People in the top quarter for friendship strength had a 24% lower risk of dying over the following eight years than those in the bottom quarter, along with a lower risk of stroke and depression, more positive mood and less loneliness. A separate 2025 analysis of the MIDUS cohort found the quality of friendships tied to lifespan more than the quality of family ties, with strained relationships mattering more than unsupportive ones.
How it was done
12,998 US adults over 50 in the Health and Retirement Study, followed for eight years. Friendship strength was a composite of three things: how many close friends someone had, how often they were in touch, and how supportive or straining those friendships were. The top quarter was compared with the bottom quarter, with a wide set of baseline health and social factors controlled for.
What it means
The composite is the point. Number of friends on its own is a weak measure; how often you are in touch and whether the friendship is a support or a strain are what the score mostly carries. That is consistent with the mortality meta-analyses, where richer measures of social integration predict survival far better than a headcount.
Vangelov, Dobson and Graham (2025, Behavioral Science & Policy) asked a different question of the MIDUS cohort: which relationship’s quality predicts how long people live. Friendships came first, family ties to a lesser degree, and a strained relationship was more closely tied to early death than an unsupportive one. We have read that paper in abstract only; the direction is clear, the sizes are not ours to quote.
Put those two next to the decay research and something falls out. Friendship is the relationship most tied to lifespan here, and it is also the one that fades when contact stops while family ties hold. The relationship with the most riding on it is the one that needs the upkeep.
