Zinc

Zinc has a real claim to relevance here: it is essential, it sits in the active site of hundreds of enzymes, and connective tissue cannot be built without it. The intuition behind supplementing it for joints — more zinc, better matrix — is not unreasonable. It just happens that when researchers finally worked out what zinc does inside arthritic cartilage, the answer pointed the other way.

The cartilage mechanism runs the wrong way

The best mechanism paper in this field found that the zinc importer ZIP8 is specifically switched on in osteoarthritic cartilage — human and mouse — and that the resulting zinc influx into chondrocytes activates a zinc-dependent transcription factor (MTF1) that turns on the enzymes that digest cartilage matrix: MMP3, MMP9, MMP12, MMP13 and ADAMTS5. Forcing ZIP8 expression in healthy mouse cartilage was enough to cause OA-like destruction; deleting it protected mice from surgically induced OA.

Two things keep this from being a simple "zinc is bad" story. First, the same chemistry cuts both ways: the matrix-degrading enzymes are themselves zinc metalloenzymes, and cartilage cannot run without zinc at all — the follow-up work on the zinc-buffering metallothioneins found that deleting them also worsens cartilage destruction. Zinc handling in a chondrocyte reads as a tightly buffered system, not a dial where more is better. Second, none of this tested oral zinc. The experiments manipulated transporters and transcription factors, not diet, so they do not show that a zinc capsule harms a human joint. What they remove is the mechanistic case that it helps.

The human trials — old, small, and net null

No trial of supplemental zinc has ever measured an osteoarthritis, cartilage, or joint-structure outcome. None is registered either. The joint trials that exist are from rheumatoid arthritis, and they ran a familiar arc.

In 1976, a preliminary double-blind trial in 24 patients with refractory RA — zinc sulphate 220 mg three times a day for 12 weeks — reported that zinc-treated patients did better on joint swelling, morning stiffness, walking time and their own rating of disease activity. The author called it preliminary and asked for confirmation.

Confirmation did not come. A 6-month randomised double-blind trial at the identical dose found no significant change in any measured parameter — pain, stiffness, articular index, grip strength, ESR — in either group, with the only zinc-specific laboratory change being a rise in alkaline phosphatase. A long-term open trial in 22 patients with severe RA saw brief subjective improvement in a few, deterioration in all, and most participants asking to stop because of the taste and the nausea. The question was quietly abandoned, and it has stayed abandoned for forty years.

The observational data points both ways

Modern epidemiology has not rescued the idea. A 2024 meta-analysis with Mendelian randomization found serum zinc no different in OA patients than controls, with no causal signal — the trace element that did stand out was copper, which was higher in OA. On intake, the two best datasets disagree: in 474 knee-OA patients from the Osteoarthritis Initiative, higher zinc intake was associated with slightly slower deterioration of subchondral bone microarchitecture — a bone surrogate, not cartilage, with tiny per-milligram effect sizes — while in 4,200 NHANES adults, higher zinc intake was associated with more osteoarthritis, not less. Different designs, different endpoints; neither is a trial, and nothing in either supports supplementing.

Deficiency is a different question

Everything above concerns adding zinc to a replete diet. Actual zinc deficiency is a separate matter: zinc is essential for skeletal growth, and zinc status is one of the long-suspected environmental co-factors in Kashin-Beck disease, an endemic cartilage disease of low-selenium regions — with evidence its own reviewers call inconsistent, though zinc reportedly helped repair bone lesions in affected children. Correcting a deficiency and supplementing past sufficiency are different interventions with different evidence, and only the first has any cartilage-adjacent support.

The picolinate form

Picolinate's reputation as the best-absorbed zinc rests on one 1987 crossover in 15 healthy volunteers, in which hair, urine and red-cell zinc rose during the picolinate month but not during citrate, gluconate or placebo. The markers are indirect, the authors were affiliated with naturopathic practice, and no one has repeated the experiment in thirty-eight years. If the absorption edge is real, it is worth noticing what it buys at high doses: a more efficient delivery of the excess.

Safety

The adult tolerable upper intake level for zinc is 40 mg/day from all sources. The concern above that line is specific: zinc induces an intestinal protein that binds copper and blocks its absorption, and the consequences are documented at supplement-realistic doses.

At 50 mg/day — barely above the UL — ten weeks was enough to significantly depress a red-blood-cell copper enzyme (Cu,Zn-superoxide dismutase), along with ferritin and hematocrit, in healthy women. That is a biochemical change, not a symptom. But it is the first step of the one serious zinc toxicity syndrome: chronic copper deficiency, whose end state is a myeloneuropathy — spastic gait and sensory ataxia resembling B12 deficiency, often with anemia and neutropenia. The well-documented modern cases came from people swallowing zinc-laden denture cream for years, at intakes far beyond any supplement label; copper repletion stops the progression, but neurologic recovery is often incomplete.

Stated precisely: a chronic intake of 50–100 mg/day sits at 1.25–2.5 times the UL, inside the range where copper status measurably declines within weeks, and well below the intakes in the reported neurological cases. The mitigations discussed in the clinical literature are unglamorous — periodic copper checks, or a dose at or below the UL — and there is no joint benefit on the other side of the ledger to weigh them against.

What would change this

A single adequately powered trial of zinc supplementation with an OA or cartilage endpoint in zinc-replete adults would turn this from an inferred null into a measured one — none is registered as of this writing. On the mechanism side, a feeding study in the mouse surgical-OA model would answer whether dietary zinc reaches the ZIP8 pathway at all, in either direction. Until one of those exists, zinc belongs in the diet at nutritional doses, and the cartilage case for taking more is not thin — it is absent, with the available biology leaning the other way.