Do Happy People Live Longer?

My heart sank whenever a patient said to me that his or her ambition was to be happy: I knew that he or she was destined for misery because happiness is not the kind of thing that can be aimed at like the bullseye of a target, and that trying to aim at it is therefore an exercise in futility.

There is no subject upon which it is easier to be banal than happiness and happiness research, so-called, fills me with immediate gloom. There was a paper, however, in a recent edition of the Lancet that filled me with glee: it asked whether happiness was good for health and found that it wasn’t.

It is true that previous research, according to the authors, had found that happy people live longer, but that does not mean that it was their happiness that caused them to live longer. People who have bad health tend to be less happy at the outset than people with good health; the authors therefore set out to examine the question of whether happiness was an independent factor in mortality.

As is now the custom, their research was on a huge scale. They took 719,671 women in England and Scotland between the ages of 55 and 63 and followed them up for 10 years, during which 4 percent of them died. At the outset of the study they asked the women how much of the time they were happy, and they controlled for such potentially health-damaging variables as obesity, smoking and drinking, in order to find out whether happiness per se prolonged life.

I was mildly surprised to read that 39 percent of the women reported being happy “most of the time,” while another 44 percent were “usually” happy; only 17 percent reported being unhappy. Moreover, those who reported being happy rarely became unhappy, and vice versa: it was as if happiness were almost a matter of disposition (at least in English and Scottish conditions).

The factors that were most closely associated with happiness were living with a spouse or partner, strenuous exercise, participation in a religious group, age (the older the happier) and lack of education, with the most educated being less happy. Ignorance, it seems, really is bliss, though it does not actually cause bliss. There was a slight association of relative poverty with unhappiness, but not as great as some people might have anticipated.

As for unhappiness, it was most strongly associated with poor sleep and smoking. Whether smoking causes unhappiness, or unhappiness causes smoking, or they are both associated with something else, the paper does not attempt to answer.

Here I must issue the warning which everyone ignores whenever he has his own particular axe to grind: statistical association does not denote causation.

Women who had reported being unhappy at the outset of the study had a rate of death 1.29 times greater than that of happy women: a small increase for this kind of research, but statistically significant (i.e. it was unlikely to have arisen by chance). But when the figures were controlled for various initial risk factors for disease, the association disappeared. Happiness by itself did not preserve life.

As with most happiness research, the authors could not escape banality altogether. The opening sentence of the summary of their paper reads “Poor health can cause unhappiness and poor health increases mortality.” Is anything worth saying whose denial would be fatuous, for example that poor health cannot cause unhappiness or that poor health does not increase mortality?

The results of this study, even if believable, cannot be extended to all demographic groups. For example, among males aged 20 – 49 in England and Scotland suicide is the commonest cause of death, accounting for nearly a quarter of all deaths. Happiness would presumably preserve their lives.