Evidence, with its limits showing.
We would rather under-promise. Here is what the research actually says.
Start your visitAntidepressants work, modestly and unevenly
The largest analysis to date pooled 522 randomised trials and 116,477 participants. All 21 antidepressants outperformed placebo for acute depression. Effects were moderate on average, and individual response varies widely (Cipriani et al., The Lancet, 2018).
The first medication is often not the last
In STAR*D, the largest real-world depression trial, roughly one in three participants reached remission on their first medication. Remission rates rose with subsequent steps, though later re-analyses suggest the cumulative figure is lower than first reported. The practical lesson stands: follow-up and adjustment matter.
Therapy is not second best
For anxiety and depression, cognitive behavioural therapy performs comparably to medication for many people, and the combination often outperforms either alone. For insomnia, CBT-I is the recommended first-line treatment, ahead of any medication.
Quitting smoking
In EAGLES (8,144 participants), continuous abstinence at weeks 9 to 24 was 21.8% with varenicline, 16.2% with bupropion, 15.7% with nicotine patch and 9.4% with placebo. No significant increase in serious neuropsychiatric events was found (Anthenelli et al., The Lancet, 2016).
Drinking less
In COMBINE (1,383 participants), naltrexone with medical management reduced the risk of a heavy drinking day (hazard ratio 0.72) compared with placebo (Anton et al., JAMA, 2006).
Off-label use
Some uses described on this site are off-label, meaning the medication is FDA-approved for a different condition: propranolol for performance anxiety, low-dose trazodone and hydroxyzine for sleep, bupropion for ADHD. Off-label prescribing is legal and common, and we label it every time.
Boxed warning
Antidepressant-class medications carry an FDA boxed warning about increased suicidal thoughts and behaviour in people under 25. Monitoring in the early weeks is essential.
All figures are published trial results for the medications. They are not okay. patient outcomes.
It is okay to ask. Start there.
A few questions, then a licensed physician takes it from here.
Start your visit
