He had two bottles in his hands. A man in his late sixties, standing in the supplement aisle of a Madison drugstore, comparing labels the way someone reads a map in a foreign city. He’d been there about four minutes when I noticed him. One bottle in each hand, looking from one to the other, and then back again. He finally put one back, dropped the other in his basket, and walked away.

I know exactly what he was doing, because I’ve done it. The labels on both bottles were written to imply that the science is settled, that the product works, and that someone else has already done the thinking for you. None of that is accurate. I watched him and thought: this is the problem in a single image. A forty-billion-dollar industry built on the gap between what people are afraid of and what science has actually managed to prove.

The Dietary Supplement Health and Education Act of 1994 means the FDA doesn’t evaluate supplements for efficacy before they reach a pharmacy shelf. Manufacturers are responsible for safety, but the agency can’t require proof that a product works before it goes to market. That’s not a technicality. It’s the structure that makes the supplement aisle what it is.

I’m not here to tell you that nothing in that aisle is worth buying. Some of it is. But joint pain is among the most common complaints I hear from people managing chronic discomfort, and the supplements marketed for it vary enormously in how much honest evidence actually exists behind them. Here’s what the research shows.

Glucosamine and chondroitin

These are the flagships. They’ve been in that aisle longer than most people can remember, and the theory behind them is genuinely reasonable. Glucosamine is a building block for cartilage. Chondroitin helps cartilage retain water and resist compression. Supplementing both, the argument goes, might slow cartilage degradation in arthritic joints and reduce pain.

The GAIT trial, published in the New England Journal of Medicine in 2006, was the largest and most rigorous attempt to test this. It was a multicenter, randomized, placebo-controlled trial involving roughly 1,600 people with knee osteoarthritis. The headline finding was that neither glucosamine alone, chondroitin alone, nor the two in combination reduced pain significantly more than placebo across the overall group.

That sentence is what the marketing leaves out.

What the marketing tends to lead with is the subgroup finding. Among participants with moderate-to-severe knee pain specifically, the combination of glucosamine and chondroitin showed a meaningful benefit: about 79 percent reported significant pain reduction compared to 54 percent on placebo. That’s a real number from a real, well-run trial. It also applies to a specific population, not everyone with aching joints.

If you have mild general joint discomfort, the evidence for these supplements is thin. If you have moderate-to-severe knee pain from osteoarthritis, there’s something here that’s genuinely worth discussing with your doctor. The supplements are considered safe at standard doses, and the cost is modest. But the marketing tends to flatten the distinction between “this worked for a subgroup with severe knee OA” and “this works for joint pain,” and those aren’t the same claim.

Fish oil (omega-3 fatty acids)

The mechanism here is real and well-documented. Omega-3 fatty acids, particularly EPA and DHA, reduce production of certain inflammatory compounds. The anti-inflammatory pathway isn’t in dispute. What’s more variable is how much that mechanism translates into meaningful joint pain relief in the doses most people take.

The evidence is stronger for rheumatoid arthritis than for osteoarthritis. Several trials in RA populations have shown modest reductions in morning stiffness, tender joint counts, and NSAID use. For osteoarthritis, the data is thinner and more inconsistent.

What I can say without much hedging is this: a diet higher in fatty fish (salmon, sardines, mackerel, herring) has well-supported benefits that go beyond any single supplement study, and fish oil is among the safer options in this category. The mechanism is credible. The evidence for joint pain specifically is modest, not absent. The risk is low. If you’re going to take something in this category, this one has a more honest story than most.

One thing worth knowing: fish oil at higher doses can affect how certain blood thinners work. If you’re on warfarin or another anticoagulant, ask before starting.

Collagen peptides

The marketing on collagen has gotten aggressive. The story is that supplemental collagen peptides travel to the joints and stimulate cartilage repair, reducing pain and improving function. Some small, recent trials have found improvements in joint pain scores, particularly in people with activity-related joint discomfort rather than established osteoarthritis.

“Small” is carrying significant weight in that sentence. When I describe a study as small, I mean it enrolled somewhere between twenty and a hundred people, often over weeks rather than months. Small trials can detect real signals. They can also reflect statistical noise, placebo effect, or features of the specific population that don’t generalize. They’re hypothesis-generating, not conclusions. The follow-up trials needed to confirm or refute what the small collagen studies found haven’t been completed at the scale necessary to call this evidence-based in the same way that fish oil’s anti-inflammatory mechanism is evidence-based.

The downside risk of collagen peptides is genuinely low, and the cost per serving is modest. I’m not saying don’t take them. I’m saying you should understand what kind of evidence you’re actually relying on when the label cites “clinical studies,” because that phrase can describe a fifty-person trial that’s never been replicated.

Turmeric and curcumin

Curcumin is the active compound in turmeric, and it has legitimate anti-inflammatory properties. The problem isn’t the mechanism. It’s getting the compound to work inside a human body.

Curcumin is poorly absorbed when taken orally. Most of what you swallow doesn’t enter the bloodstream in concentrations sufficient to do much. Supplement manufacturers have worked around this to varying degrees with formulations that include piperine (from black pepper, which improves absorption substantially), phospholipid complexes, or nano-delivery systems. Some of these do help. The evidence on even better-absorbed formulations for joint pain remains limited and inconsistent.

The research pipeline here is genuinely interesting. Scientists are working on delivery systems that could make curcumin a meaningful therapeutic option. That work is ongoing. What’s on the pharmacy shelf right now is mostly a compelling premise that the science hasn’t fully caught up to. Whether that changes in five years, I can’t say. What’s true today is what the current trials show, and the current trials are preliminary.

Vitamin D

This one belongs on the list but in a different category. Vitamin D isn’t a joint pain remedy in the direct sense. What it is, is a nutrient whose deficiency produces musculoskeletal weakness, reduced bone density, and a general decline in how efficiently the whole system works, including the muscles and structures around joints.

I live in Wisconsin. My vitamin D level drops in winter every year, and my bloodwork proves it. That’s why I take a supplement in the colder months. I’m not taking it because it will reduce my knee pain. I’m taking it because deficiency is genuinely common in people who live in northern latitudes or have limited sun exposure, and deficiency makes the musculoskeletal system work less well across the board.

Getting your level checked is a low-cost, low-effort thing to do before assuming you’re fine. A 25-hydroxyvitamin D blood test is standard and inexpensive. If your level is normal, supplementing won’t give you more joint function. If it’s low, correcting it matters.

Boswellia

This is the one most people in that drugstore aisle have never heard of, and it has more credible evidence behind it than almost anything else on this list.

Boswellia serrata, sometimes called Indian frankincense, contains acids that inhibit 5-lipoxygenase, an enzyme involved in a specific inflammatory pathway relevant to joint pain. That mechanism is well-characterized. Multiple randomized, placebo-controlled trials have found meaningful reductions in knee pain and improved mobility in people with osteoarthritis. The results have been more consistent than what exists for glucosamine in the overall population, and the safety profile is generally clean.

It doesn’t have a decades-long marketing presence in the US, which is partly why it hasn’t become a household name. That’s not a reason to discount it. The evidence is what it is, and by the standard of evidence relevant to the rest of this list, boswellia holds up reasonably well. It’s worth knowing about.

If you’re going to take something for joint pain and you want to be guided by the evidence rather than by packaging, here’s how I’d think through it. Start with your vitamin D level, because deficiency is fixable and it matters for more than your joints. If you have moderate-to-severe knee pain from osteoarthritis, the glucosamine-chondroitin combination has a specific and honest evidence base, but for that population specifically. For something with a well-documented mechanism and reasonably consistent trial results, boswellia is the sleeper here. Fish oil is reasonable if you’re not already eating fatty fish regularly. Collagen and turmeric aren’t wrong, but they’re getting ahead of where the science currently sits.

Don’t take all of them simultaneously. The interaction research in this area is thin, and more supplements don’t add up to more benefit. Your liver and kidneys are processing all of it, and the cost compounds faster than the evidence does.

Questions to bring to your doctor

What supplement interactions do I need to know about given my current medications? Fish oil and boswellia can affect certain anticoagulants, and this is worth knowing before starting either.

Is my vitamin D level actually low? Ask specifically for a 25-hydroxyvitamin D blood test rather than assuming your level is fine. The test is routine and the result matters.

What does the research say about glucosamine for my specific type of joint pain? The GAIT trial evidence applies to knee osteoarthritis in particular. The picture may look different for hip involvement or for inflammatory forms of arthritis.

Are there treatments for my joint pain with stronger evidence than supplements? Physical therapy, weight management if relevant, and certain medications have more robust evidence for osteoarthritis than most supplements do. This question is worth asking before deciding the supplement aisle is where to start.

Carol Gifford spent fourteen years as a registered nurse before moving into health communication and writing. She covers medicine and wellness for the Sunday Evening Review. She does not offer medical advice. Talk to your doctor about your specific situation.