Search tips
Search criteria 


Logo of bmjThis ArticleThe BMJ
BMJ. 2003 September 13; 327(7415): 574–575.
PMCID: PMC194072

Self esteem and health

Autonomy, self esteem, and health are linked together
Michael Marmot, director

The starting point for Richard Sennett's recent book, Respect in a World of Inequality, is that society is riddled with inequality: of natural endowment and talent, of opportunities and life chances, and of achievement.1 We respect achievement. Hence these inequalities will be accompanied by inequality of respect. This, in turn, will be accompanied by inequalities in self esteem. Do such inequalities in self esteem matter? And if they do, is there anything to be done given that there will always be individual differences in earned respect?

The answer to both questions is probably yes—they do matter, and something can be done. There is a view that human needs form a hierarchy: keeping life and limb together takes precedence over such concerns as self esteem and respect. Doyal and Gough criticise this concept of hierarchy of needs and replace it with the idea that there are two basic human needs—health and autonomy.2 Autonomy is closely linked with self esteem and the earning of respect.1 Individuals do not worry about the means of achieving good health and only then concern themselves with autonomy. Both are basic and, I would argue, linked. Low levels of autonomy and low self esteem are likely to be related to worse health.

One way this can operate is through people's behaviour. Consider this example. The levels of obesity and diabetes among the Pima Indians of Arizona have long been recognised to be high. A small study tested the efficacy of lifestyle interventions. Two groups were identified. The Pima action group had a familiar mix of interventions on nutrition and physical activity. The Pima pride group looked remarkably like a control group for a health education trial—they received printed leaflets about activity and nutrition—but in addition they had regular discussions with local leaders on Pima culture and history. At the end of 12 months, much was going in the wrong direction for the action group, but the pride group had either less deterioration of risk factors or improvements. Compared with the action group the pride group looked favourable on weight, waist circumference, and blood glucose and insulin levels two hours after a glucose load.3 A tentative conclusion was that increasing pride in their identity had a more favourable impact on health behaviours and risk than focusing on how to change diet and exercise.

Turning to more direct pathways between psychosocial influences and ill health, ample data show the link between low self esteem and depression.4 The problem here, of course, is distinguishing causes from consequences. Low self esteem may be part of depressive illness rather than a step on the way.

Where the results are fatal, it is harder to argue that low self esteem was a consequence rather than a cause, especially if challenges to self esteem have dramatic results on the health of others. This is one interpretation of the close link between income inequality and homicide that has been noted internationally and among American states. Even in one city, Chicago, among 77 neighbourhoods a close relation was found between the degree of income inequality and rates of homicide.5 What is the link with self esteem? Accounts of life in the inner city emphasise the salience of respect and self esteem. “No small amount of mayhem is committed every year in the name of injured pride.”6 Putting this together with income inequality, the hypothesis is that unequal distribution of resources leads to increased competition for status among young men who have little to lose other than their self esteem and the respect of others. The results are violent confrontation and homicide.

If we link, as Sennett does, the concepts of respect, self esteem, and autonomy the theory implies that all people have a basic need for autonomy and self esteem. The effects of self esteem on health will depend on context. Where inequality is high people at the bottom of the scale may express their response to threats to their self esteem in violent ways. At the other end of the income scale the effects may also be dramatic. How else are we to interpret the finding that actors who have won an Oscar have a life expectancy that is four years longer than that of those who were nominated and did not win?7

Several studies have shown the links to increased coronary risk of low control in the work place and imbalance between efforts and rewards.8-10 This has been elaborated into a general framework.11 Appropriate reward for efforts expended and control over life circumstances are crucial, among other things, for the enhancement of self esteem. Threats lead to health damaging behaviours and to activation of biological stress mechanisms that increase risk of diseases such as coronary heart disease. These threats are unequally distributed in society and hence may contribute to inequalities in health.

But if inequalities are part of the human condition, what is to be done? Tawney wrote that “to criticise inequality and to desire equality is not, as is sometimes suggested, to cherish the romantic illusion that men are equal in character and intelligence. It is to hold that, while their natural endowments differ profoundly, it is the mark of a civilised society to aim at eliminating such inequalities as have their source not in individual differences but in (social) organisation.”12 The UK government has set targets for reduction in health inequalities. Achievement of these will require changes in social organisation, but changes that are sensitive to the issue of self esteem. Encouraging people off welfare and into work sounds like a step in the right direction. But the quality of jobs matters. No one can read Polly Toynbee's demeaning experiences of low paid jobs in the contracted out workforce that serves the public sector and relax with the comfortable nostrum that any job is better than none.13


Competing interests: None declared.


1. Sennett R. Respect in a world of inequality. New York: Norton, 2003.
2. Doyal L, Gough I. A theory of human need. Macmillan, 1991.
3. Venkat Narayan KM, Hoskin M, Kozak D, Kriska AM, Hanson AM, Pettitt DJ, et al. Randomized clinical trial of lifestyle interventions in Pima Indians: a pilot study. Diabet Med 1998;15: 66-72. [PubMed]
4. Cheng H, Furnham A. Personality, self-esteem and demographic predictions of happiness and depression. Pers Individual Differences 2003;34: 921-42.
5. Wilson M, Daly M. Life expectancy, economic inequality, homicide and reproductive timing in Chicago neighbourhoods. BMJ 1997;314: 1271-4. [PMC free article] [PubMed]
6. Newman KS. No shame in my game. New York: Alfred A Knopf and Russel Sage, 1999.
7. Redelmeier DA, Singh SM. Survival in Academy Award-winning actors and actresses. Ann Intern Med 2001;134: 955-62. [PubMed]
8. Marmot MG, Bosma H, Hemingway H, Brunner E, Stansfeld S. Contribution of job control and other risk factors to social variations in coronary heart disease. Lancet 1997;350: 235-40. [PubMed]
9. Theorell T, Karasek R. The demand-control-support model and CVD. In: Schnall PL, Belkic K, Landsbergis P, Baker D, eds. The workplace and cardiovascular disease. Philadelphia: Hanley and Belfus, 2000: 78-83.
10. Kuper H, Marmot M. Job strain, job demands, decision latitude, and the risk of coronary heart disease within the Whitehall II study. J Epidemiol Community Health 2003;57: 147-53. [PMC free article] [PubMed]
11. Siegrist J. Place, social exchange and health: proposed sociological framework. Soc Sci Med 2000;51: 1283-93. [PubMed]
12. Tawney RH. Equality. London: Unwin, 1964,
13. Toynbee P. Hard work—life in low pay Britain. London: Bloomsbury, 2003.

Articles from The BMJ are provided here courtesy of BMJ Publishing Group