Table 2.
Study | Design | Participants | Dietary intervention and duration | Adherence and, where available, confirmation by measurement and presence of ketosis | Main mood and psychiatric symptom outcomes | Adverse effects |
---|---|---|---|---|---|---|
Observational studies | ||||||
Campbell & Campbell (2019)50 | Online observational analytic study | n = 274; bipolar disorder (age and gender n.r.) | Ketogenic diet (85 participants, 165 online posts) compared with omega-3 or vegetarian diet (94 posts) Carbohydrate intake n.r. Duration not consistently reported, but where available (or derived) varied: 1–5 months, >6 months and >12 months |
92.9% of participants ‘very likely’ to have achieved a state of ketosis | Remission or significant mood stabilisation reported in 56.4% of posts on ketogenic diet and 14.9% of posts on vegetarian diet or omega-3 supplementation In ketogenic diet, many detailed reports of the improvements experienced and several lasting for extended periods (months to years) In ketogenic diet, fewer episodes of depression (41.2%), improved clarity of thought and speech (28.2%), increased energy (25.9%) and weight loss in 25.9% Any mood destabilisation reported in 4.8% of posts on ketogenic diet and 10.6% of posts on vegetarian diet or omega-3 supplementation |
No instances of hospital admission or care-seeking reported in posts for either diet group Difficulties in keto-adaptation reported in 10.6% of posts |
Cohort studies | ||||||
Danan et al (2022)59 | Retrospective cohort study | n = 31; psychiatric in-patients: 13 bipolar II disorder, 12 schizoaffective disorder, 7 MDD (mean age: 50 years; 71% female) | Ketogenic diet (carbohydrate <20 g/day) Duration: at least 14 days, mean 59.1 days |
3 participants discontinued before 14 days. Ketosis (measured via urinary ketones) achieved by 64% of remaining participants. Dietary adherence characterised as excellent (39%), good (43%) or fair (18%) | Symptoms of depression and psychosis and overall clinical severity significantly reduced All patients who completed the HRSD achieved a reduction of ≥4 points and 95% achieved a reduction of ≥6 points. All patients who were assessed with the MADRS achieved a reduction of ≥6 points Number or dose of psychotropic medications was reduced in 64% of participants |
Most patients reported initial symptoms of keto-adaptation (headache, insomnia, irritability, excitation, dizziness, carbohydrate cravings), which resolved within 2 weeks; beyond this, 13% reported adverse effects (fat intolerance, diarrhoea, vomiting) |
Kunin (1976)68 | Cohort study | n = 73; psychiatric out-patients with symptoms of anxiety, depression and ‘dysperception’ (age and gender n.r.) | Ketogenic diet (carbohydrate 0 g/day) until ketosis reached (or maximum 5 days); then ‘optimal carbohydrate level’/low carbohydrate (<120 g/day, mean 52 g/day), duration n.r.; then higher carbohydrate diet (>120 g/day), duration n.r. | 2 participants discontinued within 5 days. Ketosis measured via urinary ketones | Symptoms of anxiety, depression and dysperception improved with ketogenic and ‘optimal carbohydrate level’ diets in 82% of participants | 60% of participants reported transient adverse effects, including fatigue, nausea, weakness, headache and palpitations; improved after administration of potassium salts |
Case reports | ||||||
Chmiel (2021)64 | Case report | n = 1; ultra-rapid cycling bipolar disorder (not further specified but suggestive of bipolar II disorder) (27-year-old male) | Low carbohydrate high fat diet for 1 year; then ketogenic diet (carbohydrate <30 g/day, 2500 kcal, 15% protein, 80% fat, 5% carbohydrate) for 1 year; then ketogenic diet with 1 day fast every 7–10 days for 1 year | Average reported blood ketones: year 1 β-HB = 0.3–0.5 mmol/L; year 2 β-HB = 1.5–3 mmol/L; year 3 β-HB = 5 mmol/L on fasting days | Depression reduced, mood stabilised, increased energy, improved sleep, cognitive function and concentration, and elimination of anxiety; no hypomania Remission of depression in year 3. Discontinuation of quetiapine (previously up to 300 mg/day) and reduction of lamotrigine to 100 mg/day (previously up to 300 mg/day) |
n.r. |
Cox et al (2019)65 | Case report | n = 1; type 2 diabetes and comorbid MDD (65-year-old female) | Ketogenic diet (65% fat, 25% protein, 10% carbohydrate) for 3 months | Blood ketones averaged 1.5 mmol/L by week 12 | PHQ-9 score decreased from 17 (baseline) to 0 (week 12); reported increased self-efficacy and self-confidence, increased energy, improved sleep, stability in mood and clearer cognition | n.r. |
Ehrenreich (2006)60 | Case report | n = 1; panic disorder (47-year-old female) | Atkins diet (carbohydrate intake n.r.) for 4 weeks | Not assessed, but 17 lb (7.7 kg) weight loss reported | Increase in baseline anxiety level over the course of the diet | Internal sensation of ‘shakiness’, frequent panic attacks; resolved on cessation of diet |
Kraft & Westman (2009)66 | Case report | n = 1; schizophrenia and comorbid depression (70-year-old female) | Ketogenic diet (carbohydrate <20 g/day) for 12 months | Reported 2–3 isolated episodes of non-adherence lasting several days; 10 kg weight loss reported | Reported increased energy and no longer experienced auditory or visual hallucinations | n.r. |
Palmer (2017)58 | Case report | n = 2; schizoaffective disorder (33-year-old male, 31-year-old female), both with comorbid MDD | Ketogenic diet (carbohydrate intake n.r.); male for 12 months; female for 4 months | Male discontinued diet on 5 occasions and experienced relapse of positive and negative symptoms within 1–2 days; female discontinued diet once and developed paranoia and delusions | Reductions in hallucinations and delusions, improved mood, energy and concentration. PANSS scores reduced from 98 to 49 in the male and from 107 to 70 in the female |
n.r. |
Phelps et al (2012)63 | Case report | n = 2; bipolar II disorder (69-year-old female, 30-year-old female) | Ketogenic diet (70% fat, 22% protein, 8% carbohydrate for 30-year-old; details n.r. for 69-year-old); 69-year-old for 2 years; 30-year-old for 3 years | Ketosis measured via urinary ketones: ranged between 0 and 80 mg/dL (over course of 7 months in 69-year-old; ketosis measurement n.r. for 30-year-old) | Sustained mood stability; diet enabled both to discontinue lamotrigine; no increase in anxiety. | None |
Pieklik et al (2021)67 | Case report | n = 1; ‘mood disorder’ (not specified) with comorbid emotion dysregulation, body dysmorphic disorder and an eating disorder (21-year-old female) | Very low calorie ketogenic diet (details n.r.) (carbohydrate intake n.r.) for 4 weeks | Did not fully comply with diet (details n.r.); no measures of ketosis; weight reduced from 113.5 kg to 102 kg, some metabolic parameters improved | Partial improvement in well-being, mood stabilised, decreased anxiety, no suicidal thoughts. BDI score improved from 40 (severe depression) to 23 (moderate) Other interventions included sertraline, trazodone, metformin (timing, doses and duration n.r.), psychotherapy and psychoeducation |
Did not experience adverse effects but did not want to continue nutritional intervention; overall ‘therapeutic cooperation’ reported as ‘difficult’ |
Saraga et al (2020)62 | Case report | n = 1; bipolar I disorder (60-year-old female) | ‘Mildly ketogenic’ diet (ratio of grams of fat to grams carbohydrate + protein of 2–3:1); duration n.r. | Ketosis measured via urinary ketones: 0.05–0.4 g/L | Decreased anxiety, maintenance of euthymia; enabled discontinuation of Sertindole; patient describesd clear improvement on both depressive and manic symptoms | n.r. |
Yaroslavsky et al (2002)61 | Case report | n = 1; in-patient with bipolar I disorder (49-year-old female), treatment-resistant, rapid cycling | Ketogenic diet (carbohydrate intake n.r.) for 4 weeks | Urinary ketosis was not confirmed, nor was there any weight loss. Patient adherence reported as very good | No clinical improvement | n.r. |
β-HB, beta-hydroxybutyrate; BDI, Beck Depression Inventory; GAD-7, seven-item Generalised Anxiety Disorder questionnaire; HRSD, Hamilton Rating Scale for Depression; MADRS, Montgomery–Åsberg Depression Rating Scale; MDD, major depressive disorder; n.r., not reported; PANSS, Positive and Negative Syndrome Scale; PHQ-9, nine-item Patient Health Questionnaire (depression module).