Alcohol’s Impact on IHD Mortality Hinges on Socioeconomic Status

Moderate drinkers with low SES—unlike those with high SES—didn’t see a survival benefit, with binge drinking harmful for all.

Alcohol’s Impact on IHD Mortality Hinges on Socioeconomic Status

People who consume no more than 20 g of alcohol per day—the equivalent of around 1.5 standard drinks—are less likely to die from ischemic heart disease (IHD) than those who abstain over a lifetime, but only if they have high socioeconomic status (SES).

Researchers say their results, based on self-reported data from more than half a million US adults over a two-decade period, suggest that drinking alcohol may be “less safe or not safe” for individuals with lower education, which they used as a proxy for SES.

“Alcohol use has long been identified as an important health behavior that contributes to socioeconomic inequalities in mortality,” they point out in their paper, published earlier this week in JAMA Network Open.

Health problems due to drinking, like liver cirrhosis and alcohol-attributable mortality, are more common in people with low versus high SES. “It is well known that there’s this alcohol paradox that says people with low SES experience more alcohol-related health conditions, even though they drink less than people with high SES,” lead author Yachen Zhu, PhD (Public Health Institute, Emeryville, CA), told TCTMD. But less is known, she said, about how this relates specifically to cardiovascular disease.

At moderate levels, alcohol is associated with less IHD—with a U-shaped curve showing higher risks among people who drink in excess or have abstained entirely throughout their lives. The question, which hadn’t yet been looked at in the US population, is how SES might influence that relationship.

Commenting for TCTMD, Kenneth J. Mukamal, MD (Beth Israel Deaconess Medical Center, Boston, MA), said while this study isn’t the first to look at the link between socioeconomics and alcohol-related outcomes, it’s built on a solid foundation: the National Health Interview Survey (NHIS). Usefully, “we know that the people being interviewed are representative of Americans more generally and we can pull together a number of years of data,” resulting in a large sample size and reliable follow-up, he said.

The notion that the link between alcohol consumption and health is inconsistent has previously been examined in various ways, Mukamal pointed out. “Even across countries you see some of the same sort of thing: in countries that tended to have higher socioeconomic status, that were more developed, you see a stronger, clearer relationship . . . than in other countries that have lower degrees of economic development.”

NHIS Data: 1997-2018

Zhu and colleagues turned to the NHIS to obtain data on 524,035 adults (mean age at baseline 50.3 years; 51.5% women) for the years 1997 through 2018, with follow-up lasting until 2019. A total of 13,266 deaths due to IHD occurred, amounting to a mortality rate of 204.8 per 100,000 person-years.

The researchers chose education level as the marker for SES because it “is less likely to be negatively affected by alcohol use and be subject to reverse causality compared with income and occupation,” the paper notes. Low level of education was defined as high school degree or less, middle level as some college, and high level was defined as a bachelor’s degree or higher. In their calculations, they adjusted for race/ethnicity, marital status, smoking, body mass index (BMI), physical activity, and survey year.

Compared with lifetime abstainers, men and women with high education levels who drank ≤ 20 g per day had a lower risk of IHD mortality. The relationships between alcohol and death were not significant for men and women with low education levels.

IHD Mortality With ≤ 20 g/d of Alcohol vs Lifetime Abstinence: Adjusted HR (95% CI)


High Education

Low Education


0.66 (0.57-0.77)

0.95 (0.82-1.09)


0.50 (0.41-0.62)

0.77 (0.54-1.10)


The protective association of drinking ≤ 20 g per day versus lifetime abstinence was larger in the high versus low SES group for men (interaction term HR 1.22; 95% CI 1.02-1.45), with an even more pronounced difference for women (interaction term HR 1.35; 95% CI 1.09-1.67). The impact of SES was only seen in people who engaged in heavy episodic drinking (at least five drinks on one occasion) less often than once per month.

For women, but not men, drinking ≤ 20 g per day versus lifetime abstinence was more protective for those who earned a bachelor’s degree versus attended some college (interaction term HR 1.35; 95% CI 1.06-1.72).

Men with low education who drank ≥ 60 g per day were at highest risk of dying from IHD among all the various combinations of consumption, SES, and sex. However, this harmful association was mostly explained by other behavioral risk factors like smoking, higher BMI, and lower physical activity.

“The poorer protective association of drinking less than 20 g per day with IHD mortality in the low-SES group may, in part, be explained by the fact that people with fewer resources have less access to healthcare services,” the researchers suggest. “Also, people with low SES are more likely to experience chronic stress across the life course, which is associated with increased blood pressure and the risk of IHD mortality.”

It's also possible that, as in the men who consumed the most, there is interplay with other risk factors or unhealthy behaviors and alcohol’s effects, they note. “In addition, individuals with higher SES may have healthier behaviors, be more aware of the risks associated with heavy drinking, and have better coping mechanisms or support systems.”

What to Tell Patients?

Mukamal said that as a primary care physician, the new findings are unlikely to change what he tells his patients about alcohol consumption. “In clinical care, we’re asking people about their drinking and trying to make sure they’re not exceeding currently recommended limits—whether they have higher SES or low SES isn’t going to change whether we give that advice or not. So, it isn’t so practical,” he commented.

What’s more interesting, he said, are the data on binge drinking, a habit that appeared to negate any apparent benefit of moderate consumption. These spikes in consumption, beyond their impact on health, could also detract from SES, for example by making it harder to hold down a job.

“If you ask people how much do you drink on average, they tell you what they usually do and they tend not to incorporate into that their episodes of binge drinking,” because they view that as a different kind of activity that’s an exception to their normal behavior, Mukamal said. “But we’ve shown previously, as have others, that even occasional binge drinking is really bad: it’s bad for all kinds of reasons. . . . I think that’s a really important message.”

Another key message is not to start drinking with the expectation that it’s good for your health, since there are easier and more evidence-driven ways to do that, he added.

Zhu stressed that their study is observational, which makes it tricky to directly apply the findings to clinical practice. “It is unknown whether the differential associations we identified in our study is due to [alcohol itself or confounders] such as access to healthcare,” she noted.

Up next, said Zhu, the research team is planning a follow-up study to measure how various risk factors—including alcohol use, smoking, BMI, and physical activity—mediate the link between SES and IHD mortality.

  • Zhu and Mukamal report no relevant conflicts of interest.