The right kind of help beats more help, and giving it counts too

Key takeaways

• In a nine-year study of 6,928 California adults, people with the fewest social ties were 2.3 times (men) to 2.8 times (women) as likely to die as those with the most¹.
• Support comes in four kinds (emotional, practical, information and companionship), and it protects most when it matches what a stressful situation demands².
• Older adults who helped others had a lower risk of dying over five years, while getting help showed no such link, though studies like this can’t prove cause⁷⁸.
• Help that is too visible can backfire, but the research on hidden help is mixed, so the safest move is to ask what someone wants⁵⁶.
• Use the support map in this article to spot your gaps, then make one specific ask this week.

Researchers led by psychologist Niall Bolger studied couples at a predictably tense moment: the weeks when one partner was preparing for the bar exam. Both partners kept daily diaries⁵⁶.

The results were not what anyone expected. On days when exam-takers felt their partner had been emotionally supportive, they tended to feel more anxious and down the next day⁵. Distress rose least on days when the partner reported helping and the exam-taker hadn’t noticed⁵⁶.

The team called this “invisible support.” Years later, Bolger and colleagues ran a bigger version with more than three times as many couples. The pattern didn’t show up⁵.

That twist captures what four decades of research on social support teach. Support is good for health. But the kind of help, its timing and the way it’s offered can matter as much as the amount.

Here is what the evidence shows about the kinds of support, why giving help may help the giver, how England began prescribing connection, and how to ask for and offer help that works.

People with the fewest ties died at more than twice the rate

In 1965, researchers surveyed a random sample of 6,928 adults in Alameda County, California. Lisa Berkman and S. L. Syme then tracked who died over the next nine years¹.

People with the fewest social and community ties were far more likely to die. Compared with the most connected, their risk was 2.3 times as high for men and 2.8 times as high for women¹.

The link held after the researchers accounted for health in 1965, income and social class, smoking, drinking, weight, exercise and use of preventive care¹. So the gap wasn’t simply explained by poor health or low income at the start.

This was an observational study, which follows people over time without assigning them to groups. It can’t prove that friendship itself adds years. But it helped launch a field, and later reviews of many studies found the same pattern³. Our article on why people join wellness groups covers that larger body of work.

How could relationships reach the body? A review by UCLA psychologist Shelley Taylor describes several routes³.

Support can calm the body’s stress hormones, including cortisol³. Socially connected people also tend to carry less C-reactive protein, a blood sign of inflammation³. The popular idea that the “bonding hormone” oxytocin drives all this rests mostly on animal research so far³.

Support comes in four kinds, and fit matters more than amount

In 1985, psychologists Sheldon Cohen and Thomas Wills reviewed the research on stress and support. They sorted support into four jobs it can do, and noted that their scheme agrees with earlier ones²:

Kind of support	What it gives you	Everyday example
Emotional (esteem)	The sense that you are valued and accepted, flaws and all	A friend who listens without judging after a bad diagnosis
Practical	Money, goods, time or labor	A neighbor who drives you to physical therapy
Information	Advice or facts that help you understand and cope	A cousin who has had the same surgery and explains recovery
Companionship	Time together, belonging and a break from worry	A Saturday walking partner

Their review found two separate effects. Being part of a wide web of people, such as clubs, congregations and neighbors, was linked to better well-being in general². That is called a main effect.

Close, trusted relationships did something more specific. They seemed to shield people during hard times, which researchers call buffering². Buffering showed up most when people believed help would be there if they needed it and when the help matched the problem². Simple counts of how many people someone knew rarely predicted much².

The fit point is practical. Cohen and Wills expected emotional and information support to help with many kinds of stress². Practical help and companionship, they argued, protect mainly when they suit the specific problem².

When help doesn’t fit, it can hurt. Taylor’s review notes that controlling help, poor advice, or advice offered when someone wanted comfort can raise distress³. Negative exchanges also tend to weigh more heavily than kind ones³.

The bar-exam research shows fit in action. In a diary study of 68 law graduates and their partners, practical help from a partner was linked to less fatigue and more energy⁴. Emotional support that exam-takers noticed was linked to more anger, anxiety and low mood⁴. In that pressured month, hands-on help did more good than noticed pep talks.

Giving help is tied to a longer life, with a catch

Most advice about support focuses on getting it. Research on giving points somewhere surprising.

University of Michigan researchers followed 423 older married couples, first interviewed in 1987, for five years⁸. Some gave practical help to friends, relatives or neighbors, such as errands, child care or rides. Some gave emotional support to a spouse⁸.

Those who gave help had a risk of dying nearly 60% lower than people who gave neither kind⁸. Receiving help showed no link to survival once giving was taken into account⁷. The findings held after adjusting for health, mental health, personality and the quality of the marriage⁷.

Volunteering shows a similar pattern. A 2013 review of older adults found that volunteers had a 47% lower risk of death⁹. After adjusting for health and other factors, the gap shrank to 24%⁹.

Here is the catch. Healthier people are more able to volunteer and help, so part of the gap may run the other way. A second 2013 review, published in BMC Public Health, put this to a sharper test¹⁰.

It found 40 papers, including five randomized trials, where people are assigned by chance to one group or another¹⁰. Long-term tracking studies linked volunteering to less depression, more life satisfaction and a lower risk of death (about 22% lower across five studies)¹⁰. But the experiments didn’t confirm the mental health gains, and the authors judged the evidence too thin to call volunteering a proven health treatment¹⁰.

The fair reading: giving help is consistently linked to better health, and it may help. It costs little and helps someone else either way.

Mutual aid is an old natural-health practice

Long before clinics, people kept each other well by bringing food to the sick, minding children and sitting with the grieving. Natural health has always valued that kind of community and self-reliance as a group.

The research above suggests those habits did real work. Being able to give, not only to receive, may be part of what keeps a person well. A healthy support life runs in both directions.

“Invisible” help looked like the best kind, until a bigger study

Why would noticed support make stressed people feel worse? In lab experiments, Bolger and a colleague tested one answer⁵.

They randomly assigned college students to give a speech that would be judged. Some got visible help, such as advice or reassurance. Students who received visible help felt more distressed than those who got none, or who got help they didn’t recognize as help⁵. The likely reason: obvious help can signal that the helper thinks you can’t handle the task⁵.

A 2010 study of 85 dating couples added a twist⁶. Each couple talked on camera for seven minutes about something one partner wanted to change about themselves. Subtle, skillful practical help that the person receiving it barely noticed eased their anger and anxiety and raised their confidence⁶. But that worked only when it was both skillful and unnoticed⁶.

Then came the larger diary study of bar-exam couples, with more than three times the original number of couples⁵. It did not repeat the invisible-support pattern, and the authors stepped back from the simple message⁵. In some couples emotional support carried a cost; in others it didn’t⁵.

So what should you take from this? Not that you should help in secret. A better lesson is to help in ways that protect the other person’s sense of being capable. Keep the spotlight off the help, and ask what they want.

England now prescribes connection, but the evidence is still catching up

In social prescribing, a doctor refers a patient to a “link worker,” who connects them with nonmedical sources of support in the community. England’s National Health Service funds link workers in primary care¹².

Working from the records of 1.2 million patients, researchers at University College London estimate that about 1.3 million people were referred by family doctors in 2023 alone¹¹. That beat the 2019 goal of 900,000 referrals by 2023 to 2024¹¹. More than 3,300 link workers now work in England¹¹.

The service has also shifted toward people with the least. The share of referred patients from the most deprived areas rose from 23% before the 2019 rollout to 42% in 2023¹¹.

Does it improve health? A 2022 review in BMJ Open, funded by Ireland’s Health Research Board, found eight studies with about 6,500 people, five from the U.K. and three from the U.S.¹². Link workers made no difference to quality of life in four studies, and three of four found no effect on mental health¹².

There was one bright spot. Two U.S. trials of an intensive program called IMPaCT, for people with several chronic conditions in low-income areas, reported fewer days in the hospital, though the drop was significant in only one¹². The reviewers rated most of the evidence low or very low in certainty¹².

A 2026 study in eClinicalMedicine, funded by England’s National Institute for Health and Care Research, tracked 168,699 people referred between 2019 and 2022¹³. In the first six months, they averaged about one extra family-doctor appointment, and the authors say the early rise in care may fit with unmet needs being addressed¹³. By one year, the extra cost was about 26 pounds per person, which the authors call clinically insignificant¹³. The data showed who was referred, not who took part¹³.

In the U.S., ask your clinic whether it works with community health workers or patient navigators. Our article on peer support describes those connecting roles.

What the evidence says

Claim	Evidence level	What it rests on
People with more social ties tend to live longer	Strong evidence (for a link)	Large long-term studies, starting with Alameda County1
Feeling that help is available softens the effects of stress	Strong evidence (for a link)	Many studies reviewed in 1985, mostly observational2
Help that fits the problem works better than help that doesn’t	Promising but early	Theory plus diary studies24
Giving help or volunteering lengthens life	Promising but early	Consistent tracking studies; trials haven’t tested survival or confirmed mental health gains710
“Invisible” support beats visible support	Promising but early, and mixed	Lab and couple studies; a larger replication failed56
Social prescribing link workers improve health	Not enough evidence	Eight studies, low certainty12

Map your support and make one ask this week

This support map is our own tool. Fill it in on paper. It takes about ten minutes.

Kind of support	Who gives me this now?	Is it enough?	One small ask this week
Emotional: someone who listens
Practical: rides, meals, a hand
Information: someone who has been there
Companionship: someone to do things with
Giving: someone I help

How to read it. A blank row is a gap, not a failure. You might find, for example, that you have several people to ask for advice but no one to call when you need to vent. Pick the emptiest row and fill it first.

How to ask so people say yes:

1. Name the kind of help. People can’t guess whether you want advice or a listener.
2. Make it small and specific. “Can you drive me Tuesday at 2?” is easier to answer than “Let me know if you can help.”
3. Make it easy to say no. A graceful exit takes pressure off both of you and keeps the friendship safe.
4. Offer something back, now or later. Keeping the give-and-take balanced protects the friendship, and giving may help you too⁸.

Scripts you can borrow:

• Emotional: “I don’t need advice right now. Could you just listen for ten minutes?”
• Practical: “Dad’s surgery is Thursday. Could you pick up our groceries that day? No pressure if it doesn’t work.”
• Information: “You went through a knee replacement. What do you wish someone had told you first?”
• Companionship: “I could use some company. Want to walk the park loop Saturday morning?”

How to offer help that helps

The research on fit and visibility points to a few habits. This guide is ours, built from those studies.

• Ask before you fix. Try, “Do you want ideas, or do you want me to listen?” Advice given when someone wants comfort can raise distress³.
• Offer one concrete thing. “I’m making soup Sunday. Can I drop some off?” beats “Call me if you need anything.”
• Lean on practical help under pressure. In the bar-exam month, hands-on help did more good than noticed pep talks⁴.
• Keep the spotlight off yourself. Skip the speech about how much you’re helping. Obvious help can signal that you think someone can’t cope⁵.
• Don’t take over. Controlling help backfires³. Let the person decide what happens next.
• Respect a no, then check back. Offer again in a week or two.

When support isn’t enough

Sometimes the person struggling needs more than a friend. The National Institute of Mental Health (NIMH) advises professional help when severe symptoms last two weeks or more¹⁴. These include feeling hopeless, worthless or guilty, losing interest in things you enjoy, and having trouble concentrating or finishing tasks. Changes in sleep, appetite or weight count too¹⁴.

Some signs mean act now. NIMH lists talking about being a burden or wanting to die, and feeling trapped, empty or hopeless¹⁵. Pulling away from friends, giving away important things, and using more alcohol or drugs are also warning signs¹⁵.

If you or someone you’re supporting is in crisis or thinking about suicide, call or text 988 to reach the 988 Suicide & Crisis Lifeline, or chat at 988lifeline.org. It’s available 24 hours a day¹⁴. In a life-threatening emergency, call 911¹⁴.

Questions readers ask

The bottom line

Support is one of the best-studied links between daily life and health. Its power seems to come less from how many people you know than from believing help is there, and from getting help that fits the problem.

Ask for the kind of help you need, plainly and specifically. Offer help the same way: ask first, keep it practical, and keep the spotlight off yourself. And keep giving. The habit of helping others may be one of the quiet ways people stay well.

Keep exploring: loneliness.

Related reading

• People come to wellness groups for answers and stay for belonging
• What a peer who’s been there can do that a clinician can’t
• Online health groups offer reach, while in-person groups have stronger evidence
• Pressure makes loved ones hide their habits, and other kinds of help work better
• What eases loneliness, which depends less on how many people you see

Sources

1. Berkman LF, Syme SL. “Social networks, host resistance, and mortality: a nine-year follow-up study of Alameda County residents.” American Journal of Epidemiology, 1979. academic.oup.com
2. Cohen S, Wills TA. “Stress, social support, and the buffering hypothesis.” Psychological Bulletin, 1985. doi.org
3. Taylor SE. “Social Support: A Review.” In: The Oxford Handbook of Health Psychology, Oxford University Press, 2011. taylorlab.psych.ucla.edu
4. Shrout PE, Herman CM, Bolger N. “The costs and benefits of practical and emotional support on adjustment: A daily diary study of couples experiencing acute stress.” Personal Relationships, 2006. ovid.com
5. Shrout PE, Bolger N, Iida M, Burke C, Gleason MEJ, Lane SP. “The Effects of Daily Support Transactions During Acute Stress: Results From a Diary Study of Bar Exam Preparation.” In: Support Processes in Intimate Relationships, Oxford University Press (author copy). psychology.columbia.edu
6. Howland M, Simpson JA. “Getting in under the radar: A dyadic view of invisible support.” Psychological Science, 2010. socialinteractionlab.psych.umn.edu
7. Brown SL, Nesse R, Vinokur AD, Smith DM. “Providing social support may be more beneficial than receiving it: Results from a prospective study of mortality.” Psychological Science, 2003. psychologicalscience.org
8. University of Michigan. “People who give, live longer: U-M study shows.” Michigan News, 2002. news.umich.edu
9. Okun MA, Yeung EW, Brown S. “Volunteering by older adults and risk of mortality: A meta-analysis.” Psychology and Aging, 2013. pubmed.ncbi.nlm.nih.gov
10. Jenkinson CE, Dickens AP, Jones K, et al. “Is volunteering a public health intervention? A systematic review and meta-analysis of the health and survival of volunteers.” BMC Public Health, 2013. pubmed.ncbi.nlm.nih.gov
11. Social Biobehavioural Research Group, University College London. “New study tracks five years of social prescribing growth.” UCL, 2025. sbbresearch.org
12. Kiely B, Croke A, O’Shea M, et al. “Effect of social prescribing link workers on health outcomes and costs for adults in primary care and community settings: a systematic review.” BMJ Open, 2022. pubmed.ncbi.nlm.nih.gov
13. Wilding A, Munford L, Sutton M, et al. “Changes in healthcare use and cost associated with referrals to social prescribing link workers in England.” eClinicalMedicine, 2026;104102. doi:10.1016/j.eclinm.2026.104102 (pre-publication copy hosted by Voluntary Health Scotland). vhscotland.org.uk
14. National Institute of Mental Health. “My Mental Health: Do I Need Help?” NIH, revised 2025. nimh.nih.gov
15. National Institute of Mental Health. “Warning Signs of Suicide.” NIH, revised 2025. nimh.nih.gov