If you are in crisis, stop reading and get help now. Samaritans, free from any phone in Ireland and the UK: 116 123, 24 hours, every day. At immediate risk: A&E, or 999 or 112. Thoughts of ending your life are a medical emergency and they are treatable — tell someone tonight, not eventually.
Depression carries a risk of suicide. That is the most important fact here, and it is also routinely misquoted. The 15% lifetime figure still repeated in textbooks came from a 1970 analysis of severely ill inpatients. A reanalysis found a hierarchy instead — 8.6% among those ever hospitalised for suicidality, 4.0% hospitalised without it, 2.2% across mixed inpatient and outpatient populations. Lower than the number you have heard, far above the general population, and, as those authors noted, no risk factor reliably predicts suicide in an individual. Which is why the response is asking, not calculating.
What follows is ranked by strength of evidence, not by what we sell. Supplements come last, with the reasons stated. We did the same for anxiety, and the ranking was no more flattering there.
The DSM-5 threshold has three parts. Five or more symptoms from a list of nine, present most of the day nearly every day for at least two weeks, causing real impairment. Plus a gateway: at least one must be either persistently low mood or loss of interest and pleasure. That second one is anhedonia, the underrated symptom — plenty of people with major depression do not describe themselves as sad. They describe everything as flat.
The rest of the list: sleep, appetite, energy, concentration, physical slowing or agitation, worthlessness or guilt, and thoughts of death or suicide.
The two-week floor does real work. A grim fortnight after a bereavement is not automatically an illness — grief has its own shape and course. What separates the disorder is duration and pervasiveness: low mood that lifts when something good happens behaves differently from low mood that never lifts. Flatness and shutdown in autistic adults, meanwhile, is often autistic burnout, which needs a different response.
Nine questions, scored 0 to 3, total 0 to 27 — one item per DSM criterion. The original validation in 6,000 primary care and obstetrics patients set the bands still in use: 5, 10, 15 and 20 for mild, moderate, moderately severe and severe. Against a clinician interview in 580 of them, 10 or above had 88% sensitivity and 88% specificity.
A 2019 individual-participant meta-analysis of 58 studies and 17,357 participants confirmed 10 as the best cut-off: sensitivity 0.88 (95% CI 0.83 to 0.92), specificity 0.85 (0.82 to 0.88) against semi-structured interviews. One caveat from the same paper: accuracy depends on the reference interview, with sensitivity 5 to 22% lower against fully structured lay-administered ones.
So 10 or above is a threshold for a proper conversation, not a diagnosis — a meaningful minority who screen positive do not have major depression. The questionnaire's real jobs are opening that conversation and tracking whether treatment works.
Item 9 asks about thoughts of being better off dead or of hurting yourself. It exists so that risk gets asked about rather than hoped about, and it is the most under-reported item. If you marked it above zero, say so out loud to whoever gave you the form.
Several treatable conditions look exactly like depression. An underactive thyroid produces fatigue, low mood and slowed thinking, and is regularly managed as depression first. B12 deficiency does much the same, though the evidence on correcting it is mixed. Iron deficiency causes profound fatigue before anaemia shows on a blood count, which is why ferritin rather than haemoglobin is the number to ask for. Untreated sleep apnoea produces exhaustion and poor concentration, and the STOP-BANG questions take two minutes. The short blood panel covers most of it. Alcohol and some prescribed medications belong on the list too.
For depression that is not severe, NICE is explicit: do not routinely offer antidepressant medication as first-line treatment. Its menu opens with guided self-help, group and individual CBT, and behavioural activation; SSRIs sit ninth.
Behavioural activation is worth naming because it targets depression's central trap. Depression removes motivation, doing less removes reward, less reward deepens it. It breaks the loop from the wrong end — you schedule the activity first and let motivation follow, rather than waiting for what the illness is suppressing.
For more severe depression the ordering changes: the first-listed option is a combination of individual CBT and an antidepressant, not either alone. That is the honest answer to anyone wanting this framed as therapy versus drugs. The real barrier is access — self-referral to NHS Talking Therapies is possible without a GP, and guided digital programmes can run while you wait.
The largest network meta-analysis pooled 522 trials and 116,477 participants across 21 antidepressants. All 21 beat placebo, with odds ratios from 2.13 (95% credible interval 1.89 to 2.41) for amitriptyline down to 1.37 for reboxetine. SSRIs and SNRIs are used first, with tricyclics and mirtazapine where those do not suit.
Three findings from that analysis get dropped from summaries. Differences between drugs were small with wide credible intervals on most head-to-head comparisons, so league tables are less useful than they look. Only agomelatine and fluoxetine had fewer dropouts than placebo; clomipramine had more. And the trials were short-term, which is not how most people take these drugs.
They take weeks. NICE puts it plainly: if an antidepressant is going to work it usually does so within four weeks, with response assessed at about four weeks for medication and four to six for therapy. Stopping at day ten is the commonest avoidable failure.
The first one often is not the right one, and that is normal. STAR*D followed 3,671 outpatients through up to four successive steps and found remission rates of 36.8%, then 30.6%, then 13.7% and 13.0%, with a cumulative 67%. Read that both ways: about a third get there on the first drug, most who persist get there eventually, and those needing more steps relapsed more often afterwards. Do not stop abruptly — coming off is something to plan with a prescriber.
A 2024 BMJ network meta-analysis covered 218 studies and 14,170 participants with clinically significant depression. Against active controls it found moderate reductions for walking or jogging (Hedges' g −0.62, 95% credible interval −0.80 to −0.45), yoga (−0.55, −0.73 to −0.36), strength training (−0.49, −0.69 to −0.29), mixed aerobic exercise (−0.43) and tai chi or qigong (−0.42). Effects scaled with intensity; yoga and strength training were best tolerated.
The caveat the headlines dropped is large. Of those 218 studies, exactly one met the Cochrane criteria for low risk of bias, and the authors graded their own confidence as low for walking or jogging and very low for everything else — you cannot blind someone to whether they are exercising. The direction is convincing; the magnitude should be held loosely. It still outranks supplements, and group exercise appears on both NICE menus. We have gone through these trials in more detail separately.
Some ingredients have real data for depressive symptoms. Omega-3 has the largest literature, and the picture is mixed once you separate EPA-heavy from DHA-heavy formulations. Saffron has head-to-head trials against antidepressants, all small and mostly from one research community. Zinc has adjunctive evidence thinner than the enthusiasm. St John's Wort has real efficacy data and an interaction profile serious enough that we stay cautious.
Three things hold across all of them: trial doses usually exceed what any general-purpose product contains, the trials are small and short, and none of it makes a supplement a treatment for major depressive disorder.
No Proco product is a treatment for depression and none is formulated as one. Cognitive Support contains ten actives, of which only niacin at 8 mg NE and vitamin B6 at 5 mg are individually disclosed — the other eight sit inside a 255 mg proprietary blend whose amounts our manufacturer will not let us publish. That limitation alone tells you where this belongs. The longer answer on supplements and mood is a heavily qualified yes.
If you take an antidepressant, check before adding anything. St John's Wort and 5-HTP both interact with serotonergic medication and both are in our Sleep Support formula; we keep a full list of those interactions. Raise it with a pharmacist rather than working it out alone.
Treatment-resistant depression has its own pathway, and reaching it is not the end of the road. The regulatory definition is inadequate response to at least two antidepressants, each at an adequate dose for an adequate time, actually taken. At least 30% of people with depression qualify.
The key caveat there: a significant proportion of apparent resistance is pseudo-resistance — trials too short or too low-dose, or medication not taken as prescribed. Worth checking before escalating.
Beyond it, several options have established efficacy — intravenous ketamine and intranasal esketamine alongside an antidepressant, repetitive transcranial magnetic stimulation, electroconvulsive therapy, and certain second-generation antipsychotics added in partial responders. All are specialist referrals. The point of naming them is this: "the tablets did not work" is a reason to go back, not a verdict.
This article is not medical advice and has not been evaluated by the FDA. Nothing here is intended to diagnose, treat, cure or prevent any disease, or to start, stop or change a prescribed medication. If you are having thoughts of suicide, contact Samaritans on 116 123 or your local emergency services now.
Sources: Kroenke et al. J Gen Intern Med. 2001 · Levis et al. BMJ. 2019;365:l1476 · NICE NG222, 2022 · Cipriani et al. Lancet. 2018;391:1357-66 · Rush et al. Am J Psychiatry. 2006 · Noetel et al. BMJ. 2024;384:e075847 · McIntyre et al. World Psychiatry. 2023;22:394-412 · Bostwick & Pankratz. Am J Psychiatry. 2000;157:1925-32