Couples who share similar drinking habits tend to live longer than those who don't, according to a study of more than 9,000 people.

This might sound like a suggestion to drink more with your partner, but the researchers from the University of Michigan warn against taking it too literally.

The new findings highlight the importance of considering how relationships affect health in complex ways. Past studies have looked at how drinking habits affect couples' relationships, but the broader effects of their shared or solitary drinking behaviors on health are not as clear.

"The purpose of this study was to look at alcohol use in couples in the Health and Retirement Study and the implications for mortality," says psychology researcher Kira Birditt, first author of the published paper.

The 'drinking partnership' concept, first described in 1998, suggests that married couples' drinking habits can affect their relationship; particularly if one person drinks more heavily than the other, it is associated with more marital distress and conflict.

"The majority of research has examined younger couples and discordant heavy drinking," Birditt and colleagues write in their published paper.

"Findings from these studies indicate that discordance is associated with lower marital satisfaction, greater verbal aggression, and higher divorce rates."

The opposite may have its own benefits, though. Shared drinking habits could bring couples together, or at least not push them apart.

The 9,312 participants studied were from 4,656 different-sex couples who were married or living together. They had also completed at least three waves of the Health and Retirement Study, a survey of adults aged 50 and older in the US that includes couples who are interviewed every two years.

The researchers focused on these older couples' drinking patterns in particular, rather than what type of alcohol they consumed. If participants reported drinking alcohol in the past three months they were asked to report the average amount consumed per week.

"We found, interestingly, that couples in which both indicated drinking alcohol in the last three months lived longer than the other couples that either both indicated not drinking or had discordant drinking patterns in which one drank and the other did not," Birditt explains.

They also found that light or moderate drinking in both partners was associated with living longer, compared to both partners drinking heavily, or not at all.

It's not clear why a couple's drinking is linked to a higher chance of survival when their individual alcohol consumption was not. This latest analysis of drinking's effects on longevity took into account individual chronic conditions like heart disease and diabetes. The researchers want to learn more about how couples' drinking affects their relationship.

It does highlight the broad ways in which partners can affect one another's well-being. Alcohol consumption levels that are similar between couples may show how well their lifestyles, closeness, and relationship satisfaction are matched.

"We've also found in other studies that couples who drink together tend to have better relationship quality, and it might be because it increases intimacy," Birditt says.

The study has limitations, for example it only analyzed people in hetrosexual relationships, that only involved one other person, while at a population level there are many kinds of relationships to consider.

Research suggests that alcohol affects couples differently in same-sex and different-sex relationships, so more inclusive investigation is needed before we draw broad conclusions about this study's results. A closer look at the impact of the types of alcohol people drink could be useful too.

"There is also little information about the daily interpersonal processes that account for these links," Birdett adds. "What are their daily lives like? Are they drinking together? What are they doing when they are drinking?"

Understanding how couples influence each other's health behaviors is key for improving health policies and practices for older adults, the team says. Interventions should consider both individuals' behaviors for better outcomes.

"Overall this study moves the field forward," the researchers write, "by showing that drinking concordance is associated with longer lives but that the amount of drinking by both partners is also important to consider."

The study has been published in The Gerontologist.