Depression Recovery: What Clinical Guidelines Actually Say About Treatment, Lifestyle, and Timeline

Major depressive disorder reaches around 8.3% of US adults in any given year and roughly 21% across a lifetime, with women diagnosed at nearly twice the rate of men [Source: https://www.nimh.nih.gov/health/statistics/major-depression]. WHO data places depression as a leading global cause of disability — currently ~280 million people worldwide.

The treatment evidence base is large and unusually well-studied. What it tells us is not that “depression is treatable” in some vague sense — it tells us that which treatment, for whom, and at what severity matters substantially. This page is the framework.

Stepped Care: What NHS Got Right

The single best piece of patient-centered design in depression treatment is the stepped-care model that NHS has run for two decades and that NICE guidelines codify. The principle: match treatment intensity to severity, escalate when needed, de-escalate when stable [Source: https://www.nhs.uk/mental-health/conditions/clinical-depression/treatment/].

Mild depression: watchful waiting, self-help (specifically guided CBT workbooks like Sleepio for sleep or Devon Mind’s resources for mood), exercise on prescription, peer support. Antidepressants are not recommended first-line for mild presentations because the risk-benefit is unfavorable — Kirsch and colleagues’ 2008 meta-analysis of FDA-registered antidepressant trials showed drug-placebo difference falling below NICE clinical-significance threshold for mild-to-moderate baselines [Source: https://pubmed.ncbi.nlm.nih.gov/18303940/]. This single paper changed prescribing guidance globally.

Mild-to-moderate: low-intensity psychological interventions. Computerised CBT (cCBT), guided self-help, group CBT, or behavioral activation. Most can be NHS-prescribed in the UK and have effect sizes comparable to face-to-face therapy when delivered well.

Moderate-to-severe: high-intensity CBT or interpersonal therapy (IPT) — OR an antidepressant — OR, the option that produces the largest effect sizes in trials, the combination of both. Cuijpers and colleagues’ 2023 network meta-analysis of 331 psychotherapy trials (n>34,000) found CBT, IPT, behavioral activation, and problem-solving therapy all outperform care-as-usual; differences between active modalities were small and not clinically meaningful for most comparisons [Source: https://pubmed.ncbi.nlm.nih.gov/36697826/]. The choice of bona fide evidence-based therapy matters less than access to any of them.

Severe / treatment-resistant: psychiatric specialist referral. Augmentation strategies (lithium, second-generation antipsychotic, T3), MAOI trial, and ECT for life-threatening cases (catatonia, sustained suicidality). These tools are not first-line but they’re not exotic either; they exist for the patients other treatments fail.

A note on suicide-risk assessment: every clinical guideline says it should happen at every encounter. If you’re in crisis, please call 988 (US) or 116 123 (UK Samaritans).

The Antidepressant Landscape, Honestly

Cipriani and colleagues (2018) published the cleanest comparative analysis of antidepressants we have. Their network meta-analysis included 522 trials, 116,477 participants, 21 drugs, and presented response rates and dropout rates for each, head-to-head [Source: https://pubmed.ncbi.nlm.nih.gov/29477251/].

The most efficacious group for response: amitriptyline, mirtazapine, duloxetine, venlafaxine, paroxetine. The best-tolerated group (lowest dropout): agomelatine, citalopram, escitalopram, fluoxetine, sertraline, vortioxetine. Sertraline and escitalopram appeared in both lists — efficacious and well-tolerated. This is exactly why they end up first-line in nearly every clinical guideline written this century.

Three things SSRIs do not do, despite popular belief:

1. They don’t make you happy. They put a floor under how low you can go. That floor is what lets the rest of recovery work — therapy homework, basic hygiene, calling friends back, cooking — happen at all.
2. They don’t work in 24 hours. Onset is typically 4–6 weeks for full effect. The first 2–4 weeks frequently feel worse: jaw clenching, GI symptoms, vivid dreams, libido drop. Hang on through week 6 unless your GP advises otherwise.
3. They don’t get tapered with a single dose drop. Tapering protocol: 25% reduction every 2–4 weeks, under GP supervision. Brain zaps from abrupt discontinuation are common, especially with paroxetine and venlafaxine.

Effect sizes are honestly modest. The number-needed-to-treat for response over placebo is around 9 — meaning ~9 patients on antidepressants for one to benefit who would not have benefited from placebo. This is similar to or larger than most other primary-care medications, but the framing matters: antidepressants are useful but they are not magic, and their best use is as part of a treatment plan that also includes therapy and lifestyle.

Recovery Is Non-Linear

This is the part that nobody says out loud and that everyone has to learn. Recovery from depression is not a straight line and the longitudinal data is unsentimental about this.

Average untreated first-episode duration: ~6–8 months. People who experience one episode have a ~50% chance of a second within 5 years. Treatment doesn’t change the underlying biological vulnerability, but it shortens episode duration, reduces severity, and improves the trajectory of subsequent episodes when they occur.

Behavioral activation deserves its own paragraph. The premise — that you should engage in scheduled enjoyable activities even when you don’t want to and your mood will follow the behavior, not the reverse — sounds maddeningly close to toxic positivity at first. The actual evidence-based protocol is different from “just be positive.” It’s structured: schedule specific activities at specific times, treat them like medical appointments, do them whether or not you feel like it, and over weeks-to-months your mood tracks the behavior rather than vice versa [Source: https://pubmed.ncbi.nlm.nih.gov/32977351/]. Effect size g=0.74 vs control. BA is especially useful when CBT therapists aren’t available — it can be delivered by less specialized practitioners, including peer counselors with shorter training.

Exercise on prescription is now standard NHS care for mild-moderate depression. Schuch and colleagues (2016) pooled 25 RCTs and found large effect on depressive symptoms (SMD -1.11), still moderate-to-large after adjusting for publication bias (-0.66). Doses that produced benefit in trials: 30–60 minutes, 3–5 times/week, moderate intensity, 8–12 weeks duration [Source: https://pubmed.ncbi.nlm.nih.gov/27253219/].

Mood tracking can help, but in a specific way. Pure logging of daily scores is data collection, not treatment. Tracking becomes therapeutic when it surfaces patterns and connects to behavioral or cognitive responses — which is the framework the mood-tracking pillar covers in detail.

A year-out perspective from people who’ve been through severe episodes: recovery is not a return to who you were before. It’s a renegotiation with yourself about what flares it up and what doesn’t. You learn that you can’t run on four hours of sleep anymore; you notice when you haven’t seen friends in two weeks and act on it instead of waiting to feel like it; you check in with the GP every six months even when you feel fine. The episode leaves a kind of weather radar in your head, and that’s the unmentioned silver lining — you become substantially harder to surprise the next time.

---

For specific topics — antidepressant comparison, behavioral activation worksheets, exercise protocols — see the long-tail articles indexed from this site’s homepage. None of this is medical advice. If you’re in crisis, call 988 (US) or 116 123 (UK Samaritans).