top of page
  • Becky McCall — United Kingdom

Contagion of a social kind

Non-communicable diseases and their risk factors seem to be common among certain groups of friends and families. Genetic reasons aside, the concepts of social contagion, shared spaces and a tendency for similar people to associate (known as homophily), might explain why these diseases appear to be more infectious than their name suggests

Audio: Listen to this article


Birds of a feather flock together is a familiar and age-old proverb. Close human relationships built on a basis of similarities permeate our lives, not least similarities relating to health behaviours and risks, among them obesity, excessive alcohol intake, tobacco use or poor diet.

Such health risk factors are typically associated with a broad category of disease known as non-communicable diseases (NCDs) which include so-called “lifestyle” diseases: cancers, heart disease, chronic respiratory diseases and diabetes. As a category, NCDs are of long duration and result from a combination of genetic, physiological, environmental and behavioural factors.

As the world battles the covid-19 pandemic, NCDs seem to have paled into insignificance in the human consciousness. Sadly, it is those affected by NCDs who are at greatest risk of the serious complications of coronavirus infection.

But NCDs and infectious diseases have other commonalities. Genetic inheritance aside, clusters of people with similar health and lifestyle related habits represent a form of social contagion, a hypothesis that endows NCDs with infectious characteristics of the non-biological, social kind.

Health behaviours and outcomes related to NCDs can be influenced by social contagion in networks of friends, group activities, and co-workers connected by interpersonal relationships. With respect to public health efforts, social connectedness also offers significant opportunity for improving outcomes for those living with NCDs by harnessing the power of group dynamics.

Another older concept that partially explains some of this social connectedness is homophily, the tendency for stronger social connections to develop among people with similar defining characteristics, such as age, gender and interests, than among dissimilar people. Certain cultural, behavioural, genetic, or material information that flows through networks tends to settle in certain groups, effectively social clusters.

“We prefer our group or tribe because we find it easier to communicate and relate to people with shared cultural and experiential backgrounds,” says Kayla de la Haye, a public health advocate from the University of Southern California in Los Angeles and a researcher in the field of social connectedness in health.

If some in a social network are obese others are more likely to follow

Crowd mentality

Anette Lykke Hindhede, a medical sociologist from Aalborg university in Denmark, also supports the view that NCDs and NCD risk factors spread through society in a socially “infectious” way.

“We know that people can become overweight through social relations, meaning that social norms and behaviours are contagious,” says Lykke Hindhede.

Most often, social contagion manifests as someone copying a certain behaviour of others with whom they have had contact or are physically near.

The study of social contagion reputedly began with the classical theory of crowd mentality advanced by Gustave Le Bon in 1896, Lykke Hindhede says. Trained as a doctor, Le Bon drew a parallel with crowd mentality and infectious diseases, explaining that ideas circulate in a crowd as microbes do in the human body.

Lykke Hindhede notes that key to the theory of crowd mentality is irrationality. “A crowd is not merely a gathering of individuals but rather a mental unity of people and they constitute a collective mind that takes the same direction.”

In 2012 in the US, sociologists Nicholas Christakis of Yale University and colleague James Fowler from the University of California explored the theory of social contagion. They showed substantial clustering in relation to obesity within social networks, finding that if some individuals in a social network become obese, then others in that network are more likely to become obese owing to social influences, such as the contagion of social norms or “mirroring,” copying what others do.

Distance learning

The pair’s work also challenged the assumption that an individual needs to be near others to become “infected” with a social norm. Using data from the Framingham Heart Study, they found that a social norm can be contagious even if people are not in the same vicinity. “If your weight increases then there’s a high probability that your friends, who might live hundreds of kilometres away, will also increase their weight. There’s a statistical probability that this will occur,” explains Lykke Hindhede.

“This social contagion not only occurs via face-to-face interactions but also via conversations, social media and other non-person-to-person forms of communication and these comprise the mechanics of establishing a social norm in a particular social network,” she adds.

The reputed causes of social contagion are fluid and have been the subject of differing interpretations over the years. In a 2018 study, which sought clarity and lends support to the social contagion theory, researchers looked at whether exposure to communities with higher rates of obesity increases the body mass index (BMI) of individuals and their risk of being overweight or obese. The study found this to be the case in both parents and children. “There was no evidence to support self-selection, or shared built environments, as possible explanations, which suggests the presence of social contagion in obesity,” the data revealed.

We know that people can become overweight through social relations, meaning that social norms and behaviours are contagious

Genetic or learned?

Chirag Patel and colleague Chirag Lakhani, data scientists from Harvard Medical School in the US, explored the respective roles of genetics and environment in determining health and disease, using a database of nearly 45 million people in the US, including 57,000 pairs of twins. The researchers looked at aggregate effects of genes and environment in 560 conditions, including some common lifestyle conditions such as obesity, cardiovascular and neuromuscular disorders. Postcode-related environmental risk factors, such as socioeconomic status, pollution exposure and climate, as well as medical data such as clinical diagnoses, were also included. “In our study, we asked, how do the genetic factors we inherit and the environment we share combine to make us individuals who get certain diseases?” Patel explains.

The researchers found that nearly 40% of the diseases in the study had a genetic component, while 25% were driven at least in part by factors stemming from sharing the same household and social influences. Eye disorders were most likely to be influenced by environmental factors with 27 of 42 eye diseases showing such effect, while respiratory diseases showed such an effect in 34 out of 48 conditions.

The strongest potential link to socioeconomic status was evident in morbid obesity, defined as a BMI of 35 kg or more. “Obesity is a great example of a disease that has both strong genetic and shared environmental factors,” says Patel. “Prior studies have shown the strong genetic component, but further to this, we found the shared environment was related to socio-economic background. Socio-economic status contributed to 3% of variation in obesity in the US.”

Like minds Homophily is the tendency for stronger social connections to develop among people with similar defining characteristics

Social networks for good

At the same time, social networks can be valuable for health improvement initiatives, De la Haye of the University of Southern California explains, as both a framework to study social interactions and a focus for intervention to change health-related behaviours.

Both social networks and shared environments are interconnected and act as triggers to health risk behaviours, she says. “These relationships influence behaviours, especially daily habitual behaviours such as drinking alcohol, diets, and exercise. These are all activities influenced by subtle cues that spread through social networks and shape our ideas of what is normal.”

Changing health risk behaviours via a social network might influence NCD outcomes, since the lifestyle risks for heart disease, cancers and other chronic diseases are well established. De la Haye says studies investigating the use of social networks have found that key hubs of interaction in networks and the people best connected within them are central to outreach success. Pressing the right buttons in a network work better than a healthcare worker alone sending out a brochure with messages about diet, smoking or other health behaviour.

One randomised controlled trial of more than 10,000 secondary school pupils in England and Wales used a social network approach to successfully prevent smoking uptake. The researchers, from Glasgow University and the University of Bristol, trained the most well-connected students to act as peer supporters at gatherings outside the classroom to encourage their peers not to smoke. For all pupils, the odds of being a smoker one-year after the intervention was reduced by 23%.

Lead by example

Family social networks were the focus of another study, “Social influence and motivation to change health behaviours among Mexican-origin adults: implications for diet and physical activity.” It explored making changes to dietary behaviours aimed at reducing the risks of complex disease caused by interactions between genes and environment. The results showed that among Mexicans living in the US, having at least one social network member who encourages others to eat more fruit and vegetables or do more physical activity improved these behaviours.

Children, spouses, and mothers were found to be especially important in influencing others.

Intervening in a network of socially connected groups, or targeting the best connected people in a network is more successful than targeting individuals alone, De la Haye observed. “Change is often both created and maintained because people find themselves embedded in a social group that supports, rather than resists [the positive behaviour].”

In terms of both researching and improving NCD outcomes, she strongly advocates for the social network-based approach. Social networks and social contagion lend support to the concept of NCDs being socially ”infectious” and the evidence indicates that interventions using these networks can be effective.

The individual approach so often taken by public health encourages someone to read information, be told what to think and then returns them to a social world and environment that counteracts this advice

Real life approach

Socially based approaches are also more in touch with real lives in the real world. “The individual approach so often taken by public health encourages someone to read information, be told what to think and then returns them to a social world and environment that counteracts this advice,” De la Haye points out. It is an approach often fails because families or individuals who live in a neighbourhood or network short of economic and other resources find themselves up against a major barrier to better health. “Often unjust social factors—less education, employment opportunity—are the reason they live in these environments in the first place,” she stresses. “Telling these people to just change their behaviour in social and built environments they are unjustly exposed to is especially unethical.”

Social contagion and social networks offer an opportunity to level the playing field for health and more generally too. Moving forward, health behaviour programmes based on social network approaches strongly suggest that the individual is not an island. Stimulating and activating someone’s social environment can maximise benefits. Importantly, evidence also suggests positive effects tend to be maintained.

Social networks have been shown to be particularly beneficial in helping hard-to-reach groups, such as people who inject drugs or minority ethnic communities. They may also have particular value in reducing the prevalence of NCDs. The digital age also brings new opportunities for influencing crowd mentality. Online social network platforms are growing in popularity, especially among adolescents. Their members may be receptive to the social network approach to health improvement.

Slowly, but surely, the scientific and public health community is waking up to the largely untapped potential held within the socially infectious nature of NCDs. You could say it is starting to catch on.


TEXT Becky McCall — ILLUSTRATION Trine Natskår


bottom of page