Want to live a long life? Get married. Study after study has shown that married people, particularly married men, live significantly longer than their single friends. Some of the research is overall correlation, while other studies look at specific diseases and possible mechanisms. Doctors at Harvard tossed some more data on the pile last month, showing that married patients were more likely to identify cancer in its early stages and less likely to die from the disease than their unmarried peers. Epidemiologists refer to the well-established correlation between marriage and longevity as the “marriage protection hypothesis.”

The marriage protection hypothesis isn’t entirely surprising. Unlike the connection between alcohol and longevity, which still lacks a fully coherent explanation, there are a handful of intuitive and attractive reasons why marriage might extend your life. Having a family gives people something to live for, which may discourage risky behaviors like smoking and riding a motorcycle. Married men commit suicide at lower rates than singles, possibly for the same reason. Your spouse may urge you to get a mammogram, wear sunscreen, or have that worrisome mole checked out. A life partner provides an outlet to discuss personal stresses. (One medical argument in favor of legalizing same-sex marriage is that gay people have significantly lower stress levels when married.) Married people may remain more intellectually engaged with others, which helps avoid dementia. And healthy people may be more likely to attract a mate and marry than unhealthy people.

It’s not all good news for married people, though: Marriage also increases obesity rates. Getting married raises the risk of a woman becoming overweight by 3.9 percentage points compared to peers who did not marry, and marriage increases her risk of obesity by 1.4 percentage points. The effect is more pronounced for men. Married men see a 6.1 percentage point rise in the risk of becoming overweight and a 3.3 percentage point increase in the risk of obesity.

Obesity likely takes a sizable cut out of any longevity benefits of marriage. More than one-third of Americans are obese. It is a major risk factor for the leading cause of death, heart disease. The number of Americans living with diabetes, another illness linked to obesity, doubled from 1998 to 2013.

The proportion of adults who are obese increases with age; obesity can be caused by the cumulative effect of years of eating just a little more than we should. While less than 33 percent of us are obese between the ages of 20 and 39, nearly 40 percent of Americans older than 60 are obese. It’s a bit of a paradox that marriage is connected to both longevity and obesity, one of the primary contributors to early death.

The nature of the connection between marriage and weight gain is a matter of conjecture. One plausible explanation is the “marriage market hypothesis”: Single people remain thin for no other reason than to attract a mate. Once you’ve legally obligated someone to stick with you through sickness and health, there’s no compelling reason to eat right and exercise. But that theory flies in the face of the overall marriage protection hypothesis. If people quit smoking, wear sunscreen, and eschew suicide for the sake of their families, why would they allow themselves to become obese? (It’s important to note that mere age is not the explanation for the link between obesity and marriage. Researchers control for our tendency to gain weight as we get older, and the correlation is still easily detectable.)

There may be a simple explanation: People eat more when they’re together. A 1992 study that asked participants to record their food consumption for a week found that eating with one companion increased meal size by 41 percent compared with eating alone, while breaking bread with six or more people increased an individual’s caloric intake by a whopping 76 percent. Other studies have found somewhat more modest increases, but the effect is remarkably consistent.

Why do we eat more when we eat with others? It may be a result of social norms. Communal meals tend to last longer than eating solo, and people who sit in front of food can’t resist eating it. Another possible explanation is our habit of mimicking our dining companions: When they take a bite, we take a bite. More people means more cues to eat. The habit may, alternatively, lie deep in evolutionary history. Chimpanzees and marmosets spontaneously share food, probably to help form social bonds. From the innocent, altruistic act grew a regrettable human custom. Perhaps you feed (and feed, and feed) your spouse to strengthen your marital bond, even if you’re ultimately shortening its duration.

Sharing food may be a harmless gesture of friendship for marmosets, and it may have made sense for our hunter-gatherer ancestors, who were far more likely to die from scarcity than from the sequelae of diabetes. Our social customs, however, must evolve with us. There are better things to share with your spouse than dessert.  If that is all your sharing your definitely missing something in your marriage.

Marriage protects against mental illness and psychological distress. One study following young adults for seven years found that, even after taking into account the mental health of people prior to marriage, marriage boosted the mental health and reduced depression among young adults. Another study investigating the mental health of 13,000 men and women in older middle age (51 to 61 years), found that after controlling for race, education, family structure, income, and living arrangements, married people were less depressed and emotionally healthier than comparable singles.