Intermittent Fasting, BDNF, and Depression: What Holds Up

Intermittent fasting may help some people feel less ruled by food. It does not yet have human clinical evidence as a treatment for major depressive disorder, and the popular claim that fasting reliably raises brain-derived neurotrophic factor (BDNF) does not survive a close look at human data.

BDNF helps regulate synaptic plasticity and has understandable appeal: antidepressants, exercise, and animal models have all linked it to mood-related biology. The leap from that biology to "fasting is an antidepressant" is much larger than social media usually admits. In humans, fasting studies show BDNF going up, down, or nowhere. The one study in people diagnosed with major depressive disorder used a 72-hour fast rather than a standard intermittent-fasting plan and did not measure BDNF.

Food cravings are the most credible brain-related benefit

The best human case for intermittent fasting concerns eating behavior, not depression. A 2024 randomized clinical trial assigned 56 overweight or obese women with food addiction to an 8-hour time-restricted-feeding schedule plus a low-calorie diet (29 participants) or the low-calorie diet alone (27 participants) for eight weeks. The time-restricted group had significantly higher serum BDNF, lower food-craving scores, and better eating-behavior scores than the diet-only group; the reported BDNF difference was statistically significant at P<0.001. BDNF also correlated with cognitive dietary restraint (r=0.468, P<0.001), but not with food-addiction severity.

That pattern matters. It suggests that, in this particular population, the BDNF result tracked self-regulation more closely than compulsive-eating severity. It does not prove that BDNF caused lower cravings. Both groups ate fewer calories, and the time-restricted group also lost more weight, so the study cannot separate the effect of meal timing from negative energy balance and weight change. Still, lower cravings and improved eating behavior are patient-relevant outcomes, even when the biological explanation remains unsettled. Read the trial here.

A 2025 narrative review of food-reward neuroscience offers a plausible framework. Hyperpalatable, ultra-processed foods can repeatedly stimulate the mesocorticolimbic reward system—the network that influences motivation, craving, and reinforcement. The review describes fasting-related changes in dopaminergic signaling, opioid-receptor expression, and appetite hormones such as ghrelin and leptin. BDNF is present in reward-relevant regions including the ventral tegmental area, nucleus accumbens, and prefrontal cortex. Plausibility, however, is not proof that a fasting schedule rewires reward circuitry in every person or treats a mood disorder. The review is available here.

Human BDNF results do not point in one direction

A 2024 systematic review of 16 controlled human studies found five studies with significant BDNF increases, five with significant decreases, and six with no significant change. That is not a reliable "fasting raises BDNF" signal. The studies used different protocols, populations, and durations, which likely explains part of the disagreement, but it also means no single result can carry the broader claim.

Ramadan fasting accounted for five of the 16 studies, and most of those trials reported lower BDNF. One Ramadan investigation reported BDNF about 25% above control at day 14 and 47% above control at day 29, but Ramadan changes more than meal timing: nocturnal eating, sleep, activity, hydration, social routines, and religious motivation can all change at once. Those conditions make a clean fasting-specific interpretation impossible.

Time-restricted feeding produced more consistent increases within the reviewed studies, but the positive studies also included calorie restriction and meaningful weight loss. The review's authors specifically flagged weight and body-composition change as possible explanations for BDNF changes. No human trial has cleanly isolated an eating window from calorie deficit for this endpoint. The full review is available here.

Serum BDNF is a biomarker, not a readout of hippocampal neuroplasticity. It is substantially influenced by platelet storage and release, as well as exercise-related physiology.

This distinction blocks a common overclaim. A change in serum BDNF does not establish that fasting increased BDNF in the hippocampus, changed brain circuitry, or improved depression. No fasting study described here measured serum BDNF alongside a validated central BDNF proxy such as cerebrospinal-fluid BDNF or an established neuroimaging surrogate.

The ketone-to-BDNF story failed its most direct test

The usual mechanism runs like this: fasting raises ketones, ketones reach the brain, BDNF rises, and neuroplasticity improves mood. It is coherent enough to test. In a 2023 randomized mechanistic study, 12 participants fasted for 20 hours, with and without exercise. Fasting produced a documented cerebral fuel shift and a ninefold increase in ketone delivery to the brain. At rest, it changed none of the measured circulating BDNF metrics—serum, plasma, or BDNF per platelet—with P values of at least 0.4637.

The same experiment found that fasting did not amplify the BDNF increase produced by exercise. This does not prove that ketones can never affect BDNF inside the human brain. It does show that substantial fasting-induced ketosis did not raise peripheral BDNF at rest in this controlled setting, and it removes any basis for claiming that fasting boosts exercise's BDNF response. See the Gibbons et al. study.

Depression is the missing clinical test

A 2023 systematic review identified 15 human studies of intermittent fasting and mood. None enrolled people with confirmed major depressive disorder. The eight randomized trials relied on self-report mood questionnaires rather than diagnostic criteria or structured clinical interviews, and no study measured both BDNF and standardized mood outcomes in the same participants. Mood improvements in nonclinical populations may reflect weight loss, better sleep, or lower inflammation; those pathways deserve study, but they are not evidence that fasting treats depression. The review is available here.

The closest evidence comes from a cross-sectional study of 21 hospitalized patients with major depressive disorder and 28 healthy controls during a 72-hour prolonged fast. This was not a standard 16:8 or 5:2 intermittent-fasting intervention, and the investigators measured no BDNF. Among patients with depression, total Beck Depression Inventory-II scores did not change significantly, while the cognitive-affective subscale improved statistically significantly. That result came from the higher-severity subgroup of roughly 10 patients; the lower-severity subgroup did not show the benefit. The subgroup size is too small for a dependable treatment conclusion.

Healthy controls had higher total BDI-II scores during fasting, driven by somatic symptoms such as appetite and fatigue. The metabolic results also complicate simple extrapolation: patients with depression had a much smaller ketone response than healthy controls, 857 versus 4,312 µmol/L (P<0.0001). If ketone signaling matters, people with depression may not respond metabolically like healthy volunteers. The study appears here.

What to do with this evidence

Use intermittent fasting, if it suits you, as a dietary structure that may reduce cravings and support weight-management or cardiometabolic goals. Longer-term trials in adults with overweight or obesity have found benefits versus control diets, including a mean 2.84 kg lower body weight and 3.85 cm lower waist circumference across 24 studies; those results do not establish a depression treatment. See the 2025 meta-analysis.

Do not use serum BDNF as proof that your fasting routine is increasing brain plasticity. Do not replace depression treatment with fasting. People taking glucose-lowering medication, people with a history of eating disorders, pregnant or breastfeeding people, and anyone considering prolonged fasts should discuss the plan with a clinician first.

If raising circulating BDNF is your main goal, exercise has stronger direct human support than fasting, and fasting beforehand did not add to exercise's BDNF response in the 20-hour study. For depression, the honest conclusion is narrower: intermittent fasting remains a hypothesis that needs properly controlled trials in people who actually have the condition.