The mind does not live only in the head.
It lives in the whole body.
Around 25 controlled trials and meta-analyses suggest that exercise, sauna, sleep, nutrition, micronutrients, light and social bonds can reach effect sizes in depression and anxiety that lie in the same range as those of psychotropic medication. For cold this explicitly does not apply, there is only a single case report so far. Direct head-to-head trials are almost entirely missing, and the numbers come from very different study designs with different control groups. Why current guidelines have not yet caught up, and what an honest, holistic, integrative psychiatry can already do today.
Imagine you walk into your doctor's office with a depression and you do not just leave with a prescription for a pill. You leave with a second prescription. On it: "30 minutes of brisk movement on 4 days per week. Mediterranean anti-inflammatory nutrition. 5 minutes of coherent breathing daily. 3 short cold showers per week. 30 minutes of daylight in the first hour of the day. Iron, vitamin D, omega-3 according to laboratory check. Sauna 2 times per week. A phone call with a person who is glad to hear your voice." With companionship, with diagnostics, with a real plan. That is a picture, not a treatment plan. What makes sense for one individual person, and what does not, can only be clarified by an individual assessment.
That is the direction I work in, in my practice in Berlin. And I think psychiatry as a whole could move further in this direction if it followed today's evidence. So far it does so only hesitantly. This article is my attempt to show why.
I am a doctor. I prescribe psychotropic medication where it is needed. I see at the same time, every day, what happens when the body is finally heard as a co-cause of the mind. We talk to patients without asking their mitochondria. We prescribe medications whose average effect over placebo is small in the large meta-analyses (you will see this in plain language below). That does not mean they do not work. For many people they are important and right. It only means there is room for more alongside them. It is time to be more honest.
This article is the most extensive evidence summary I have ever written here for you. Around 25 controlled trials and meta-analyses, sorted by the pillars of movement, cold, heat, sleep, nutrition, micronutrients, stress, relationships, toxins, light, microbiome. For some of these pillars the evidence is good, for others it is thin. I say each time where we stand. Read it as knowledge, not as instructions for self-treatment. No one should stop antidepressants on their own based on a blog article, including this one. What you need is a clinician who builds an individual plan with you.
For transparency: I run a private practice and offer medical consultation, extended diagnostics and accompaniment there myself, partly as a self-pay service. I have also written a book about cold. So what you are reading is not a neutral review, it is the reasoned position of a doctor with his own interests. Please read everything that follows with that in mind.
HAMD and MADRS are questionnaires to express depression severity in numbers. SMD (standardized mean difference): 0.2 small, 0.5 medium, 0.8 large. Odds Ratio: how much more likely something is in treatment versus control. NNT: number needed to treat for one extra to benefit. Remission: symptoms below diagnostic threshold. Response: at least 50 percent symptom reduction.
- Where we stand: an honest balance sheet of classical psychiatry
- An honest comparison table: effect sizes in plain language
- The PNI lens
- The inflammation hypothesis of depression
- Exercise: the best studied pillar
- Cold: much experience, little evidence
- Sauna: the Finnish lesson
- Sleep: the gateway to the mind
- Nutrition 1: the Mediterranean phenomenon
- Nutrition 2: what the keto pilot study hints at
- Micronutrients
- Stress, HRV and breathing
- Mindfulness, meditation, yoga
- Social bonds as medicine
- Hormones and the female mind
- Toxins as underestimated triggers
- Light as therapy
- The microbiome and the mind
- The anthroposophic view
- Why the guidelines lag behind
- Psychotherapy also needs an expansion
- How a holistic psychiatry could look
- Safety notes and limits
- Four honest levers for your self-check
A pattern I meet again and again in my consulting room looks roughly like this. Someone has a depressive episode, has tried several antidepressants one after another, and feels no clear effect. The question then hanging in the room is often: "What is wrong with me, that no pill works?"
My answer to that is not a diagnosis and not a promise. It is a second question: has anyone actually looked at thyroid, iron, vitamin D, blood sugar, inflammatory markers, sleep, movement and load? Frequently the answer is no. Whether anything useful follows from that in an individual case can only be shown by an individual assessment, never by an article.
I deliberately describe no concrete illness course here, no laboratory values and no before-and-after numbers. Such stories raise expectations that cannot be met in the individual case, and German medicines advertising law sets narrow limits on reproducing case histories in advertising aimed at the public. What I want to show is only this: it is worth looking at the whole person before the fourth substance is tried.
And just as clearly in the other direction: there are severe, biologically anchored depressions where lifestyle pillars alone are not enough and medication remains essential. No path in this article is a replacement for psychiatric treatment. Please do not stop anything on your own.
1. Where we stand: an honest balance sheet of classical psychiatry
Many patients with depression leave a practice with an antidepressant. Sometimes with a referral to a therapy slot they may have to wait longer for, depending on the region. Extended diagnostics asking about inflammatory markers, micronutrients, thyroid or cortisol is as a rule not part of the initial workup. That is down to guidelines, time budget and reimbursement, not to a lack of care on the part of the clinicians.
Kirsch and colleagues analysed all FDA-submitted data including unpublished trials. On average, antidepressants performed only 1.8 HAMD points better than placebo. Below the 3-point threshold of clinical relevance.
Kirsch I et al. PLoS Med. 2008;5(2):e45. DOI: 10.1371/journal.pmed.0050045
The largest network meta-analysis ever, 522 studies, 117,000 patients, 21 antidepressants. All 21 substances beat placebo, but the odds ratios ranged between 1.37 and 2.13. Moderate effect on average.
Cipriani A et al. Lancet. 2018;391(10128):1357 to 1366. DOI: 10.1016/S0140-6736(17)32802-7
Originally reported: 67 percent cumulative remission after four steps. A 2023 reanalysis of the patient-level data, using the original protocol, arrived at a cumulative remission of 35 percent instead of the reported 67 percent. This reanalysis is professionally contested and the original study authors have disputed it. What can be said: the question of how high the remission rate after four treatment steps really is stands more open today than it seemed for twenty years.
Pigott HE et al. BMJ Open. 2023;13(7):e063095. DOI: 10.1136/bmjopen-2022-063095
Psychotherapy versus pill placebo across 10 trials with 1240 patients. Effect size g=0.25, NNT 7.14. In points: 2.66 points lower on the Hamilton scale and 3.20 points lower on the Beck Depression Inventory. Real but small.
Cuijpers P et al. Psychol Med. 2014;44(4):685 to 695. DOI: 10.1017/S0033291713000457
If we are honest, the classical tools of psychiatry work, but moderately. They are no miracle. Saying that is not anti-psychiatry. It is taking the data seriously.
An important clarification before we go through the pillars
Before we go through the pillars, a sentence that gives many patients courage and that many doctors do not like to hear. A prescription is a professionally grounded recommendation within the constraints that apply. If an experienced psychiatrist prescribes you an antidepressant, he or she has good evidence-based reasons for it. Whether it is the right thing in your case, and what can sensibly stand alongside it, can only be clarified individually, not in an article. If you feel important questions have stayed open, raise that with your own clinician.
Prescription practice often reflects new evidence with delay. Licensing status, reimbursement rules, the time budget of an appointment and the availability of trials all play a part. New evidence takes time to reach guidelines and everyday practice. If you feel something is missing, it is legitimate to ask for a second opinion or for extended diagnostics.
2. An honest comparison table: effect sizes in plain language
Let us put the central numbers next to each other. This is strictly speaking a comparison of apples and pears: study designs, control groups and patient populations differ widely, and direct head-to-head trials between most of these approaches do not exist. SMD: 0.2 small, 0.5 medium, 0.8 large. Antidepressants in large meta-analyses sit in the range of 0.2 to 0.3 over placebo, that is small to small-medium. One point is decisive here and often overlooked: the antidepressant numbers come from placebo-controlled trials, while many lifestyle numbers come from comparisons with groups receiving no treatment. A comparison against doing nothing almost always comes out larger than a comparison against placebo. The numbers in this table therefore cannot be set off against each other.
Antidepressants (Cipriani 2018)
SMD ≈ 0.2 to 0.3OR for response 1.37 to 2.13. Small to small-medium.
Psychotherapy vs pill placebo (Cuijpers 2014)
g ≈ 0.25, NNT 7Seven people treated for one extra to benefit. Real, small-medium.
Exercise (Heissel 2023 BJSM)
SMD ≈ 0.95From 41 randomised trials, compared with non-active control groups. Restricted to the twelve trials at low risk of bias the value drops to SMD ≈ 0.67. Not a direct comparison with antidepressants, and a different control condition from those trials.
Exercise SMILE vs sertraline (Blumenthal 2007)
45% vs 47%Remission under supervised exercise practically equal to sertraline. Placebo 31 percent, and the group difference missed significance (p=0.057).
Mediterranean SMILES (Jacka 2017)
32% vs 8%Remission in diet group four times as high as in the active social control.
Ketogenic diet (Sethi 2024)
Pilot, n=23In schizophrenia and bipolar patients who all stayed on their medication. Single-arm pilot without a control group, focused on metabolic markers. No efficacy can be derived from this.
Bright light (JAMA Psychiatry 2024)
41% vs 24%Remission rates under add-on light therapy versus control in non-seasonal depression.
Cold exposure (van Tulleken 2018)
Case + mechanismA published single case report. Single cases do not demonstrate efficacy. A randomised feasibility trial (OUTSIDE) is running, results pending.
Anti-inflammatory (Köhler-Forsberg 2019)
SMD up to 0.64Anti-inflammatory substances added to ongoing therapy. Many of them are prescription-only and are not free of side effects.
Omega-3 EPA-dominant (Mocking 2016)
SMD ≈ 0.4In major depression, especially as add-on.
Vitamin D (Meta 2024)
SMD ≈ 0.32Per 1000 IU per day, dose-dependent. More pronounced with existing depressive symptoms, SMD ≈ 0.57.
HRV biofeedback (Pizzoli 2021)
g ≈ 0.38For depression and anxiety, comparable to classical relaxation training. Mostly small trials.
Mindfulness (Goyal 2014 JAMA)
SMD ≈ 0.30Similar order as antidepressants.
Social bonds (Holt-Lunstad 2010)
+50% survivalObservational data. Strong social ties were linked with higher survival. Association, not proof of causation.
Several lifestyle pillars reach effect sizes in the same range as antidepressants, a few above. Because the trials are built differently, this is not proof of superiority. It also does not show that every individual patient will respond like the average of a study. Effect sizes are group means. And one thing must be said clearly: these approaches are not risk-free. Intense cold exposure, sauna, a ketogenic diet, high-dose micronutrients, light therapy and therapeutic sleep deprivation all have their own side effects and contraindications. They differ from those of psychotropic medication, but they exist. Lifestyle medicine is a therapeutic level in its own right, and like every therapeutic level it belongs in medical hands.
3. The PNI lens
Clinical psychoneuroimmunology sees the human through four lenses:
- Nervous system. Vagus, sympathetic, HPA axis, pain and reward circuits.
- Immune system. Cytokines, microglia, T-cell balance, silent inflammation.
- Metabolism. Mitochondria, insulin, fat and amino acid metabolism, glucose.
- Hormonal system. Cortisol, thyroid, sex hormones, insulin as hormone.
These four lenses are interwoven. A disturbance in one can ripple through the others. A medication usually acts on one lens. Life plays on four.
Being mentally ill does not mean the root lies in the mind. It can equally lie in metabolism, immune system, or nervous system. Whoever looks only at the mind may be missing part of the picture.
4. The inflammation hypothesis of depression
In a subgroup of patients, depression could be connected to a chronic, often invisible inflammation reaching into the brain. Peripheral cytokines like IL-6, TNFα and IL-1β can act on the brain, activate microglia and lower neurotrophic factors like BDNF. In the kynurenine pathway, tryptophan can be diverted away from serotonin synthesis, with potentially neurotoxic metabolites.
The paradigmatic review. The depression subgroup with elevated CRP, IL-6 and TNFα is real and may have different therapeutic needs.
Miller AH, Raison CL. Nat Rev Immunol. 2016;16(1):22 to 34. DOI: 10.1038/nri.2015.5
36 RCTs of anti-inflammatory substances. Add-on: SMD minus 0.64. Monotherapy: SMD minus 0.41. Important context: some of these substances are prescription-only, the authors report a high risk of bias for all included trials, and a trend towards more infections was seen.
Köhler-Forsberg O et al. Acta Psychiatr Scand. 2019;139(5):404 to 419. DOI: 10.1111/acps.13016
Which tools can lower inflammation? Movement, sleep, Mediterranean diet, omega-3, vitamin D, stress regulation, microbiome care, cold exposure, sauna. Lifestyle medicine is not only about mood. It can also move something at the immunological level.
5. Exercise: the best studied pillar
41 randomised studies with 2264 participants, pooled effect on depression against non-active control groups (waiting list or no treatment): SMD minus 0.946. A large effect. Under supervised exercise: SMD minus 1.026. From this the authors calculate an NNT of 2, meaning roughly every second person treated may benefit additionally. Restricted to the twelve studies with a low risk of bias, the effect is considerably smaller at an SMD of minus 0.666, and the NNT there is 2.8. The more cautious figure is the more meaningful one. Important for interpretation: the comparison here was not against placebo but against groups receiving no treatment. Such comparisons almost always come out larger. The authors also point to small sample sizes and high heterogeneity. The whole analysis concerns adults aged 18 and over.
Heissel A et al. Br J Sports Med. 2023;57(16):1049 to 1057. DOI: 10.1136/bjsports-2022-106282
202 adults with major depression, four arms over 16 weeks. Remission: 45 percent under supervised exercise, 40 percent under home-based exercise, 47 percent under sertraline, 31 percent under placebo. Important context: the difference between groups narrowly missed statistical significance (p=0.057), and HAM-D scores in the active groups did not differ significantly from placebo. The authors describe exercise as broadly comparable to medication and point out the high placebo response.
Blumenthal JA et al. Psychosom Med. 2007;69(7):587 to 596. DOI: 10.1097/PSY.0b013e318148c19a
Exercise reaches an effect size of SMD minus 0.95 in the large meta-analysis, and minus 0.67 in the methodologically strongest subset. Antidepressants sit at around minus 0.3 in the large meta-analyses. These numbers must not simply be placed side by side. The exercise trials compare against groups receiving no treatment, the antidepressant trials against placebo. A comparison against doing nothing almost always comes out larger than a comparison against placebo. What the numbers show: movement is well studied and effective in depression. What they do not show: that it works better than a medication.
Exercise can raise BDNF, regulate the HPA axis, lower inflammation, raise endocannabinoid tone, improve mitochondrial function, raise insulin sensitivity, strengthen the microbiome and improve sleep architecture and vagal tone. It acts on several points at once. As orientation, the trials use 30 to 45 minutes moderate intensity, 3 to 5 times per week. If you have not trained for a long time or have cardiovascular disease, have the start checked medically first. The same applies in pregnancy and in eating disorders, where intensity and volume belong under medical supervision.
Exercise is already part of several depression guidelines. If nothing of it appears in your treatment plan, you are allowed to ask about it. Not as a replacement for what you are getting, but as a second pillar next to it.
6. Cold: much experience, little evidence
In 2008 Shevchuk proposed, in a pure hypothesis paper, that adapted cold showers might activate cold receptors, raise beta-endorphin and noradrenaline and act anti-inflammatory via the vagus nerve. That is a thought model, not a measurement. The paper appeared in a journal for hypotheses, without these effects having been tested in humans.
Shevchuk NA. Med Hypotheses. 2008;70(5):995 to 1001. DOI: 10.1016/j.mehy.2007.04.052
A published single case report about a young woman with long-standing, treatment-resistant depression who swam weekly in cold water under supervision. A decline in symptoms over time is described. A single case does not demonstrate efficacy, it only justifies a research question. And more importantly: stopping antidepressants always belongs in medical hands, never in self-management.
van Tulleken C et al. BMJ Case Rep. 2018. DOI: 10.1136/bcr-2018-225007
A randomised feasibility trial of outdoor swimming for mild to moderate depressive symptoms is running in England. It plans to recruit 88 participants for an eight-session outdoor swimming course added to usual care, compared with usual care alone. So far only the study protocol is published, results are not yet available. I mention it because the question is finally being investigated properly, not as evidence of an effect.
Massey H, Denton H, Burlingham A et al. OUTdoor Swimming as a nature-based Intervention for DEpression (OUTSIDE): study protocol for a feasibility randomised control trial. Pilot Feasibility Stud. 2023;9(1):122. DOI: 10.1186/s40814-023-01358-3
I wrote a book about cold as therapy because the topic has occupied me for years, first in myself and then in my consulting room. There are no robust trials on cold in depression so far, and I can therefore give you neither an effect nor a success rate. What I can say: the question is open, and it is finally being investigated properly.
Cold immersion is not for everyone. Cold water can trigger severe circulatory reactions, the so-called cold shock, and in open water there is an additional risk of drowning. With cardiovascular disease, cardiac arrhythmia, epilepsy, severe blood pressure dysregulation or in pregnancy this must be checked medically first. Never alone in open water, and always build up slowly.
A method without a large RCT is not automatically without effect. But neither is it demonstrated. One reason for the thin evidence can be that such approaches are hard to patent and therefore hard to fund. Until better trials exist, it stays an open question.
7. Sauna: the Finnish lesson
Finnish observational study in 2138 men aged 42 to 61, median follow-up about 25 years. 4 to 7 sauna sessions per week were linked with an around 78 percent lower risk of a psychotic disorder compared with one session per week. That number is less precise than it sounds: only 184 of the 2138 men were in the 4 to 7 group, and the confidence interval ran from 0.09 to 0.58. Further analyses of the same Finnish male cohort found comparable associations for cardiovascular mortality, hypertension and dementia. Context matters here: these are observational data from a male-only cohort. People who use the sauna often tend to be healthier and more active in general. Causation is not shown.
Laukkanen T, Laukkanen JA, Kunutsor SK. Sauna Bathing and Risk of Psychotic Disorders: A Prospective Cohort Study. Med Princ Pract. 2018;27(6):562 to 569. DOI: 10.1159/000493392
Discussed mechanisms: heat can activate heat-shock proteins, lower systemic inflammation, improve endothelial function, stimulate the vagus and support deep sleep. Sauna is not for everyone either. With unstable cardiovascular disease, recent myocardial infarction, severe blood pressure dysregulation, a tendency to fainting or in pregnancy this must be checked medically first. Alcohol and sauna never belong together, the combination markedly raises the risk of circulatory collapse and sudden death. And again on the evidence: the cited cohort covers middle-aged men only. It does not transfer to women without further study.
8. Sleep: the gateway to the mind
20 randomised trials with 1162 participants. CBT-I is regarded as first-line therapy for chronic insomnia. Sleep onset shortened by about 19 minutes, wake-after-sleep-onset by about 26 minutes. Effects appeared to hold at later time points, though the estimates there were less certain. No adverse outcomes were reported.
Trauer JM et al. Ann Intern Med. 2015;163(3):191 to 204. DOI: 10.7326/M14-2841
Systematic review and meta-analysis of total sleep deprivation in acute bipolar depression. Added to medication, total sleep deprivation can lower depressive symptoms more than medication alone within one week, SMD minus 0.58. The effect was maintained over three months when medication was continued. Tolerability and the risk of switching into mania did not differ in this analysis. A very important note on this: therapeutic sleep deprivation is a psychiatric procedure with real risks, among them a switch into mania and a lowered seizure threshold. It belongs in psychiatric care, usually inpatient, and it is not a self-experiment.
Ramirez-Mahaluf JP, Rozas-Serri E, Ivanovic-Zuvic F, Risco L, Vöhringer PA. Effectiveness of Sleep Deprivation in Treating Acute Bipolar Depression as Augmentation Strategy: A Systematic Review and Meta-Analysis. Front Psychiatry. 2020;11:70. DOI: 10.3389/fpsyt.2020.00070
Sleep is not a pause, it is active work. In animal models a flushing system of the brain has been described, the glymphatic system, which could transport metabolic waste during sleep. In humans this is not yet settled and remains professionally contested. What is well studied is that sleep consolidates memory and that cortisol profile and inflammatory markers can change with sleep. Practically: cortisol day profile, 30 minutes daylight in the first hour, no blue light in the last hour, cool dark room, magnesium evening, stable sleep rhythm.
9. Nutrition 1: the Mediterranean phenomenon
67 adults with moderate to severe depression, most of them in parallel psychotherapy or on medication. The dietary counselling therefore came in addition to ongoing treatment, compared against social support of equal duration. After 12 weeks: 32.3 percent remission in the diet group, 8 percent in the control group, NNT 4.1. Behind those percentages stand 10 versus 2 people, so the trial is small and only single-blind.
Jacka FN et al. BMC Med. 2017;15:23. DOI: 10.1186/s12916-017-0791-y
Adherence to healthy dietary indices correlates with reduced depression risk. Ultra-processed food intake: HR 1.22 to 1.32 for depression.
Lassale C et al. Mol Psychiatry. 2019;24(7):965 to 986. DOI: 10.1038/s41380-018-0237-8
10. Nutrition 2: what the keto pilot study hints at
23 adults with schizophrenia or bipolar disorder and existing metabolic abnormalities, all of whom stayed on their psychiatric medication. 4 months of ketogenic diet, added to ongoing treatment. What is reported is mainly metabolic: among adherent participants about 12 percent less weight, 27 percent lower HOMA-IR, 25 percent lower triglycerides. On the psychiatric side, BPRS scores fell by 32 percent in the participants with schizophrenia, average illness severity on the CGI by 31 percent, and 79 percent of those with elevated symptoms improved by at least one CGI point. This is a single-arm pilot study in 23 people, without a control group and without blinding. Numbers like these show feasibility and justify further research. They do not demonstrate efficacy.
Sethi S et al. Psychiatry Res. 2024;335:115866. DOI: 10.1016/j.psychres.2024.115866
That a dietary change produces any signal at all in severe psychiatric illness is a reason to look more closely. So far it is no more than that. A single-arm pilot study in 23 people can raise a question. It cannot answer it.
The discussed mechanism is that a ketogenic diet shifts brain metabolism towards ketone bodies, which could activate mitochondria, lower inflammation, raise GABA tone, reduce oxidative stress and dampen glutamate excitotoxicity. Important: a ketogenic diet is a medical dietary form with side effects and contraindications, among others in certain metabolic disorders, in pregnancy and under several medications. It belongs under medical supervision, and it is never a replacement for antipsychotic or mood-stabilising treatment. Replication as a larger RCT is running.
There are indications that mental illness could also carry a metabolic component in some people. That is a research direction, not established knowledge. It is good enough, though, to look at metabolism once in the individual person.
11. Micronutrients: vitamin D, omega-3, magnesium
Dose-response meta-analysis of 31 RCTs with 24,189 participants. Per 1000 IU per day, vitamin D can slightly reduce depression scores, SMD minus 0.32. In people who already had depressive symptoms the effect was more pronounced, SMD minus 0.57. Arithmetically the largest reduction appeared at 8000 IU per day. That number has to be read with great caution: it comes from very few trials and its confidence interval is correspondingly wide (SMD minus 2.04, confidence interval minus 3.77 to minus 0.31). Equally important: in this analysis the effect was mainly short-term. At follow-up beyond 52 weeks it was no longer detectable (SMD plus 0.14). No significant effect on anxiety symptoms. A very important note: 8000 IU per day is well above the tolerable upper intake level of 4000 IU for adults stated by EFSA. Such a dose belongs exclusively in medically supervised, laboratory-monitored treatment. Chronically excessive intake can cause hypercalcaemia and kidney damage.
Ghaemi S, Zeraattalab-Motlagh S, Jayedi A, Shab-Bidar S. The effect of vitamin D supplementation on depression: a systematic review and dose-response meta-analysis of randomized controlled trials. Psychol Med. 2024;54(15):3999 to 4008. DOI: 10.1017/S0033291724001697
13 RCTs of omega-3 PUFA. SMD 0.398 in favour of omega-3. EPA-dominant formulations performed clearly better. The doses tested in the trials were mostly 1 to 2 grams of EPA per day. Higher omega-3 doses can increase bleeding tendency and need checking if you take anticoagulants.
Mocking RJ et al. Transl Psychiatry. 2016;6(3):e756. DOI: 10.1038/tp.2016.29
Magnesium is often short in heavily processed diets. Individual smaller randomised trials hint at antidepressant and sleep-improving effects, above all in documented deficiency. The evidence is thin and large trials are missing. Important: with impaired kidney function magnesium does not belong to be taken without medical assessment, and it can interfere with the absorption of antibiotics, bisphosphonates and thyroid hormone. Take it at a different time of day in that case.
12. Stress, HRV and breathing
24 studies, 484 participants. HRV biofeedback can reduce stress and anxiety symptoms, with an effect size of Hedges g=0.83 against control conditions. The authors stress themselves that these are mostly small trials and that better controlled studies are needed.
Goessl VC et al. Psychol Med. 2017;47(15):2578 to 2586. DOI: 10.1017/S0033291717001003
14 randomised trials, 794 participants. Mean effect size g=0.38 on depressive symptoms. That is the same range as the antidepressant effects reported in meta-analyses. It is not a direct within-trial comparison.
Pizzoli SFM et al. Sci Rep. 2021;11:6650. DOI: 10.1038/s41598-021-86149-7
Exercise, mindfulness and HRV biofeedback equally effective for stress reduction.
van der Zwan JE et al. Appl Psychophysiol Biofeedback. 2015;40(4):257 to 268. DOI: 10.1007/s10484-015-9293-x
Coherent breathing (four seconds in, six seconds out, five minutes daily) costs nothing. It can favourably influence heart rate variability. The trials cited above, however, examine device-supported biofeedback, not the free breathing exercise. For the free exercise alone the evidence is thin.
13. Mindfulness, meditation, yoga
47 randomised trials, 3515 participants. Mindfulness-based programmes: anxiety SMD 0.38, depression SMD 0.30 at 8 weeks. The authors found no evidence that meditation works better than another active treatment.
Goyal M et al. JAMA Intern Med. 2014;174(3):357 to 368. DOI: 10.1001/jamainternmed.2013.13018
12 randomised trials with 619 participants. Against usual care the short-term effect size was SMD minus 0.69, against relaxation minus 0.62, against aerobic exercise minus 0.59. Yoga can be an adjunctive option. Only three trials had a low risk of bias, and none reported safety data.
Cramer H et al. Depress Anxiety. 2013;30(11):1068 to 1083. DOI: 10.1002/da.22166
14. Social bonds as medicine
148 studies, 308,849 participants. Strong social ties were linked with an around 50 percent higher likelihood of survival. Social isolation: mortality risk plus 26 to 32 percent. The authors rate this magnitude as comparable to well-established risk factors for mortality. These are observational data. They show associations, not causes.
Holt-Lunstad J et al. PLoS Med. 2010;7(7):e1000316. DOI: 10.1371/journal.pmed.1000316. Plus Perspect Psychol Sci. 2015;10(2):227 to 237. DOI: 10.1177/1745691614568352
In the largest analysis of this question, the link between social integration and survival was of the same magnitude as well-established risk factors for mortality. Relationships are not comfort therapy. They belong on the list.
15. Hormones and the female mind
The hormonal dimension of female mental health deserves particular mention. In my experience it comes up less often in the initial psychiatric workup than the topic warrants. PMDD, postpartum depression and perimenopause each produce their own symptom pictures. In PMDD an altered sensitivity to progesterone metabolites such as allopregnanolone is discussed. Whether individualised hormonal accompaniment can contribute anything in perimenopause depends on the individual case and has to be weighed against risks, among them thrombosis and breast cancer risk depending on preparation, dose and history.
On conditions around pregnancy, childbirth and the postnatal period I am not permitted to make public statements. German medicines advertising law (Heilmittelwerbegesetz) restricts this. If this concerns you, please turn to your gynaecologist, your midwife, your general practitioner or a psychiatric outpatient service. In an acute crisis in Germany: emergency number 112, Telefonseelsorge helpline 0800 111 0 111.
If your psychiatric symptoms are cyclic, hormonally shifted or clearly linked to life-phase transitions, the right question is not "which antidepressant", but "which hormonal constellation produces this and what does it specifically need".
16. Toxins as underestimated triggers
Cumulative metal mixture exposure was linked with higher depression scores in US adults in NHANES data. Oxidative stress is the mechanism most discussed. These are cross-sectional and observational data. They show an association, not a cause.
Shin J, Luo Y. Heliyon. 2024;10(23):e40221. DOI: 10.1016/j.heliyon.2024.e40221 · Li Q, Liu X et al. J Affect Disord. 2025;393(Pt A):120290. DOI: 10.1016/j.jad.2025.120290
Toxins are not the trigger for most depressions. In a smaller group they could play a part. If you are chronically exhausted, several systems are affected and several therapies have not carried you, it can make sense to raise the topic of environmental exposure in a medical consultation. Whether such diagnostics can contribute anything in your case is an individual decision. Trials demonstrating a benefit of these tests in depression do not exist so far.
If three antidepressants in a row have produced no clear effect, it can be worth asking not only about the fourth substance, but also whether something in the body is contributing that has not been examined yet. Both are possible. Both deserve to be discussed.
17. Light as therapy
11 randomised trials, 858 patients. Light therapy was given in addition to ongoing treatment. Remission 40.7 percent versus 23.5 percent in controls, odds ratio 2.42. Response 60.4 versus 38.6 percent, odds ratio 2.34. The effect also appeared outside seasonal depression.
JAMA Psychiatry. 2024. DOI: 10.1001/jamapsychiatry.2024.2871
Practically: 30 minutes at 10,000 lux within the first hour of waking. Light therapy is well studied and simple to apply, but it is not harmless. In bipolar disorder bright light can favour a switch into mania, and with certain eye conditions or photosensitising medication it is unsuitable. Check with your doctor first.
18. The microbiome and the mind
This analysis included 23 randomised trials with 1401 clinically diagnosed patients, 20 of which provided data for the meta-analysis. In the 18 trials on depression a reduction in symptom scores was seen under probiotics, SMD minus 0.96, and in the 9 trials on anxiety a moderate effect, SMD minus 0.59. Prebiotics alone showed no significant effect. Heterogeneity between studies was high. What this means for an individual person cannot be derived from it. I am describing an evidence base here, not recommending a product. And a safety note: with immunosuppression, a central venous catheter or severe underlying illness probiotics are not harmless, cases of bacteraemia and fungaemia have been described. Have it checked medically first in that situation.
Asad A, Kirk M, Zhu S, Dong X, Gao M. Effects of Prebiotics and Probiotics on Symptoms of Depression and Anxiety in Clinically Diagnosed Samples: Systematic Review and Meta-analysis of Randomized Controlled Trials. Nutr Rev. 2025;83(7):e1504 to e1520. DOI: 10.1093/nutrit/nuae177
The microbiome is not a powder. It is a lifestyle: Mediterranean diet, fibre, fermented foods, fewer unnecessary antibiotics, sleep, exercise, stress regulation.
19. The anthroposophic view: the liver thinks too
Rudolf Steiner spoke in the 1920s about how a poorly working liver connects to a "defect of the will": wanting to do something without that impulse crossing into action. Antiquity already knew. Melancholia means "black bile". When the liver becomes sluggish and bile flow stagnates, the mood symbolically tips.
You can place this old image next to a present-day physiological description: disturbed phase 1 and phase 2 detoxification, insulin resistance, fatty liver, hormonal imbalance, altered BDNF expression. Those are axes that also appear in the inflammation hypothesis. That is not proof that Steiner was right. It is a parallel I find interesting.
Anthroposophic medicine sees the human as a threefold of nerve-sense system, rhythmic system and metabolic-limb system. It recognises the liver as co-bearer of the will. A century later, connections can be drawn to the language of mitochondria, inflammation and the HPA axis. That is an interpretation, not a scientific proof. For anthroposophic remedies there is no robust evidence base in this field. They can at most be an accompanying level, discussed case by case.
20. Why the guidelines lag behind
1. Research is pharma-centric
A large share of clinical drug trials is funded by manufacturers. Lifestyle interventions are hard to patent, so large trials are correspondingly harder to finance.
2. Medical training barely teaches it
Nutrition, exercise and sleep medicine take up only a small part of the curriculum in many medical schools. That is changing, slowly.
3. Guidelines follow RCTs, not effects
522 large RCTs for antidepressants. Hundreds for lifestyle, but smaller per pillar.
4. Lifestyle is hard to standardise
20 mg of citalopram is clear. "Mediterranean diet" is a spectrum.
5. Patient must be active
Lifestyle costs energy. In acute depression exactly this energy is missing.
6. Insurance reimburses pills, not lifestyle accompaniment
A pill prescription is reimbursed. A 90-minute PNI history typically not.
7. Academic psychiatry is trapped in the symptom model
DSM and ICD are symptom lists. They do not separate, for example, depression with inflammation from depression with a trauma background.
This is not a conspiracy theory but a well-known topic in health-services research. It may help explain why a well-studied building block like exercise is still rarely prescribed in a structured way in routine care.
21. Psychotherapy also needs an expansion
Classical psychotherapy works above all through thoughts, feelings and relationships. That is effective and important. If part of the depressive complaints can also relate to inflammation, metabolism, hormones, sleep and movement, then it can help when a psychotherapy thinks the body along with it. Approaches like Somatic Experiencing, EMDR, EAET, Pain Reprocessing Therapy, polyvagal-informed therapy and mindfulness-based CBT do exactly that.
A psychotherapy that convinces me sees the human as embodied, networked, hormonal, immunological, social. It talks. It also breathes. It looks at vagal tone. That is not a criticism of the established methods, it is an addition alongside them.
22. How a holistic psychiatry could look
1. Clean diagnostics before any therapy
An extended laboratory panel can make sense, depending on the picture: thyroid, vitamin D, iron, B12, folate, hsCRP, insulin and HOMA-IR, liver and lipids. Which parameters can contribute anything in an individual case is decided in the consultation. For single procedures, such as microbiome or heavy metal analyses, a benefit in depression is not demonstrated. They therefore only come into consideration after individual weighing and with a clear question.
2. Lifestyle pillars with demonstrated effect sizes
Movement 30 to 45 min, 3 to 5x/week. Mediterranean. CBT-I. Light therapy. HRV biofeedback and coherent breathing. For cold and heat applications a robust evidence base in depression is still missing, they remain an individual question.
3. Micronutrients and phytotherapy
Micronutrients guided by laboratory results and only where a need is documented, at a medically defined dose and with follow-up monitoring. Concrete target values and dosages are individual and do not belong in a blog article. High-dose supplements taken on your own can cause harm.
4. Psychotherapy trauma-sensitive and PNI-informed
EMDR, Somatic Experiencing, integrative, body-informed.
5. Pharmacotherapy when right, not because fastest
Antidepressants, mood stabilisers, antipsychotics have a firm place in severe courses. The problem is using them as first and only line.
23. Important safety notes and limits
Acute suicidality. Call the emergency number immediately, in Germany and across the EU that is 112. In Germany the Telefonseelsorge helpline is 0800 111 0 111. Lifestyle medicine has no place as first line here.
Severe bipolar, schizophrenia, severe personality disorders. Psychiatric care with medication is often life-saving. Lifestyle pillars are complementary, never replacement.
Tapering antidepressants. Self-tapering can cause rebound, discontinuation syndromes, relapses. Stepwise under medical accompaniment, ideally hyperbolic.
Therapeutic sleep deprivation, intense cold and heat applications. These are procedures with real risks, among them a switch into mania, circulatory reactions, cardiac arrhythmia and, in open water, drowning. They belong under medical supervision and not in a self-experiment.
Children and adolescents. All the trials named in this article were conducted in adults, the exercise meta-analysis explicitly from age 18. None of it transfers to children and adolescents. In minors every decision belongs in child and adolescent psychiatric hands.
Self-experimentation. Combining high-dose vitamin D, ketogenic diet, intense ice bathing and phytotherapeutics without knowing interactions risks harm. The tools are powerful. They need a competent hand.
The aim of this article is not "do all this yourself". It is: "know that these tools exist, raise them, ask your doctor about extended diagnostics and a lifestyle plan, and if you do not get further together, you are allowed to ask for a second opinion".
24. Four honest levers for your self-check
Before you read on, one sentence that matters more than everything that follows. If you are doing very badly right now, if you are thinking about suicide or feel you cannot bear it any longer, then these levers are not your next step. Then it is: emergency number 112, or in Germany the Telefonseelsorge helpline 0800 111 0 111, around the clock and free of charge. Please get help, now and not later.
First lever. Move. Today. 30 minutes, fast enough that you could no longer comfortably converse. Of all the lifestyle pillars this is the best studied one. If you have cardiovascular disease or have not moved much for a long time, check the start with your doctor first.
Second lever. Eat a meal your grandmother would recognise. Fresh ingredients, plenty of vegetables, good fats, some protein, little sugar.
Third lever. Breathe coherently five minutes a day (four seconds in, six seconds out). It can favourably influence your heart rate variability. For the free breathing exercise the evidence is thinner than for device-supported biofeedback.
Fourth lever. Call a person who is sincerely glad to hear your voice. In large observational studies, supportive relationships are linked with a longer life. Those are associations, not proven causes.
The mind is not a place in the head. It is the result of a body, a metabolism, a nervous system, a story and a relationship to other people.
Closing word
We live in a medicine that has specialised ever finer. What lies between specialisations is often seen by no one. That is exactly where much of what we today call "common condition depression" lives.
A holistic psychiatry is not the opposite of biological psychiatry. It can be a sensible extension of it. It uses pills when necessary. It does ask that before and alongside any medication the pillars are checked for which we now have good data.
There is dignity in understanding that your mind lives in the whole body. You did not "become depressive". Something in your system speaks. When you see the whole system, you have options that no pill alone can ever contain.
If you want to work on your own picture, with diagnostics that sees all four PNI lenses, with lifestyle pillars that fit your constellation, with anthroposophic and phytotherapeutic accompaniment where it makes sense, you will find the option to book an appointment below this article.
Sources
- Kirsch I et al. PLoS Med. 2008;5(2):e45. DOI: 10.1371/journal.pmed.0050045
- Cipriani A et al. Lancet. 2018;391(10128):1357 to 1366. DOI: 10.1016/S0140-6736(17)32802-7
- Pigott HE et al. STAR*D reanalysis. BMJ Open. 2023;13(7):e063095. DOI: 10.1136/bmjopen-2022-063095
- Cuijpers P et al. Psychol Med. 2014;44(4):685 to 695. DOI: 10.1017/S0033291713000457
- Miller AH, Raison CL. Nat Rev Immunol. 2016;16(1):22 to 34. DOI: 10.1038/nri.2015.5
- Köhler-Forsberg O et al. Acta Psychiatr Scand. 2019;139(5):404 to 419. DOI: 10.1111/acps.13016
- Heissel A et al. Br J Sports Med. 2023;57(16):1049 to 1057. DOI: 10.1136/bjsports-2022-106282
- Blumenthal JA et al. Psychosom Med. 2007;69(7):587 to 596. DOI: 10.1097/PSY.0b013e318148c19a
- Shevchuk NA. Med Hypotheses. 2008;70(5):995 to 1001. DOI: 10.1016/j.mehy.2007.04.052
- van Tulleken C et al. BMJ Case Rep. 2018. DOI: 10.1136/bcr-2018-225007
- Kelly JS et al. Lifestyle Med. 2022. DOI: 10.1002/lim2.53
- Laukkanen T, Laukkanen JA, Kunutsor SK. Sauna Bathing and Risk of Psychotic Disorders: A Prospective Cohort Study. Med Princ Pract. 2018;27(6):562 to 569. DOI: 10.1159/000493392
- Trauer JM et al. Ann Intern Med. 2015;163(3):191 to 204. DOI: 10.7326/M14-2841
- Ramirez-Mahaluf JP, Rozas-Serri E, Ivanovic-Zuvic F, Risco L, Vöhringer PA. Effectiveness of Sleep Deprivation in Treating Acute Bipolar Depression as Augmentation Strategy: A Systematic Review and Meta-Analysis. Front Psychiatry. 2020;11:70. DOI: 10.3389/fpsyt.2020.00070
- Jacka FN et al. SMILES trial. BMC Med. 2017;15:23. DOI: 10.1186/s12916-017-0791-y
- Lassale C et al. Mol Psychiatry. 2019;24(7):965 to 986. DOI: 10.1038/s41380-018-0237-8
- Sethi S et al. Psychiatry Res. 2024;335:115866. DOI: 10.1016/j.psychres.2024.115866
- Ghaemi S, Zeraattalab-Motlagh S, Jayedi A, Shab-Bidar S. The effect of vitamin D supplementation on depression: a systematic review and dose-response meta-analysis of randomized controlled trials. Psychol Med. 2024;54(15):3999 to 4008. DOI: 10.1017/S0033291724001697
- Mocking RJT et al. Transl Psychiatry. 2016;6(3):e756. DOI: 10.1038/tp.2016.29
- Goessl VC, Curtiss JE, Hofmann SG. The effect of heart rate variability biofeedback training on stress and anxiety: a meta-analysis. Psychol Med. 2017;47(15):2578 to 2586. DOI: 10.1017/S0033291717001003
- Pizzoli SFM et al. Sci Rep. 2021;11:6650. DOI: 10.1038/s41598-021-86149-7
- van der Zwan JE et al. Appl Psychophysiol Biofeedback. 2015;40(4):257 to 268. DOI: 10.1007/s10484-015-9293-x
- Goyal M et al. JAMA Intern Med. 2014;174(3):357 to 368. DOI: 10.1001/jamainternmed.2013.13018
- Cramer H, Lauche R, Langhorst J, Dobos G. Yoga for depression: a systematic review and meta-analysis. Depress Anxiety. 2013;30(11):1068 to 1083. DOI: 10.1002/da.22166
- Holt-Lunstad J et al. PLoS Med. 2010;7(7):e1000316. DOI: 10.1371/journal.pmed.1000316
- Holt-Lunstad J et al. Perspect Psychol Sci. 2015;10(2):227 to 237. DOI: 10.1177/1745691614568352
- Shin J, Luo Y. Exposure to multiple heavy metals and depressive symptoms in the US (NHANES). Heliyon. 2024;10(23):e40221. DOI: 10.1016/j.heliyon.2024.e40221
- Li Q, Liu X et al. Heavy metals co-exposures and depression (NHANES). J Affect Disord. 2025;393(Pt A):120290. DOI: 10.1016/j.jad.2025.120290
- Menegaz de Almeida A, Aquino de Moraes FC, Cavalcanti Souza ME et al. Bright Light Therapy for Nonseasonal Depressive Disorders: A Systematic Review and Meta-Analysis. JAMA Psychiatry. 2025;82(1):38 to 46. DOI: 10.1001/jamapsychiatry.2024.2871
- Asad A, Kirk M, Zhu S, Dong X, Gao M. Effects of Prebiotics and Probiotics on Symptoms of Depression and Anxiety in Clinically Diagnosed Samples: Systematic Review and Meta-analysis of Randomized Controlled Trials. Nutr Rev. 2025;83(7):e1504 to e1520. DOI: 10.1093/nutrit/nuae177
- Massey H, Denton H, Burlingham A et al. OUTdoor Swimming as a nature-based Intervention for DEpression (OUTSIDE): study protocol for a feasibility randomised control trial. Pilot Feasibility Stud. 2023;9(1):122. DOI: 10.1186/s40814-023-01358-3
- Guu TW, Mischoulon D, Sarris J et al. ISNPR practice guidelines for omega-3 fatty acids in major depressive disorder. Psychother Psychosom. 2019;88(5):263 to 273. DOI: 10.1159/000502652
- Marx W, Manger SH, Blencowe M et al. Clinical guidelines for the use of lifestyle-based mental health care in major depressive disorder: WFSBP and ASLM taskforce. World J Biol Psychiatry. 2023;24(5):333 to 386. DOI: 10.1080/15622975.2022.2112074
- Antonelli M, Barbieri G, Donelli D. Effects of forest bathing (shinrin-yoku) on levels of cortisol as a stress biomarker: a systematic review and meta-analysis. Int J Biometeorol. 2019;63(8):1117 to 1134. DOI: 10.1007/s00484-019-01717-x
- Steiner R. Spiritual Science and Medicine (GA 312, 1920).
- Bartelme RR. Anthroposophic Medicine. Glob Adv Health Med. 2020;9. DOI: 10.1177/2164956120973634