Soap operas are fictional television or radio serials based on the daily lives of a group of characters, often within a localised community. They enjoy immense popularity in the UK, where soaps such as EastEnders and Coronation Street attract an average of 10 million viewers per episode, making them the most watched programmes on the channels they are broadcast on.¹

A number of examples of health-related story lines in popular television programmes appear to lead to increased viewer knowledge or behaviour change in the developed and developing world—both by accident and by design.^2–7 However, well-controlled, randomised studies of the long-term effects are hard to conduct and strong evidence of causality is not available.

Although the portrayal of alcohol on UK, US and Canadian television^8,9,10 and smoking on US television¹¹ has been described previously, little is currently known about the portrayal of a wider range of health-related behaviours on UK television soap operas. To provide initial descriptive data, we assessed the portrayal of some key health-related behaviours in UK soap operas.

METHODS

The four UK-based soap operas broadcast on UK terrestrial television with highest average weekly viewing figures were identified¹ (table 1), and all episodes first broadcast between 22 April and 23 May 2005 video recorded.

From the “overarching priorities” identified by the UK government’s recent white paper on public health (Choosing health),¹² four key behavioural areas were selected for inclusion—alcohol consumption, smoking, food choice and physical activity. A list of 11 specific behaviours of interest were then generated using current national guidelines on health promotion and health behaviour^12–15 (see box 1, available at http://jech.bmj.com/supplemental). To avoid double counting, all behaviours were classified under only one heading—for example, an episode of smoking around children was also not classified as smoking, non-specific.

Video recordings were watched by a single researcher and the occurrence of any behaviour of interest was noted—with the pause and rewind facilities of the video player used as necessary. Portrayals of food and drink that could not be clearly identified were excluded.

To confirm the validity of the data collection method, 10% of episodes included in the main analysis was randomly selected for independent analysis by a second researcher.

RESULTS

Eighty episodes of soap operas were included, giving a total of 32 h of programming (table 1). Overall, 959 instances (30 per programming hour) of the behaviours of interest were identified.

A high degree of inter-rater agreement was found, with the second researcher identifying and agreeing on the category of 79 of the 80 (99%) behaviours that the first researcher recorded in the 10% validation sample—with no additional behaviours recorded by the second researcher.

Table 2 shows the absolute frequency and frequency per programming hour of each behaviour of interest. The most frequent behaviour of interest recorded was non-specific alcohol-related behaviour—identified 619 times (19.3 per programming hour). No instances of four behaviours of interest were observed: driving after drinking, drinking during pregnancy, smoking during pregnancy or smoking around children.

DISCUSSION

In this first survey of the portrayal of a wide variety of health-related behaviours on popular soap operas set and broadcast in the UK, we recorded 959 portrayals of the behaviours of interest (30 per programming hour). Of the 11 behaviours of interest, alcohol-related behaviours were the most commonly portrayed (on 619 occasions, or 19.3 per programming hour). There were no portrayals in the sample of four behaviours of interest: driving soon after drinking, drinking during pregnancy, smoking during pregnancy and smoking around children.

We have considered only a limited number of health-related behaviours in this analysis. In particular, we did not include sexual health-related behaviours or use of illegal substances. However, the four behavioural areas (alcohol consumption, smoking, food choice and physical activity) and the specific behaviours focused on were identified using current UK government recommendations^12–15 and so were of current policy relevance.

As it was occasionally difficult to categorise some behaviours—particularly food and drink—we excluded a number of observed behaviours. As these behaviours were not recorded, we do not have data on how frequently this occurred. This necessarily limits the comprehensiveness of our data. However, it is unlikely that food and drink that could not be identified by a dedicated researcher would be identified by a casual viewer or influence their beliefs and behaviours. It is not clear how more comprehensive identification of behaviours could be achieved.

Our data also lack details on the context of behaviours. Although we initially attempted to collect contextual data on characters, we found even estimating characters’ ages difficult. Summarising more detailed information on the exact circumstances of almost 1000 behaviours would have been very challenging. However, the data presented here were intended to be descriptive only. More detailed data, perhaps collected in different ways, would be necessary to test specific hypotheses.

Our data were collected over 4 weeks in spring 2005. The behaviours portrayed may, therefore, have been influenced by dominant story lines and seasonal variations in behaviours. In addition, we only included UK-based television soap operas broadcast on terrestrial channels. However, in keeping with work from the UK and elsewhere, we found that the frequency of alcohol-related behaviours was high^8,9,10 and the frequency of smoking in the sample was comparable to previous data from the US.¹¹ These similarities suggest that, despite their limitations, our data may be generalisable.

All the soap operas included in the sample include a pub or bar, as well as a café—providing common areas where characters can interact. This may explain the preponderance of alcohol-related behaviours and unhealthy foods, and it is possible that the formulaic nature of British soap operas goes hand in hand with the portrayal of unhealthy behaviours.

There was some indication that behaviours portrayed varied according to the target audience of the different soap operas included. For example, Hollyoaks is primarily aimed at a younger audience and there was a higher frequency of portrayal of healthy food and physical activity in this programme.

Given the evidence that behaviours portrayed on television may affect both the knowledge and behaviour of viewers,^2–7 the overall frequency of “unhealthy” behaviours in our sample, such as drinking and “unhealthy” food choice, is of concern. Programme makers may feel, with some justification,¹⁶ that unhealthy behaviours are the norm in the population groups they are portraying. However, popular television serials offer the chance to portray healthier behaviours as normal and so help change attitudes, shape behavioural norms and, ultimately, change behaviour among the viewing public. Engaging the makers of these programmes in a health promotion agenda may be a fruitful method of promoting healthy behaviours among a wide cross-section of the population—perhaps by the establishment of voluntary liaison with public health policy makers.