Eating with a partner lowers post-meal blood sugar, study finds

The human body constantly works to maintain a stable internal state — a process known as homeostasis. When blood sugar rises after a meal or body temperature drops in the cold, the body works to return to its baseline.

In a body where homeostasis functions well, post-meal blood sugar does not spike too high or linger too long before returning to normal levels.

A new study has found that eating with a romantic partner, rather than alone, leads to a smaller blood sugar spike that also subsides more quickly. The same pattern held across other physical stressors: people recovered faster from cold exposure and mild electric shocks when their partner was present.

The findings were published in volume 12, issue 39 of the international journal Science Advances by Monia Masalha and Shir Atzil of the Hebrew University of Jerusalem, who examined how blood sugar, body temperature and stress responses change depending on whether a person is with a romantic partner or a mother.

The research team had 54 couples — 108 people in total, aged 18 to 37, none with diabetes — eat 102 grams of white bread containing 50 grams of carbohydrates while their blood sugar was monitored for two hours.

Each participant ate the bread twice: once alone and once with their partner. When together, the couples were free to talk.

The researchers measured both how high blood sugar climbed above the normal fasting threshold of 100 milligrams per deciliter and how long it stayed elevated. The overall blood sugar burden was on average 14.6 percent lower when participants ate with their partner than when they ate alone.

Among those who ate together, blood sugar was lower at every measured point from 45 to 120 minutes after the meal. When comparing individuals who ate alone with those who ate with a partner, the difference reached as high as 23.9 percent.

The research team said stress levels during the meal did not differ between the alone and together conditions, and that the precise mechanism behind the blood sugar difference remains unclear.

The research team applied the same approach to other bodily functions.

They had 116 participants enter a room cooled to 12 degrees Celsius for five minutes and measured how much their skin temperature dropped and how quickly it recovered.

When participants entered the cold room with their partner, the drop in skin temperature was 6.8 percent smaller than when they were alone.

In a separate experiment involving mild electric shocks to the arm, participants who had their partner present showed a stress response 6.0 percent smaller than those who were alone.

The team also conducted an experiment with 56 infants aged 2.5 to 13 months, exposing them to a loud noise. When their mother was present, the infants' stress response was 2.6 percent smaller than when the mother was in another room.

The research team said the findings may help explain why social animals — not just humans — go to considerable lengths and accept real risks to stay close to others of their kind.

"Being together may serve not only the purposes of protection, cooperation and reproduction, but also a more fundamental function of enhancing homeostasis itself," the team wrote in the paper.

The team said the differences between being alone and being together were modest, in part because the experiments captured only a narrow slice of bodily function over a short period. They added that even small effects could become meaningful if they accumulate over time.

According to the research team, individual variation was also substantial. Some participants gained a clear benefit from being with their partner, while others actually showed weaker homeostasis when together than when alone.

The team said the results reflect what was observed in close relationships and should not be taken to mean that being with anyone, in any context, is beneficial to the body.

DOI: 10.1126/sciadv.aeg4937

Monia Masalha et al., "Social Physiology: Human homeostasis is more efficient in social proximity." Sci. Adv. 12, eaeg4937 (2026).