Mental health

Being generous linked with lower death risk

"Being generous can give you more than a warm glow... it protects health and helps you live longer," the Mail Online reports. 

The news is based on a study that looked into the effect that being caring has on our physical and mental health. This was analysis of research done in the 1980s and 1990s that looked at the health and lifestyles of older married couples.

As part of the original study, people were asked:

  • how much of their time they spent helping others
  • whether or not they had recently experienced stressful life events

Researchers traced their risk of dying over a five-year follow-up period. They then assessed the association between caring, stress and subsequent death.

In line with previous studies, the research found that people who experienced stressful events had a higher risk of death. However, people who experienced stressful events and spent time helping others did not have a higher risk of death. The researchers interpreted this as meaning that helping others buffers against the negative impact of stressful life events.

However, the study has considerable limitations that make it difficult to say whether these findings apply to most people and what the precise nature of the relationship is. 

Despite these limitations, the findings do add to our understanding of how helping others might also improve our mental – and possibly physical – wellbeing.

Where did the story come from?

The study was carried out by researchers from the University of Buffalo, Grand Valley State University and Stony Brook University in the US, and was funded by the US National Institute on Aging.

It was published in the peer-reviewed American Journal of Public Health.

The study was covered appropriately, if uncritically, by the Daily Mail.

What kind of research was this?

This research was a secondary analysis of data from a previous prospective cohort study that examined the relationship between providing help and support to others and the risk of dying. The study specifically aimed to determine whether experiencing stressful events changed this relationship.

The original study was carried out between 1987 and 1994. The participants were drawn from the Detroit area in the US. They were all married couples, and in each couple the husband's age was 65 or more. The current study analysed data from just over half of this cohort (846 of the original 1,536). 

Researchers traced their risk of dying over a five-year follow-up period. They assessed the association between caring, stress and subsequent death.

The researchers report that there is a known relationship between social connectedness (the amount of people you interact with in a meaningful way) and health. Over the last two decades, multiple studies have been carried out attempting to describe this relationship more fully.

Research into the link between receiving social support and health or mortality has come up with inconsistent results. The current study turns the relationship on its head and examines whether providing rather than receiving social support is the source of this link. The authors thought that helping others would reduce the association between experiencing stressful events and dying.

This observational research can describe the associations between helping, stress and longevity, but cannot determine whether helping behaviour directly causes changes in the risk of dying.

What did the research involve?

The researchers used data collected during interviews with 846 married people (423 couples) conducted between 1987 and 1988. The original study, called "Changing Lives of Older Couples", had been designed to study spousal bereavement, but also collected data on a range of other psychological, social and health variables. This means that it is also possible to carry out secondary analysis separate from the original goal of the initial study.

Participants were asked whether they had recently experienced stress or had provided help to their close network of friends and family.

For the purposes of the current analysis, the researchers classified participants as having recent stress if their interviews revealed that during the past year they had experienced:

  • serious non-life threatening illness
  • burglary
  • job loss
  • financial difficulties
  • death of a family member

To categorise providing help or support to others, the researchers used data on how much time over the past year the participants had spent on any of four specific activities for friends, neighbours or relatives who they did not live with:

  • transportation, errands or shopping
  • housework
  • child care
  • other (unspecified) tasks 

The researchers then used previously collected data to determine which of the participants died during the five-year follow-up period. This data was originally collected by scanning obituaries published in one of three local newspapers each day, as well as death records provided by the state of Michigan.

The researchers statistically tested the relationship between helping others, stress and mortality over time. This analysis was intended to determine whether experiencing stress influenced the relationship between helping others and the risk of death.

This analysis included several other variables that could potentially account for or confound the relationship, including demographic and socioeconomic factors, social interactions, self-rated health, health behaviours and mental health.

What were the basic results?

The average age of the selected group of 846 participants was 71. Overall, 134 people (~16%) died during the follow-up period.

At the beginning of the study, 74% of the sample reported having helped a friend, neighbour or relative, with the average amount of time spent in such activities ranging from 20 to 39 hours over the previous year. These participants were, on average, younger, healthier, of higher socioeconomic status, had more social contact and experienced more social support than people who reported not helping others.

Helping others was found to be predictive of a decreased risk of dying over the five-year follow-up period (hazard ratio [HR] 0.41, 95% confidence interval [CI] 0.29 to 0.57).

Overall, 70% of the participants reported experiencing none of the five stressful life events assessed in the study, while 26% reported experiencing one event, and 4% reported that they experienced two or three events over the previous year. Experiencing a stressful life event was associated with increased mortality over the follow-up period (HR 1.56, 95% CI 1.22 to 1.99).

When the researchers considered the association between helping others and mortality in the context of stressful events, they found that there was a significant interaction between helping and experiencing stress and mortality risk over time, even when controlling for other potentially confounding variables.

Stress was not significantly associated with mortality among individuals who reported helping others (HR 0.96, 95% CI 0.79 to 1.18).

Conversely, among individuals who did not report helping others each additional stressful life event was associated with a 30% increase in the risk of dying during the five-year follow-up period (HR 1.30, 95% CI 1.05 to 1.62).

How did the researchers interpret the results?

The researchers report that their study helps "to clarify what kinds of social connections are beneficial and why."

The authors point out that as this is an observational study (as opposed to an experimental design), they cannot rule out the possibility that confounding variables not assessed in this study could explain the observed relationship between helping others, stress and mortality.

They report that they included likely confounding variables, including "health and functioning, health behaviours, psychological well-being, personality traits, and social engagement and received social support" in their analyses.

The researchers concluded that, "helping valued others predicts reduced mortality specifically because it buffers the association between stress and mortality."


This study suggests that helping others is associated with a decreased risk of dying. Looking at the role of providing support to others is an interesting approach to examining the relationship between social connectedness with health and longevity.

At the risk of being ungenerous, however, this study has several weaknesses that should be considered. A major drawback is the fact that variables such as health and functioning, social engagement and psychological wellbeing were measured on a subjective, self-reported basis.

While it is important that the interviews included questions related to health and wellbeing, there is always a risk that subjective measures do not completely or accurately measure the variables of interest.

This is especially important, as the type of helping behaviours measured in this study were all considered "tangible aid" that require a degree of physical functioning and ability. Providing assistance with transportation, errands, shopping, housework or childcare may be less likely among people with mobility issues or health problems.

This doesn't necessarily mean that people with mobility issues or health problems can't or don't provide social support, just that they may find it more difficult to provide the specific types of support measured in this study.

There is a risk that focusing on tangible aid doesn't take into account other means of social and physical support that people provide. The researchers note this limitation and say that, "it is possible that expressing warmth and caring or emotional support" (for example, via a phone call or friendly email) is also beneficial.

The study also only assessed the support provided to people that did not live with the participants, which does not include any help given to spouses or ill relatives living in the same home. Again, this may have failed to encapsulate important sources of helping behaviour.

It is also important to note that this study was conducted among a very specific population: older married couples. We should therefore not assume that we can apply the results to people of all ages, and they may also not be applicable to non-married couples.

Finally, as the data from this study was collected in the US 25 years ago, it is worth considering whether the results apply in today's Britain.

Despite these limitations, this is an interesting study that provides something of an insight into an often neglected field of research – whether altruistic behaviour also brings individual benefits.

NHS Attribution