Let me start where I start in the consulting room: bipolar disorder is treated with medication. Mood stabilisers and, where indicated, antipsychotics prevent episodes, and no dietary pattern substitutes for them. Anyone suggesting otherwise is proposing something dangerous.
Inside that frame, nutrition earns its place for three concrete reasons — and they are more practical than the usual anti-inflammatory-diet article suggests.
Why inflammation comes up in bipolar disorder at all
Bipolar disorder is associated with elevated inflammatory markers, more so during acute episodes than in euthymia. Whether inflammation drives the illness, results from it, or reflects shared risk factors — disturbed sleep, metabolic syndrome, smoking, medication effects — is unresolved. What is not in question is that people with bipolar disorder carry substantially higher cardiometabolic risk and, on average, a shortened life expectancy driven largely by cardiovascular disease rather than by the psychiatric illness itself.
So the nutritional argument does not depend on winning the inflammation debate. Reducing cardiometabolic risk is worth doing regardless, and it happens to overlap almost entirely with what anti-inflammatory eating means in practice.
The three reasons I raise food
1. Medication side effects run through metabolism. Several effective mood stabilisers and antipsychotics cause weight gain, insulin resistance and dyslipidaemia. These side effects are a leading reason people stop taking medication that was working — which is the single most dangerous thing that can happen in bipolar disorder. Anything that makes the medication easier to stay on protects against relapse.
2. Blood glucose stability affects how the day feels. Large refined-carbohydrate meals produce energy and mood swings that are unpleasant for anyone and confusing in an illness where you are monitoring mood closely. Steadier intake makes self-monitoring cleaner.
3. There is real, if early, trial evidence for dietary patterns in mood. Mediterranean-pattern dietary interventions have improved depressive symptoms in randomised trials in unipolar depression. That work has not been replicated specifically in bipolar depression, and I say so — but the pattern is safe and its cardiometabolic benefit is established.
What we actually recommend
Not a protocol. Five changes, in the order they usually matter:
- Protein and fibre at every meal. The most reliable lever on blood glucose and satiety, and the one that helps most with medication-related appetite increase. Eggs, fish, poultry, legumes, Greek yoghurt, tofu; vegetables, beans, whole grains.
- Oily fish twice a week. Salmon, sardines, mackerel, herring. Omega-3 evidence in bipolar disorder is stronger for depressive symptoms than for mania and is modest either way — but oily fish is worth eating on cardiovascular grounds alone.
- Olive oil as the default fat, with nuts and seeds. This is the core of the Mediterranean pattern and the part with the best cardiovascular data.
- Cut sugar-sweetened drinks first. Of all single changes, removing liquid sugar gives the most metabolic return for the least effort. Diet drinks are a reasonable interim step.
- Alcohol: less, and honestly accounted for. Alcohol is a mood destabiliser, a sleep disruptor and a relapse risk in bipolar disorder, and it interacts with the medications. This is the nutritional change with the largest psychiatric effect, and the one patients most often leave out of the food diary.
Foods to reduce rather than eliminate: ultra-processed snack foods, refined baked goods, processed meats, and deep-fried food. Elimination frameworks tend to fail in an illness where routine and simplicity matter, and where an episode can flatten anyone's cooking capacity for a fortnight.
The safety points that are specific to bipolar disorder
This is the part generic diet advice misses, and it matters:
- Lithium and sodium are coupled. A sudden large reduction in salt intake raises lithium levels; a large increase lowers them. Dehydration — from illness, heat, intense exercise or a very low-carbohydrate diet's early diuresis — raises lithium levels and can push you toward toxicity. If you are on lithium, keep salt and fluid intake steady and tell your prescriber before any major dietary change.
- Fasting and very-low-carbohydrate diets need supervision. Ketogenic diets are being actively studied in bipolar disorder and some early results are interesting, but they alter fluid and electrolyte balance and can affect lithium levels. This is not a self-start intervention on a mood stabiliser.
- Caffeine interacts with lithium clearance and, separately, degrades sleep — the most important variable in episode prevention. Stopping caffeine abruptly can raise lithium levels.
- Grapefruit affects some medications metabolised by CYP3A4, including certain antipsychotics. Worth checking your specific list.
- Rapid weight loss attempts can destabilise sleep and mood. Slow beats dramatic here, always.
Where supplements fit
Modestly. Omega-3 has the most evidence, mainly for depressive symptoms, and is reasonable as an adjunct. Vitamin D, B12 and folate are worth measuring and correcting if low. Beyond that, caution is warranted: St John's wort can precipitate mania and interacts with numerous medications, and any supplement marketed as a mood stabiliser alternative should be treated as a red flag. Everything you take belongs on the list your prescriber sees — see our approach to nutrition and supplements.
How this looks over a year
The patients who do well with this do not overhaul their diet. They change two things, keep them through an episode, and add a third when the first two are automatic. That is also the honest expectation to set: in an illness with a relapsing course, the nutritional plan that survives a bad month is worth more than the optimal one that collapses.
In our bipolar disorder treatment we build this into medication review — weight, blood pressure, lipids and HbA1c monitored on the medications that require it, with dietary work aimed at the specific risk the medication carries. My clinical focus is reproductive and perinatal psychiatry alongside mood disorders, and if you are pregnant, planning a pregnancy or postpartum, both nutritional needs and medication decisions change materially; you can read more about how I work.
FAQs
Can diet replace medication for bipolar disorder?
No. Bipolar disorder is managed with mood stabilisers and, where indicated, antipsychotics, and stopping them carries a high risk of relapse and hospitalisation. Nutrition is an adjunct that reduces cardiometabolic risk and can make medication easier to tolerate.
Is the ketogenic diet safe for bipolar disorder?
It is being studied and some early results are interesting, but it should not be started unsupervised on a mood stabiliser. Ketogenic diets shift fluid and electrolyte balance, which can raise lithium levels toward toxicity. Any trial of it needs your prescriber involved and lithium monitoring.
Does salt intake affect lithium?
Yes, directly. Reducing salt intake raises lithium levels and increasing it lowers them, and dehydration raises levels too. If you are on lithium, keep sodium and fluid intake reasonably consistent and tell your prescriber before any significant dietary change, including starting a low-salt or low-carbohydrate diet.
Do omega-3 supplements help bipolar disorder?
The evidence is modest and stronger for depressive symptoms than for mania. Omega-3 is reasonable as an adjunct alongside medication, and eating oily fish twice a week is worthwhile for cardiovascular reasons regardless of its psychiatric effect.
Which foods should I avoid on antipsychotic medication?
There is no forbidden list, but sugar-sweetened drinks and ultra-processed foods are worth reducing first, because the main risk these medications add is metabolic — weight gain, insulin resistance and lipid changes. Grapefruit can affect the metabolism of some antipsychotics, so check your specific medication.
