Does drinking to relax make you more anxious?
An alcohol researcher of forty-five years explains the push-pull: the calm is real and immediate, and the same drink is winding the stress system tighter underneath it. The bill arrives tomorrow.
What this episode covers
- Alcohol genuinely does calm you down.
- It also activates the brain's stress systems at the same time, so tomorrow you start from a worse place, and over years you can end up drinking to quiet a system the drinking wound up.
- That mechanism is the best answer on this site to why 'a drink to take the edge off' stops working.
- The rest is practical: what actually helps when you try stopping, from two addiction psychiatrists who do this for a living.
Confidence in this episode
Everything about how much to believe this episode, in one place.
The most heavily credentialed expert on any alcohol episode here — forty-five years at UNC, the Bowles Center for Alcohol Studies, principal investigator on the pivotal naltrexone trial — and he behaves like it. Asked for breast cancer figures he does not have, he says the risk is real, small at one drink a day, and that he cannot give the number. Asked about the host's face-photo metric, he suggests a randomised comparison. He also declines the no-safe-level framing others take, giving a genuinely mixed picture instead. Confidence is Moderate rather than High because the episode's own measurements are uncontrolled before-and-afters, and because the behavioural half, while sensible, is clinical experience rather than trial data.
- Alcohol produces acute anxiolysis while activating brain stress systems, so stress reactivity increases the following day. This is well established and is the strongest content here.
- Alcohol is sedating but disrupts sleep architecture later in the night — more waking, less deep sleep, less total sleep time.
- Acetaldehyde is the first metabolite, is toxic, has carcinogenic properties, and drives inflammatory and immune activation.
- Blood pressure often normalises in heavy drinkers who cut down or stop — his own trial observation, and he names its limit unprompted.
- Alcohol raises breast cancer risk in a dose-dependent way. The 7% per drink figure the host cites from the 53-study pooled analysis matches the literature.
- Cue conditioning is real: the time, place and people you habitually drink with trigger anticipatory dopamine before any decision is made.
- That the before-and-after face photographs show anything measurable. Uncontrolled, unblinded, and no comparison group — as Garbutt himself gently points out.
- That the challenge's lab markers reflect the alcohol specifically. Participants change several things at once, and in earlier years some markers did not move because people swapped alcohol for sugar.
- That 40 days is a meaningful threshold. It is the length of Lent, which is the actual reason, and nobody claims otherwise.
- That behavioural strategies transfer from addiction treatment to social drinkers. Plausible and sensible; the psychiatrists say themselves this is not their patient population.
- That 'the average number of tries to quit smoking is seven' applies here. A real figure from a different behaviour, offered as encouragement rather than evidence.
Why it matters
Almost everyone who drinks to unwind believes it works, and in the moment they are right. This episode explains why they are also wrong, and it does it without telling anyone to quit. James Garbutt spent forty-five years at UNC studying exactly this, and his account is the mechanism most alcohol coverage skips: the anti-stress effect and the stress-system activation happen together, not in sequence. So the relief is real, and the cost is real, and they arrive at different times. Do it long enough and a stressor that should read as a three reads as an eight. What makes the episode unusually useful is that it is the only one on this site where someone says out loud: if you try this and it is hard, that is information, and here is where to get help. Two addiction psychiatrists spend the second half on what actually works — one day at a time rather than forty, planning for the hours you always drink, replacing the behaviour rather than just removing it. The catch is that they work for the health system that runs the challenge and the rehab programme they refer you to. That is context, not a reason to discount them.
What stands out
- The calm and the cost are the same event, not a sequence. Alcohol relaxes you while it activates the stress systems that make tomorrow worse. Nobody connects the drink on Tuesday to the short fuse on Wednesday, because the relief arrives on time and the bill arrives a day late.
- The most qualified person here refuses to say alcohol has no safe level. Garbutt says small amounts may lower heart attack and ischaemic stroke risk while raising blood pressure and breast, liver and other cancer risk, and that people should weigh it themselves. Other experts on this site reach a harder conclusion from the same evidence.
- He is asked for the breast cancer numbers and says he doesn't have them. 'The risk that increase is small and I don't have the precise percentage.' In a genre built on confident figures, a specialist declining to produce one is the most reliable moment in the episode.
Best-supported action
The single highest-leverage move from this episode, anchored in the strongest evidence the speaker presents.
If you drink to take the edge off, notice how you feel the next afternoon rather than the same evening. The relief is real and immediate; the cost lands roughly a day later, which is exactly why almost nobody connects the two. That gap is the whole mechanism, and you can test it on yourself this week without giving anything up.
Where to start
Small low-friction starters covering the main moves from this episode.
- Try one day rather than forty. Both psychiatrists say the same thing, and it is the opposite of how these challenges are marketed: the number is the obstacle. 'I am not drinking today' is a decision a person can actually make, and the days accumulate on their own.
- Name your witching hour and plan for it in advance. The transition home from work, or walking into a bar — your brain starts releasing dopamine in anticipation before you have decided anything. A plan you made when calm beats willpower you do not have at 6pm.
- Replace the behaviour, do not just subtract it. Their own challenge learned this the hard way: participants swapped alcohol for ice cream and their markers stayed flat. Decide what you are drinking instead before you are standing at the bar.
- Say it plainly and blame someone else if it helps. 'I'm not drinking today,' with eye contact, beats 'I don't think I want to.' Their observation that most people do not notice or care, and that whoever pushes hardest may have their own reason, costs nothing to test.
Other supported actions
Further actions discussed in this episode, ordered from strongest to weakest evidence. This is one expert's view, the full topic compares and ranks across experts.
- Test the mechanism on yourself before deciding anything. Drink as usual, then rate your stress the following afternoon rather than the same evening. Garbutt's whole point is that the relief and the cost are separated by about a day, which is why almost nobody links them. If the pattern is there, you will see it in a fortnight without giving anything up.Strong evidence
- If you drink daily or heavily, talk to a doctor before you stop. The episode routes you to treatment if the challenge is hard, which is better than most, but it never says that withdrawal from daily heavy drinking can cause seizures and is occasionally fatal. It is manageable with support.Strong evidence
- Aim at today, not at the number. Both addiction psychiatrists independently make the same point and it is the opposite of how challenges are sold: forty days is the obstacle. Decide about today, plan for the hour you usually drink, and pick what is in the glass instead before you are standing there.Moderate evidence
Full context, impact ratings, and timing — available in related topics
Questions to take to your doctor
- Given how much I drink, is it safe for me to stop on my own, or do I need supervision?
- My blood pressure is high and I drink most days. Is it worth seeing what happens to it if I stop for a while before we talk about medication?
- I drink to manage anxiety. Is that making it worse, and is there something better?
- Are there medications for alcohol that would be worth considering for someone like me?
Full doctor prep with ranked questions available in the full topic page
Context
The most heavily credentialed voice on any alcohol episode in this library, and the one who is most careful with what he does not know. Asked directly for the breast cancer statistics, he says the risk is real, that it is small at one drink a day, and that he does not have the precise figure — rather than producing one. Asked about the host s before-and-after face photographs, he says she is ahead of the medical community and then, gently, that the clinical trial methodologist in him would want people randomly assigned with a comparison group. He is also the only speaker here who declines the no-safe-level framing: he says small amounts may reduce heart attack and ischaemic stroke risk while raising blood pressure, breast cancer and other cancer risk, and that people should weigh it themselves. That is a more careful reading of the evidence than the harder line taken elsewhere. And he volunteers the limit of his own best finding: blood pressure often normalises when heavy drinkers stop, but essential hypertension will not, and he does not want to give that impression.
One thing to be clear about, because the episode is not: if you drink heavily every day, stopping abruptly can cause seizures and delirium and is occasionally fatal. Alcohol is one of the few drugs whose withdrawal kills. Nobody here names that danger. What this episode does do — alone among the alcohol pages on this site — is route you somewhere: two addiction psychiatrists say plainly that if forty days proves very hard, that is worth paying attention to, and tell you where their programme is. That is better than silence and it is not the same as the warning. If stopping feels frightening, that is a medical conversation first.
The core neurobiology comes from the independent expert. Garbutt's push-pull account of stress — acute relief and simultaneous activation of stress systems, with the reckoning arriving the next day and compounding over years — is established neurobiology and the best mechanistic explanation on this site of why drinking to cope stops working. His sleep account is standard and correct. His acetaldehyde chemistry is textbook. None of it requires trusting him, which is the point.
He is also the most careful person here, and the moments where he holds back are worth more than the moments where he explains. Asked for breast cancer statistics, he says the risk is real, small at one drink a day, and that he does not have the precise figure — so the host looks it up and cites the 53-study pooled analysis at roughly 7% per drink, which is accurate. Asked about blood pressure, he offers a genuine clinical observation, that heavy drinkers who stop often normalise and some come off their medication, and then immediately fences it: essential hypertension will not respond, and he does not want to give that impression. And he declines the no-safe-level framing entirely, describing a mixed picture where small amounts may lower heart attack and ischaemic stroke risk while raising blood pressure and several cancers. That is a more conservative reading than David Nutt or Jeremy London reach on this site from the same literature, and it is not obviously wrong.
The episode's own evidence is weaker than the science it reports, and it never quite says so. The challenge measures inflammatory and liver markers, blood pressure, body composition and close-up photographs of participants' faces, before and after. That is an uncontrolled before-and-after in people who have volunteered, know what is being tested, and are changing several things at once. Garbutt handles this more gracefully than most would — he calls the face photographs interesting and says she is ahead of the medical community, then notes that the clinical trial methodologist in him would want people randomly assigned with a comparison group. That is a polite way of saying the photographs are not a finding, and it should be read as one.
The behavioural half is sensible, experienced and untested at this dose. One day at a time, HALT, planning for cues, replacing rather than subtracting — this is standard addiction practice and there is no reason to think it is harmful for a social drinker. But both psychiatrists say explicitly that people doing a forty-day challenge are not their patients, and strategies developed for alcohol use disorder are being extrapolated, in good faith, to a different population. Their candour about that boundary is a mark in their favour.
The episode is an in-house production and it does not announce itself as one. The host directs Lifestyle Nutrition at Ochsner Health; the challenge is run by her Ochsner Eat Fit team; the two in-person guests are Ochsner clinicians; and the close refers listeners to Ochsner's rehabilitation programme. The independent expert is the one calling in from UNC. This does not invalidate Garbutt's neurobiology or the pooled breast cancer analysis, and the two are worth judging separately — Eat Fit is a nonprofit initiative, and nothing here is being sold to a listener at the point of listening. It is still worth knowing that the challenge, the host, two of the three guests and the treatment programme all belong to a single organisation, and that the one voice outside it is the one that keeps declining to overstate.
Overall evidence profile: excellent, well-sourced neurobiology from an independent specialist who is careful about his limits, wrapped around a challenge whose own measurements cannot support conclusions and a set of behavioural strategies borrowed in good faith from a more severe population. What survives is the mechanism and the tactics, both of which cost nothing. The reason to try this is that you will find out what alcohol was doing to your baseline. The reason the challenge lasts forty days is Lent, not evidence for a biologically meaningful threshold.
Where people go wrong
- Reading the before-and-after photographs as evidence.They are the most persuasive thing in the episode and the least reliable: no control group, no blinding, and everyone knows what is being tested. Garbutt says as much, politely, by suggesting a randomised comparison. The underlying changes may well be real — but if the photo is what convinced you, your reason for stopping will not survive the first person who shows you a set that did not change.
- Subtracting the alcohol and changing nothing else.Their own challenge ran into this: participants' markers stayed flat because the wine had been replaced by ice cream. If the drink was doing a job — winding down, marking the end of the day, being social — the job does not disappear when the drink does. Something fills it, and if you do not choose what, it chooses itself.
What to expect over time
- The first few daysOften worse before better, and this is where people quit. Sleep can be harder without the sedative, and Garbutt is explicit that the improvement takes time — a week or more — and warns against the obvious trade of alcohol for a sleeping pill. The host's own experience ran the other way: she noticed the anxiety change within a week, before any lab could.
- One to two weeksWhere the stress systems start to settle and the thing people report is usually not physical. Less irritable, less reactive, arguments that do not happen. Garbutt notes it often gets pointed out by someone else first — a spouse noticing before you do.
- Around forty daysThe challenge's endpoint, and its honest origin is Lent rather than biology. This is when their measurements happen, and when the failure mode shows up: markers that did not move because sugar replaced the alcohol. That finding is worth more than the ones that did move.