About 4,000 adults in Pomerania, in northeastern Germany, were followed for more than two decades. Low income and heavy drinking each came with the same 31% higher risk of death. In the poorest group, the two risks may have added up.

I've written before about the alcohol-harm paradox: poorer communities often drink the same as richer ones, or less, yet they take more of the damage. Most of that research looks at deaths caused directly by alcohol. This study asks a broader question. Does the same pattern hold for death from any cause?

What they did

The data comes from SHIP-START (Study of Health in Pomerania). Between 1997 and 2001, 4,307 adults went through a baseline survey and medical exam. Researchers then tracked deaths through January 2023.

They measured socioeconomic status two ways, by income and by education, and analyzed each one separately. That left two final samples: 3,771 people with income data and 3,939 with education data. During follow-up, 981 people in the income sample died. In the education sample it was 1,047.

At baseline, they also recorded:

  • how much each person drank, including former drinkers who had stopped
  • blood biomarkers linked to long-term heavy drinking
  • age, sex, marital status, smoking, and BMI (body mass index), which they adjusted for

They used two kinds of survival models. One gives relative risk, the familiar "X% higher" figure. The other gives absolute added risk. You need both when the question is how two factors combine.

What they found

After adjusting for age, smoking, and the rest:

  • Low income: 31% higher risk of death compared with high income (HR, or hazard ratio, 1.31).
  • Heavy drinking: also 31% higher, compared with light drinking (HR 1.31).
  • Former drinkers: people who used to drink and had stopped had the highest risk of any group, 56% higher (HR 1.56).
  • Biomarkers: higher levels of the drinking-related blood markers went with higher mortality.

The gap between rich and poor shrank once smoking, BMI, and drinking were accounted for. It didn't close. Having less money stayed linked to dying earlier, beyond every risk factor they measured.

Then they looked at combinations. There was some evidence that the risks interact. Low-income former drinkers, and low-income people who drank moderately or heavily, showed more risk than income or alcohol alone would predict. The authors are careful here, and I'll be careful too. Split 4,000 people by income and then by drinking level, and some groups get small. These interaction estimates are shaky.

What it means

The matching numbers are what stayed with me. In this cohort, being in the low-income group carried about the same mortality risk as heavy drinking. I've seen the study summarized as "social status matters more than alcohol." The data doesn't say that. It says they're roughly equal, and that they may make each other worse.

I don't read the former-drinker result as "stopping is dangerous." There's a known pattern in this research called the "sick quitter": people often stop drinking because they're already ill. Drinking was measured once, at baseline, so the study can't tell who quit for health reasons and who quit for other ones. I wrote about how the same bias inflates claims of low-dose health benefits. When the nondrinker group includes sick ex-drinkers, moderate drinkers look healthier than they are.

A few things limit how far I'd take this:

  • It's one region in Germany, measured around the year 2000.
  • Income and drinking were recorded once and could have changed over 25 years.
  • It's observational. It shows associations, not cause and effect.

Still, the core finding fits a lot of earlier work. The same drink seems to weigh differently depending on what else is going on in your life. Money buys things that soak up risk: better food, safer housing, less chronic stress, an easier path to a doctor. The study didn't measure any of those, so this is my guess at the mechanism. The authors only say the gap held after their adjustments.

On the drinking side, "high intake" in a study like this is a long-term average. But averages build up one evening at a time, and a lot of the drift happens on nights when the pace picks up without you noticing. That part you can actually see in real time. AlcoBalance shows where a night is heading while it's still going, so you can catch the speed-up and ease off on purpose. You're still steering the evening. A weekly total is mostly made of moments like that.

Income is harder to change from inside a single evening. That's what makes this study uncomfortable. It puts a factor you can adjust tonight next to one you mostly can't, and finds them about equal.

Source: Alcohol, Clinical and Experimental Research, DOI