The effect was confined to hospital-based case-control studies, and the authors themselves conclude it is likely false, a product of selection bias rather than protection. It marks the boundary of the claim available here: not that smoking causes osteoarthritis, but that it is associated with faster cartilage loss and worse repair outcomes. Those are the outcomes the MRI and surgical studies actually measured.

What I take from it

No randomized evidence exists here, and none ever will — no one randomizes people to smoke. But the asymmetry is total: smoking cessation is unambiguously beneficial on every other health axis, so acting on imperfect evidence costs nothing and the dose-response in the ACI data suggests reduction helps and cessation helps more. The highest-leverage moment is before a cartilage procedure, where the graft-failure data are starkest. No safe-exposure threshold is defined anywhere in this literature, and whether quitting before surgery restores outcomes to non-smoker levels — and over what abstinence window — remains an open question. Vaping and nicotine replacement are essentially unstudied for cartilage.