What Nutrition Can and Cannot Do
Diet genuinely affects body weight, systemic inflammation and general health, and all three matter for joints. Eating well is worth doing.
It does not regrow cartilage, and it is not a substitute for diagnosis. If a joint is locking, giving way, hot and swollen, or acutely painful after an injury, that needs a clinician rather than a food diary. Persistent joint pain with morning stiffness lasting more than an hour, or affecting many joints symmetrically, can indicate inflammatory arthritis — a different condition with different treatment, where delay causes avoidable damage.
Nothing on this page should postpone that assessment.
The Two Things With the Best Evidence
Weight, if you are carrying extra
This is the intervention with the largest and most consistent effect in knee osteoarthritis.
In a study of 142 overweight and obese older adults with knee OA undergoing three-dimensional gait analysis, each unit of weight loss was associated with an approximately fourfold reduction in knee-joint compressive force during walking. [1]Losing 10 pounds means roughly 40 pounds less force through the knee with each step.
The Arthritis, Diet, and Activity Promotion Trial (ADAPT) randomised 316 overweight and obese adults aged 60+ with knee OA to healthy lifestyle, diet only, exercise only, or diet plus exercise, for 18 months. The diet-plus-exercise group showed significant improvements in self-reported physical function, six-minute walk distance, stair-climb time and knee pain relative to control — improvements that diet alone or exercise alone did not reliably produce. [2]
⚖️ A note on how this is meant. Body weight is one factor among many, it is harder to change than most advice admits, and joint pain itself makes movement harder — which is a loop, not a failure of character. This is here because the effect size is real and the mechanism is simple physics, not because anyone's pain is their fault.
And exercise counts on its own. WHO recommends 150–300 minutes of moderate or 75–150 minutes of vigorous activity weekly, plus muscle-strengthening on two or more days, and notes that some activity is better than none. [3]
A Mediterranean eating pattern
Among people in the Osteoarthritis Initiative cohort, greater adherence to a Mediterranean diet was associated with better quality of life, including physical function. [4]In the same cohort, higher adherence was associated with better knee cartilage measures on MRI. [5]
⚠️ Both are observational: people who eat this way differ in other ways too, and association is not proof that the diet caused the difference. It remains a sensible pattern with no downside.
Key components:
- Olive oil as the primary fat
- Abundant vegetables and fruit
- Legumes and whole grains
- Moderate fish and poultry
- Limited red meat
- Wine is traditionally part of the pattern; if you don't drink, there is no health reason to start
Anti-Inflammatory Foods
Omega-3 fatty acids
Sources: fatty fish (salmon, mackerel, sardines, anchovies), algae supplements, walnuts, chia, flaxseed.
The evidence: EPA and DHA compete with omega-6 fatty acids and give rise to specialised pro-resolving mediators — resolvins, protectins and maresins — that actively switch off inflammatory responses. [6]
A meta-analysis of 17 randomised trials found that omega-3 supplementation for 3–4 months reduced patient-reported joint pain (SMD −0.26, 95% CI −0.49 to −0.03), morning stiffness (SMD −0.43, 95% CI −0.72 to −0.15), number of tender joints (SMD −0.29, 95% CI −0.48 to −0.10) and NSAID consumption (SMD −0.40, 95% CI −0.72 to −0.08). [7]
⚠️ Three qualifications the headline usually omits. The effects are small. Physician-assessed pain did not improve significantly (SMD −0.14, 95% CI −0.49 to 0.22), nor did the Ritchie articular index. And the trials were in rheumatoid arthritis and other inflammatory conditions — not osteoarthritis, so this evidence does not transfer straightforwardly to wear-related joint pain.
Amount: 2–3 servings of fatty fish weekly, or 1–2 g combined EPA/DHA daily from supplements. See the safety section on bleeding.
Colourful fruit and vegetables
Berries, leafy greens, cruciferous vegetables and orange vegetables. Polyphenols and anthocyanins have antioxidant and anti-inflammatory activity, and cherry anthocyanins inhibit COX-1 and COX-2 in laboratory work. [8]
A randomised crossover study found that strawberry consumption reduced pain and inflammatory markers in obese adults with radiographic knee osteoarthritis. [9]
Turmeric and ginger
Curcumin modulates multiple inflammatory pathways including COX-2 and NF-κB. [10]
The most substantial clinical trial randomised *367 people with knee OA to Curcuma domestica extract 1,500 mg/day or ibuprofen 1,200 mg/day for four weeks. Turmeric extract was non-inferior to ibuprofen on WOMAC total, pain and function scores, with fewer episodes of abdominal pain*. [11]
⚠️ What that does and does not show. It was four weeks, there was no placebo arm — so both groups improving tells you less than it appears — and it used 1,500 mg/day of turmeric extract, which is not the same as the 500 mg of isolated curcumin sold in most capsules. Encouraging, not settled. Curcumin also has poor oral bioavailability, which is why formulations add piperine or use liposomal delivery.
Green tea
EGCG inhibits IL-1β-induced nitric oxide synthase expression and COX-2 activity in cultured human chondrocytes. [15][16]These are cell-culture findings — they describe a plausible mechanism, not a demonstrated benefit in people. Green tea is a pleasant drink with no downside at 2–3 cups daily; that is the honest basis for including it.
Collagen
A 24-week randomised, placebo-controlled trial gave 10 g daily collagen hydrolysate to athletes with activity-related joint pain. Of 147 randomised, 97 provided evaluable data; the collagen group showed reduced joint pain during activity. [12]
⚠️ The participants were athletes with no evidence of joint disease, so this does not speak to osteoarthritis, and the attrition (147 → 97) weakens it further.
Vitamin D — read this before supplementing for your joints
Vitamin D receptors are present in chondrocytes, and low vitamin D status has been associated with radiographic knee osteoarthritis in cohort studies. [13]That association led to an obvious hypothesis, which was then tested properly.
It failed. A two-year randomised, placebo-controlled, double-blind trial gave 146 people with symptomatic knee OA cholecalciferol 2,000 IU/day, escalated to raise serum 25-hydroxyvitamin D above 36 ng/mL. Serum levels rose by a mean 16.1 ng/mL. And:
- Knee pain fell in both groups, with no significant difference at any time point
- Cartilage volume loss was identical: −4.30% versus −4.25%, P = .96
- "There were no differences in any of the secondary clinical end points"
The authors concluded that vitamin D supplementation at a dose sufficient to raise levels above 36 ng/mL did not reduce knee pain or cartilage volume loss. [21]
What follows from that:
- If you are deficient, correcting it is worthwhile — for bone and muscle, which matter for joints indirectly. Ask for a blood test.
- If you are already replete, taking more will not help your joints. Targets above the standard sufficiency threshold have been tested for exactly this purpose and produced nothing.
- More is not safer. Sustained high-dose vitamin D can cause hypercalcaemia. Do not chase a high number.
Foods to Limit
Processed and refined foods — high in omega-6 fats, trans fats and advanced glycation end-products, which form when food is cooked at high temperature and promote oxidative stress. [19]Limit fast food, processed snacks, refined carbohydrates and processed meats.
Sugar — high-glycaemic-index carbohydrate increases NF-κB activation in mononuclear cells. [18]Aim for under 25 g added sugar daily; watch sugary drinks, desserts, and corn syrup/maltose/dextrose on labels.
Excessive omega-6 oils — the modern ratio of omega-6 to omega-3 is far higher than historical intakes, which favours pro-inflammatory eicosanoid production. [20]Prefer olive and avocado oil.
Nightshades — some people report that tomatoes, peppers, aubergine and potatoes worsen their symptoms. There is no good evidence that nightshades cause joint inflammation. If you want to test it for yourself, eliminate for 2–4 weeks, reintroduce one at a time and keep a symptom diary. ⚠️ If symptoms are worsening, see a clinician rather than narrowing your diet further — progressive joint pain has causes that food elimination will not find.
⚠️ Supplements: Safety First, Then Evidence
Before you start anything
Tell your doctor or pharmacist what you are taking. Supplements are not inert, and the interactions below are specific rather than theoretical.
Bleeding and clotting. A review of dietary supplements and haemostasis identified fish oil, garlic, ginkgo, ginger, ginseng and vitamin E as capable of interfering with clotting, principally by reducing platelet aggregation. [14]This matters if you:
- take warfarin, a DOAC (apixaban, rivaroxaban, edoxaban, dabigatran), aspirin or clopidogrel — tell the clinician who manages it before adding fish oil or high-dose turmeric
- have surgery or a procedure planned — tell the surgical team everything you take, and ask when to stop
- bruise or bleed easily, or have a bleeding disorder
⚠️ This applies with particular force if you have been told to stop NSAIDs or aspirin before a procedure or trip. Stopping an antiplatelet drug while adding antiplatelet supplements is not a net reduction in risk, and neither instruction is aware of the other unless you tell someone.
Warfarin and leafy greens. Green vegetables are high in vitamin K, which opposes warfarin. The right advice is consistency, not avoidance — keep your intake steady and tell your anticoagulation clinic if you change your diet substantially, so your INR can be monitored. Do not cut out vegetables on your own.
Upper limits. More is not better and can be harmful. Vitamin D in sustained high doses can cause hypercalcaemia; vitamin A is teratogenic in pregnancy; selenium has a narrow safe range. Stay within labelled doses unless a clinician has told you otherwise.
Other sensible practice:
- Choose third-party tested products (USP, NSF, ConsumerLab) — supplement quality is inconsistent
- Start one at a time, so you know what is doing what
- Allow 8–12 weeks before judging
- Tell every clinician you see, including dentists and surgeons
What the evidence actually supports
Glucosamine and chondroitin, specifically. The largest trial — GAIT, 1,583 patients, randomised to glucosamine 1,500 mg, chondroitin 1,200 mg, both, celecoxib or placebo for 24 weeks — found that glucosamine and chondroitin were not significantly better than placebo for the primary outcome. [17]A benefit appeared in an exploratory subgroup with moderate-to-severe pain (n=354), which has not reliably replicated and should not be treated as an established finding. Several clinical guidelines now recommend against routine use. Harm is low, and cost is not nothing.
Food first. Whole foods provide fibre, phytochemicals and nutrient combinations supplements do not reproduce. Supplements address specific deficiencies; they are not a parallel diet.
Hydration
Synovial fluid and cartilage are largely water, and dehydration is worth avoiding on general grounds. Two to three litres daily is a reasonable target, adjusted for body size, activity and climate — more in hot weather. Pale yellow urine is a practical guide.
⚠️ Hydration will not treat joint disease, and there is no evidence that drinking more than you need improves joints.
Sample Anti-Inflammatory Day
Breakfast — Greek yoghurt with blueberries, walnuts and chia seeds; green tea.
Alternative: steel-cut oats with tart cherries and almond butter.
Lunch — Large salad with mixed greens, cherry tomatoes, cucumber, avocado and grilled salmon; olive oil and lemon dressing.
Alternative: lentil soup with wholegrain bread.
Snack — A handful of almonds and an apple.
Alternative: carrots with hummus.
Dinner — Grilled mackerel or chicken; roasted broccoli and Brussels sprouts with garlic; quinoa or brown rice; turmeric-ginger tea.
Evening — Tart cherry juice; chamomile or ginger tea.
FAQs
How quickly will dietary changes help?
Weight change and its effect on joint load happen on the timescale of weeks to months. Effects on symptoms are variable and often modest. Be sceptical of any specific promised timeline, including ours — the honest answer is that it differs between people and some people notice nothing.
Do I need to be perfect?
No. Consistency over time matters more than perfection, and an 80/20 approach is sustainable where perfectionism is not.
Can diet replace medical treatment?
No. Diet supports general health and can reduce joint load through weight change. It does not treat inflammatory arthritis, repair structural damage, or substitute for diagnosis. If you have persistent or worsening joint symptoms, see a clinician — and tell them what you are taking.
What about supplements versus food?
Food first. Supplements address specific deficiencies or deliver doses hard to reach through diet. They are not a replacement, and each one is a decision with its own evidence and its own interactions.
Is eating this way expensive?
Not necessarily. Cutting processed food, buying frozen fish, choosing seasonal produce, buying nuts in bulk and cooking at home often reduces overall food cost.
Can I drink alcohol?
If you drink, keep it moderate. Excess alcohol promotes inflammation and disrupts sleep and recovery. If you don't currently drink, there is no health reason to start, and claims about red wine and resveratrol are weaker than they are usually presented.
What about coffee?
Coffee contains polyphenols and, for most people, 1–3 cups daily is compatible with this pattern. If it disrupts your sleep, that matters more for recovery than the polyphenols do.