My comments address three general topics, the fundamental importance of social support in human behaviour and health, the strategic roles of peer support in prevention and health care, and some more general thoughts about behaviour change and health.

Social support as fundamental

Psychologists used to think that the basis for the connection between the infant and the mother was that the mother was the source of milk food. Harlow showed that, except when it’s hungry, an infant monkey went to the relatively warm, terrycloth mother on the left rather than the wire surrogate mother that was the source of milk. From this and a number of other studies, Harlow made the point that ‘contact comfort’, as he put it, is a powerful and fundamental characteristic of human behaviour.

A vast amount of research shows that having someone you can call on for a favour, with whom you can discuss personal matters, and who knows you and understands you helps you do better in all sorts of areas, from parenting adolescent children to aging well. The risk of death associated with social isolation, that is the lack of social contact, is comparable to that from smoking cigarettes.

We sometimes think of social support or peer-support programmes as frivolous influences with marginal benefit. The reality, however, is that these can be very powerful in harnessing the fundamental influence of social support. If we take them seriously, they can have real impacts in prevention and care.

Strengths of peer support

Harnessing social support, interventions by community health workers, lay health advisers, promotores de salud, ‘Lady Health Workers’ in Pakistan, ‘Village Health Volunteers’ in Thailand, or peer supporters by many other names, entail non-professionals helping each other lead healthy lives.

They encourage prevention and disease management, provide emotional support and encouragement, help people get the clinical care they need or find resources in their communities, and provide ongoing support for the lifelong tasks of avoiding disease or managing those we acquire. A growing literature shows peer support is effective in prevention as well as disease management, improving quality of life, encouraging ‘the right treatment at the right time’, reducing costly, avoidable care, and reducing overall costs. Four specific strengths of peer support are especially pertinent to ‘bending the curve’ of health care to achieve higher quality care that our societies can afford.

Peer support:

1. Reaches populations

For example: 85 per cent of 3,787 low-income, Latino adults with diabetes in a community clinic in Chicago.

2. Reaches and benefits those too often hardly reached

For example: reaches 89 per cent of low-income mothers of children hospitalised for asthma; in diabetes, benefits are greater among those initially low on medication adherence or health literacy.

3. Reduces psychological or emotional distress

Even when it is not designed to address these – the medium of peer support seems to provide implicit psychological support.

4. Reduces costly, avoidable care

For example: hospitalisations for low-income children with asthma; ‘normalises’ hospitalisation rates among distressed adults with diabetes in Hong Kong.

General considerations

What do Copernicus, Darwin, Watson and Crick, and Einstein have in common? No control groups.

Nevertheless, health research and agencies like the Cochrane Collaborative have enshrined the randomised control trial as the ‘gold standard’ for testing interventions. Well suited to testing the efficacy of a pill or a discreet medical procedure, it is poorly suited to testing broad, multidimensional interventions which are highly contingent on their contexts and often recruit those contexts to enhance their impacts. Interestingly, genetics may provide better models. It too entails impacts that are complexly dependent on multiple interacting influences.

Figure 1: Van de Vijver et al. New England Journal of Medicine. 2002 347: 1999-2009

Figure 1 portrays genetic ‘signatures’ with poor and good prognosis for survival of breast cancer. It is clear that no element of either signature is necessary or sufficient for survival or death. Rather, complex arrays of elements alter the probability of survival. Similarly, in communities and the settings of individuals’ lives, complex arrays of elements alter the likelihood of healthy or unhealthy behaviours. Studying those elements in their complexity, instead of trying to isolate them by randomisation, may be a valuable strategy.

From the variety of interventions and considerations presented during the day, two generalisations occurred to me as being of value:

1. Context trumps content

The programme presented from Stockport was fascinating in this regard. It did not focus on the content of collaborating organisations’ programmes. Rather, it focused on linkage among them. Within broad limits, the leaders in Stockport almost didn’t care what the interventions entailed as long as linkages among them were improved. A context of linked and coordinated services may provide greater benefit than the content of any one service.

2. Structure trumps messaging

If we don’t have structures in place, if we don’t have choices realistically available to people, if we don’t have incentives in place to encourage the behaviours we want to encourage, then perfecting messaging is unlikely to have great effect. On the contrary, salient, available, reinforcing healthy choices attractive healthy alternatives will tend to command individuals’ choices with minimal messaging to encourage them. Many years ago, I asked an advertising executive to identify the key ingredient of a successful advertising campaign. His wise answer: “The product.” The range of ideas and programmes and approaches presented during the meetings suggested to me the shift in emphases portrayed in Figure 2.

Figure 2: Weltanschauugen – Two world views

To explain a few of these, the old model on the left sees illness as a micro, discreet event at the level of the individual. In contrast, the new model on the right sees health as a macro, complex and social set of events. Accordingly, the two models go from illness and the focus on clinical indicators to patterns of living in the focus on quality of life in a community or population. In community and people-powered movements to enhance health, we go to from data on individual events like a blood pressure reading to mass data – either a mass of data points across an individual’s experience or across many individuals.

We go from looking at the individual as the site of disease and as the locus of change to the individual as socially situated; as in Hillary Clinton’s book title, It takes a village. In going from micro to molar effects, we are not so interested in what the individual with diabetes eats today but in what she or he eats over 30 or 40 years of living with diabetes. Instead of looking for a unique and dominant effect of a single intervention, we recognise that interactions among influences are the rule and seek to understand the complexity of those interactions.

From a statistical perspective, the world is not orthogonal. Finally, we go from seeing motivation and energy as based in the individual to seeing energy as emergent among individuals and from communities. More generally, instead of putting responsibility – and too often blame – on the individual to guide their own health, we seek to understand how communities enable individuals to lead healthy lives and look for ways to help communities do that better.