Sobriety Isn’t the Hard Part (It’s Everything After)

People think getting sober is the hard part. And yeah, detox sucks. Early sobriety is uncomfortable. Breaking the habit is genuinely difficult. But that’s not what pulls people back to the bottle or the pills six months in.

What pulls them back is not knowing what to do with themselves once they’re sober. The substance was doing a job. Multiple jobs, actually. Managing anxiety. Filling evenings. Lubricating social situations. Hobby, coping skill, off-switch for whatever the day put in your head. You take it away and all of that is still sitting there with no replacement in sight.

Detox you can do in a week. Withdrawal you can do in two. Building a life that doesn’t require numbing out to tolerate takes years, and the first one is the worst.

Why month six is worse than month one

Acute withdrawal ends. Post-acute withdrawal does not, at least not on the timeline anybody wants. PAWS is the thing where your brain chemistry is still recalibrating months after your last drink, and the symptoms look like personality flaws if you don’t know what they are. Sleep that won’t stay together. Mood that swings on nothing. Cravings on a Tuesday afternoon when you were fine an hour ago. Cognitive fog. Anhedonia, which is the clinical word for “nothing feels good anymore.” That one is the killer.

For alcohol, PAWS can run six months to two years. For opioids, similar. For benzodiazepines it can be longer, and the protracted anxiety component is genuinely awful. Most people who relapse around month six are relapsing into PAWS symptoms they didn’t know were normal. They think they’re broken. Their brain is just slow to come back online.

I had a guy come in last spring, eight months sober from alcohol, convinced he was bipolar. Mood swings, irritability, crashing afternoons. He’d done AA, white-knuckled through every birthday and wedding, and was sitting in my office thinking he had a separate psychiatric illness on top of everything else. He didn’t. He had PAWS. We got him on a low-dose SSRI, fixed his sleep, and the “bipolar” went away over the next four months. The brain takes its time.

What the substance was actually doing

Almost nobody drinks just to drink. The substance was medicating something, and when you pull it out, whatever it was medicating comes back with interest.

In clinic, the most common things underneath are anxiety, depression, untreated ADHD, and trauma, usually in some combination. Alcohol is a fantastic short-term anxiolytic. It works on GABA the same way benzos do, just sloppier. If you’ve spent fifteen years using two drinks at 6 PM to come down from work, your nervous system has built its whole afternoon around that. Take it away and the anxiety that was there at 5:55 every day is still there, except now it’s at 5:55, 6:30, 8:00, midnight, and 3 AM.

Opioids are an emotional anesthetic. They blunt physical pain and psychological pain, and people don’t always distinguish. Plenty of people on opioids are medicating depression or trauma and don’t know it because the opioid was working. Stimulants get used to self-treat undiagnosed ADHD constantly, which is why some people in recovery from cocaine or meth respond well to actual Vyvanse or Adderall once the dust settles. Careful conversation, that one.

Stopping is the first step. Building something that doesn’t require numbing out to tolerate is the rest of the staircase.

Sobriety surfaces the original problem. Most people in their first year sober are dealing with anxiety or depression that’s been there their whole adult life, just chemically muffled. That’s the project of recovery. You can’t fix what the drinking was hiding until you stop drinking, and you can’t stay stopped until you fix what the drinking was hiding.

MAT exists and it works

Medication-assisted treatment still gets weirdly moralized, mostly by people who’ve never had to live inside the brain it’s treating. Naltrexone, acamprosate, buprenorphine. These work. The data isn’t subtle.

Naltrexone is underused for alcohol. Daily oral or the monthly Vivitrol shot. It blocks the opioid receptors that mediate the reward from alcohol, so if you drink on it, you feel the drink physically but you don’t get the emotional payoff. A surprising number of people on naltrexone just lose interest over a few months because the brain stops associating alcohol with reward. Some people don’t respond. But it’s prescribed at maybe a tenth of the rate it should be.

Acamprosate (Campral) is the other one. Three times a day dosing, which is annoying, but it directly addresses the glutamate dysregulation that drives PAWS. If you’ve got a patient who’s six months sober and miserable, acamprosate is often what gets them through the next six.

Buprenorphine for opioids isn’t optional in 2026. The data on overdose mortality is overwhelming. Anybody telling someone in opioid recovery that they need to taper off Suboxone to be “really sober” is, statistically, telling them to die. Long-term maintenance saves lives. The brain on opioids doesn’t fully heal in six months or a year, and plenty of people do best on indefinite buprenorphine the same way diabetics do best on indefinite insulin. The moralism around this kills people.

The social network problem

Everyone you used to drink with still drinks. Every place you used to go, drinking happens there. The wedding. The work happy hour. The Sunday football thing at your buddy’s house. The dinner at your in-laws where wine is just on the table like furniture. Your entire adult social infrastructure was built around something you’re not doing anymore.

For the first year, that infrastructure is a relapse risk in human form. Your friends aren’t bad. The routes of association are just wired into your brain. You walk into Steve’s living room and your nervous system says “this is the place where the IPA happens.” Normal cued-recall, nothing more dramatic than that.

This is the actual argument for AA and the twelve-step rooms, separate from any theological piece. The rooms give you a parallel social network where the default activity isn’t drinking. Coffee at 7 PM with people who get it. A reason to be somewhere on a Friday night that isn’t a bar. For a lot of people that’s the whole thing right there. For others the higher-power language is a dealbreaker, and SMART Recovery and Refuge Recovery exist for exactly that reason. No rule says you have to do AA. There’s a strong rule that says you have to do something. Isolated sobriety has worse one-year outcomes than almost any other version.

The relapse curve, honestly

Relapse rates for substance use disorders run between 40 and 60 percent in the first year. The dangerous windows are around the 90-day mark, six months, and twelve months, and what they share is the emotional gap. Three months in, the “look at me I quit” energy fades and life is just life again. Six months in, PAWS is peaking. Twelve months in, the brain has decided you’re recovered, which is exactly when overconfidence sets up the slip.

A woman I worked with last year, two years sober from alcohol, picked up again at a work conference because she figured she could handle one glass with dinner. She drank for three months before she got back into treatment. The brain’s memory for the substance doesn’t go away. It just goes quiet.

So if you’re in year one, the relevant questions are: what are you doing instead of drinking on Friday at 7 PM, what’s your plan for the next wedding, who do you call when the craving lands on a Wednesday for no reason, and what underlying thing are you treating that the substance was treating before. If you can’t answer those, the math isn’t on your side. If you can, it bends substantially in your favor. Sobriety is the entry fee. Most of the work is on the other side of the door.