Good company is good health.
- One person you can tell anything
- A few friends you see on a schedule
- One group that meets without you organizing it

The protocol in seven steps.
Food is medicine. So is company. Across 148 studies with more than 300,000 people, those with stronger relationships were 50 percent more likely to be alive at follow-up, about seven and a half years later on average.
Most men treat friendship as something that happens when there is time left over. There rarely is. This protocol treats it like training: a few fixed habits, repeated. No medication does this part for you, and we do not sell one that claims to.
Count your people.
Start with an honest count. Who would you call with bad news? Who did you last laugh with until it hurt? Which group would notice if you stopped showing up? Write the names down. Most men find the list shorter than they expected.
You are not alone in that. In a 2021 survey of U.S. adults, 15 percent of men said they had no close friends, five times the share in 1990. Men with at least six close friends fell from 55 percent to 27 percent over the same period.
There is no proven right number of friends, so do not chase one. Look for the gaps: no confidant, no regular company, or no group. The steps below are sorted by those gaps.
Put friends in the calendar.
Friendship is built from time, and adult life does not hand that time out. Between 2003 and 2020, the time Americans spent in person with friends fell from 60 minutes a day to 20, according to the U.S. Surgeon General's 2023 advisory.
The fix is boring and it works: a standing appointment. Thursday dinner, Saturday morning run, the first Sunday of the month. A fixed slot removes the planning, which is where most friendships quietly end.
The hours add up faster than you think. One evening a week is more than 150 hours a year. Pick people you laugh with. Laughing together raised pain tolerance in a set of small experiments, a hint that it does something physical, though that evidence is thin.
Call on your walk.
A call does not need an hour or an appointment. Ten minutes on the way to the train is enough. Open with the truth: "I was thinking of you and wanted to hear your voice for a few minutes."
Two sets of experiments back this up. When 200 people reconnected with an old friend, those assigned to call felt a stronger bond than those who emailed. They had expected the call to be awkward. It was not, and it took about the same time as writing the email.
The unplanned part matters as well. In experiments with 5,900 participants, people underestimated how much a check-in would be appreciated, and the gap was largest when the contact came as a surprise. The call you hesitate to make is usually the one that lands best.
Do things together.
Meeting a friend does not have to mean sitting across a table and reporting on your life. Lift together. Do the grocery run together. Cook, walk the dog, fix the bike, watch the game. Doing something side by side takes the pressure off the conversation, and the conversation tends to come anyway.
It also solves the time problem, because these are hours you spend regardless. In the friendship studies from step 2, the hours that counted were leisure hours: hanging out, joking around, playing games. Hours spent working together counted for less.
We found no trial comparing side-by-side with face-to-face time. Treat this step as practical advice that fits the research, not as a tested method.
Tell one person the real thing.
Company is one thing. Being known is another. Pick one person and tell them what is really going on: the worry about your health, the job, the marriage. It does not have to be solved. It has to be said.
Men do this less. In the 2021 survey, 21 percent of men had received emotional support from a friend in the past week, against 41 percent of women.
If you have a partner, start there, and do not stop there. Among married men, 85 percent said their spouse is the first person they turn to with a personal problem. One person carrying everything is a fragile setup, for both of you. Add one friend or a brother to the list. If there is nobody, a therapist is a good first listener, and a sound choice on its own merits.
Join one group.
One-to-one friendships depend on two calendars. A group keeps going when you are busy, and it brings new people without any effort from you. A run club, a five-a-side team, a choir, a volunteer shift, a faith community. The topic matters less than the fixed day.
In a study of 424 people in England followed through their first six years of retirement, those who kept both of their groups had a 2 percent risk of death. Those who lost one had 5 percent, and those who lost both had 12 percent. The authors found the effect comparable to that of regular exercise. It is an observational study, so it cannot prove that the groups were the cause.
Talk to the people you pass.
Not every contact has to be deep. Learn the barista's name. Ask the neighbor how the move went. Nod to the regular at the gym, then say something next time.
In a set of small studies, people reported more happiness and a stronger sense of belonging on days when they had more of these casual exchanges than usual. It is the cheapest step in this protocol, and it is where new friendships start.

What to expect, and what this will not fix.
Expect this to be slow. A close friendship takes more than 200 hours, so count in months, not weeks. The first calls may feel clumsy. In the studies they went better than people predicted.
This protocol will not fix depression, grief or social anxiety. Loneliness that does not lift, or low mood that lasts more than two weeks, is a reason to see a clinician. It will not repair a relationship that is harmful either. The research is about good relationships, not about having people around at any price.
And it will not replace the rest of your protocol. Company supports sleep, nutrition and training. It does not stand in for them.

Quick answers.
Is loneliness bad for your health?
The data say yes. In a 2015 analysis of more than 3.4 million people, social isolation was linked to a 29 percent higher risk of early death and loneliness to a 26 percent higher risk. These studies observe people over time, so they cannot prove cause.
How many close friends do you need?
No number has good evidence behind it. Aim for one person you can tell anything, a few friends you see regularly and one group you belong to.
Is a partner enough?
A good partner is a strong start. Among married men, 85 percent turn to their spouse first with a personal problem. Add at least one friend, so one person does not carry everything.
How long does it take to make a friend?
About 50 hours together for a casual friend, about 90 for a friend and more than 200 for a close friend, in two studies of people who had recently moved.
Do calls and messages count?
Voice counts for more than text. In an experiment with 200 people, a phone call to an old friend built a stronger bond than an email and took about the same time.
The evidence, if you want to go deeper.
The central finding has been repeated for more than a decade. The 2010 analysis pooled 148 studies and 308,849 people and found 50 percent higher odds of survival with stronger social relationships. A 2015 analysis from the same group separated three measures.
A 2023 analysis of 90 cohort studies with 2.2 million people came to a similar result: social isolation was linked to a 32 percent higher risk of death from any cause, loneliness to a 14 percent higher risk. For the heart, a review in the journal Heart linked poor social relationships to a 29 percent higher risk of coronary heart disease and a 32 percent higher risk of stroke.
The longest view comes from the Harvard Study of Adult Development, which has followed men since 1938: first 268 Harvard students, later 456 men from inner-city Boston. Its director, Robert Waldinger, sums it up: "The people who were the most satisfied in their relationships at age 50 were the healthiest at age 80."
How company gets under the skin is less settled. The Surgeon General's advisory points to stress hormones, inflammation and behavior: people with close ties tend to sleep better, move more and see a doctor sooner.
Where this protocol stands on thin ice.
Nearly all of this evidence is observational. Nobody can assign people to twenty years of loneliness, so the studies follow people and compare. Illness can also cause isolation, which makes the link look stronger than it is.
The effect sizes are probably somewhat too high. An independent 2024 reproduction of the 90-cohort analysis confirmed its main claims but found smaller effects after adjusting for publication bias. And in the 2010 analysis, living alone on its own was not a clear risk (odds ratio 1.19, with a confidence interval that includes no effect). Living alone and being lonely are different things.
No trial shows that any step in this protocol makes you live longer. Steps 3, 4 and 7 rest on short experiments that measured how connected people felt, some of them with students. Step 4 has no direct trial at all. The friendship-hours figures come from people who had just moved, most of them young.
Programs designed to reduce loneliness have a mixed record. In a review of 50 of them, the randomized trials showed smaller effects than the weaker study designs, and the approach that worked best changed how people think about social situations, not how many people they met.
The line that loneliness is as harmful as 15 cigarettes a day is the Surgeon General's comparison. We have left it out of the protocol because it compares statistics from different kinds of studies.
Sources (17)
- Holt-Lunstad J, Smith TB, Layton JB. "Social Relationships and Mortality Risk: A Meta-analytic Review." PLoS Medicine, 2010. journals.plos.org/plosmedicine
- Holt-Lunstad J, et al. "Loneliness and Social Isolation as Risk Factors for Mortality: A Meta-Analytic Review." Perspectives on Psychological Science, 2015. scholarsarchive.byu.edu/facpub/7516
- Harvard T.H. Chan School of Public Health. "From Loneliness to Social Connection: Lessons from research and a global pandemic." hsph.harvard.edu/health-happiness
- Wang F, et al. "A systematic review and meta-analysis of 90 cohort studies of social isolation, loneliness and mortality." Nature Human Behaviour, 2023. nature.com/articles/s41562-023-01617-6
- Molo F, Pawel S, Fraga González G. "A Robustness Reproduction of a Systematic Review." I4R Discussion Paper Series 169, 2024. rwi-essen.de/i4r-discussion-paper-series
- Valtorta NK, et al. "Loneliness and social isolation as risk factors for coronary heart disease and stroke." Heart, 2016. heart.bmj.com/content/102/13/1009
- Office of the U.S. Surgeon General. "Our Epidemic of Loneliness and Isolation." Advisory, 2023. hhs.gov/.../surgeon-general-social-connection-advisory.pdf
- Cox DA. "The State of American Friendship: Change, Challenges, and Loss." Survey Center on American Life, 2021. americansurveycenter.org/research
- Cox DA. "Why Men's Social Circles Are Shrinking." Survey Center on American Life, 2021. americansurveycenter.org/why-mens-social-circles-are-shrinking
- Hall JA. "How many hours does it take to make a friend?" Journal of Social and Personal Relationships, 2019. University of Kansas summary. news.ku.edu/news/article/2018/03/06
- Kumar A, Epley N. "It's surprisingly nice to hear you." Journal of Experimental Psychology: General, 2020. University of Texas at Austin summary. news.utexas.edu
- Liu PJ, et al. "The surprise of reaching out." Journal of Personality and Social Psychology, 2022. News summary of the study. earth.com/news
- Sandstrom GM, Dunn EW. "Social Interactions and Well-Being: The Surprising Power of Weak Ties." Personality and Social Psychology Bulletin, 2014. repository.essex.ac.uk/15300
- Steffens NK, et al. "Social group memberships in retirement are associated with reduced risk of premature death." BMJ Open, 2016. bmjopen.bmj.com/content/6/2/e010164
- Masi CM, et al. "A Meta-Analysis of Interventions to Reduce Loneliness." Personality and Social Psychology Review, 2011. pmc.ncbi.nlm.nih.gov/articles/PMC3865701
- Dunbar RIM, et al. "Social laughter is correlated with an elevated pain threshold." Proceedings of the Royal Society B, 2011. pmc.ncbi.nlm.nih.gov/articles/PMC3267132
- Mineo L. "Good genes are nice, but joy is better." Harvard Gazette, 2017, on the Harvard Study of Adult Development. news.harvard.edu/gazette
This article is educational and is not medical advice, diagnosis or a treatment recommendation. Loneliness that does not lift, or low mood that lasts more than two weeks, should be evaluated by a clinician. If you are in crisis in the U.S., call or text 988. Study findings describe group averages and do not predict your result. No faber medication is prescribed for loneliness or to improve relationships.
The medicine is one part.
The protocol is the rest.
Every prescription comes with a plan for the habits around it.
Get started →