Antidepressants (SSRIs & SNRIs)

This inference is the most common error in the entire debate, and it is a logical mistake rather than an empirical one. A drug can work without the illness being a deficiency of what the drug acts on: aspirin relieves headaches, and headaches are not caused by an aspirin deficiency. Whether antidepressants beat placebo is a separate question, tested separately, in hundreds of randomised trials — and the answer is that they do.

Yes. The largest network meta-analysis of the field — over 500 randomised trials of 21 antidepressants — found all of them more effective than placebo for acute major depression. The finding is robust. The argument in this field is about the SIZE of the effect, not its existence.

On average, no — it is modest. Roughly one additional person responds for every six to eight treated compared with placebo, and the average symptom-score improvement over placebo is small-to-moderate. This is a real effect at a scale worth having, and it is not the transformation the advertising implied.

The evidence points this way consistently. Benefit over placebo grows with baseline severity, and in mild depression the drug-placebo difference is small enough that psychotherapy and exercise are reasonable first choices. This is also why blanket statements — 'they work' or 'they don't' — are both wrong: it depends who is taking them.

Yes, and this is underappreciated. SSRIs and SNRIs are first-line pharmacological treatment for generalised anxiety disorder, panic disorder, and social anxiety disorder, with good randomised evidence. The name of the drug class undersells what it treats.

Not in the ordinary sense: they do not produce craving, euphoria, compulsive use, or dose escalation to chase an effect. But this reassurance was used for years to wave away a real problem — physical dependence and withdrawal — and that was a mistake. Not addictive, and not easy to stop, are both true at once.

Yes, and the profession got this badly wrong for a long time. Discontinuation symptoms — dizziness, 'brain zaps', insomnia, irritability, flu-like symptoms, anxiety — are common, and in a minority they are severe or protracted for months. Guidance has been revised to acknowledge this. Estimates of frequency and severity remain contested; the existence of the problem is not. Slow, gradual tapering with a clinician is the answer, and it can take far longer than the two weeks people are often given.

Very. Reduced libido, delayed or absent orgasm, and erectile difficulty affect a large share of users — plausibly the majority on some agents — and they are the single most under-discussed harm at the point of prescribing. People deserve to be told before they start, not to discover it and assume it is the depression. Emotional blunting is similarly common and similarly under-warned.

Regulators carry a warning about increased suicidal thoughts and behaviours in people under 25, particularly in the first weeks. It is a real signal and it warrants close early monitoring. It must be read against the fact that untreated depression is itself a leading driver of suicide, and that the warning's introduction was followed by concerns about under-treatment. This is a reason for careful supervision — not a reason to leave a depressed young person untreated.

Not on your own, and not abruptly. Continuing for a period after recovery reduces relapse, and stopping suddenly is the most reliable way to produce withdrawal symptoms — which are then frequently misread, by patients and clinicians alike, as the depression returning. If you want to come off, that is a legitimate goal: do it slowly, with your prescriber.

For moderate-to-severe depression the combination outperforms either component on its own in randomised comparisons. It is also the option least often offered, because it is the most expensive and the hardest to arrange.