Sunlight exposure: risks and benefits

Background

The committee agreed that sunlight offers risks and benefits according to the population group and a range of other variables. It also agreed that the order of the words 'risks and benefits' does not imply a hierarchy but is used to ensure consistency throughout the document, in line with NICE's house style.

Determining and quantifying the contribution sunlight makes to vitamin D status (and how high‑protection sunscreen may reduce this) was beyond the remit of this guideline. Committee members were aware that the Independent Advisory Group on Non‑ionising Radiation (AGNIR) was considering the links between sunlight and vitamin D during development of this guideline. Any new findings from the AGNIR report will be taken into account when this guideline is updated. In addition, the committee noted that NICE has published a guideline on how to increase vitamin D supplement use among at‑risk groups. Members hoped that these 3 pieces of work will provide the basis for clear, consistent advice to reduce the risk of low vitamin D status among all at‑risk groups.

The causal relationship between vitamin D status and musculoskeletal health is well established. However, the nature of the association between vitamin D levels and other chronic diseases, such as cancer and multiple sclerosis, is unclear. The committee was aware that the Scientific Advisory Committee on Nutrition (SACN) was reviewing vitamin D and health outcomes, and the recommendations in this guideline should complement SACN's final conclusions.

The committee considered SACN's draft report on vitamin D when it was issued for consultation, towards the end of the guideline development process. SACN indicated that sunlight is the major source of vitamin D for most people. But it could not quantify how much exposure people need in the summer to maintain vitamin D levels in winter. SACN did not, therefore, include sunlight in calculations to establish a reference nutrient intake (RNI) for vitamin D. In its draft report, SACN proposes an RNI of 10 ug per day for the general UK population as a precautionary approach, to take account of variable exposure to sunshine and diet. To reach this proposed RNI, people would need to take a daily supplement of vitamin D. The committee noted that not everyone will choose to take a supplement or make a point of consuming (natural or fortified) dietary sources of vitamin D. Furthermore, some people may prefer to get their vitamin D from sunlight. Members concluded that there is still a need to provide clear advice on how to safely get vitamin D from sunlight, regardless of SACN's final recommendations.

The committee acknowledged that people at risk of overexposure to sunlight and those at risk of not having enough vitamin D may be in different groups. So it recommended the need to adapt messages for different groups and individuals. But members also noted that consistent universal messages will help change attitudes and behaviour. The committee aligned messages in this guideline with national advice from NHS Choices to achieve some consistency.

It is not possible to provide a simple definitive message telling different groups how often and how long they can be exposed to sunlight to ensure minimum risk but maximum benefit. That is because the amount of UV someone gets from sunlight depends on a range of biological, environmental and behavioural factors. But the committee agreed that advice on preventing both skin cancer and low vitamin D status can be combined. It heard (from expert papers 4 and 5) that short (less than the time it takes for skin to redden or burn), frequent periods of sunlight exposure are best for vitamin D synthesis. In addition, this type of exposure is less likely to result in skin cancer.

Advice on sunlight exposure is available from many organisations and sources (as summarised in expert paper 7). However, the information given is often inconsistent and potentially confusing. PHAC considered that this was likely to be due to a lack of consensus in the evidence. The committee agreed that if a consensus could be achieved on the approaches and messages needed, it would minimise public confusion and increase the likelihood of behaviour change. It would also minimise the duplication of effort by different organisations. The committee also agreed that a central source of messages on sun exposure would be helpful.

The committee noted that both practitioners and the public find it difficult to judge 'skin type I–VI'. To overcome this problem, the recommendations refer to both skin types and 'lighter and darker' skin.

The committee acknowledged the importance of adults or young people 'knowing their own skin'. Ideally, nobody should experience sunburn, but many young people and adults will have done so in the past (albeit inadvertently). Members agreed that, in such cases, people could use the experience to understand how their skin reacts in sunlight and to adequately protect themselves in future. Members hoped that children would never experience sunburn and therefore would not need to learn from such an experience.

Approaches to skin protection

The committee emphasised the need to use suitable clothing, shade and sunscreen in combination. Members did not want any of these methods to be promoted individually as the 'main' way of providing protection.

Members viewed advice on how to use sunscreen as particularly important because often it is not applied effectively– and people overestimate the protective effect. The committee acknowledged that the use of sunscreen may encourage people to spend a long time in the sun and that will, in turn, increase the risk of sun damage.

Expert testimony confirmed that frequent, liberal use of high‑protection sunscreen may prevent vitamin D synthesis. But this is only true under research conditions. Evidence suggests that it is unlikely to be the case in practice. This is because people may inadvertently miss some areas of skin and they also tend to apply much less sunscreen than the manufacturers recommend.

The committee debated whether to recommend SPF15 or 30 sunscreen. If SPF15 is applied liberally, according to the manufacturer's instructions, it should offer adequate protection in most cases. However, this level of coverage is difficult to achieve. Using SPF30 or higher may partially overcome problems arising from inadequate application. But it does not mean people can spend longer in strong sunlight without risking burning. Therefore the committee did not want advice about choice of sun protection factor to be separated from practical advice on sunscreen application.

The committee recognised that a range of sunscreen products are available at different costs. However, members were concerned that the cost of sunscreen or specialist protective clothing could be off‑putting for some people and limit their ability to protect their skin.

Behaviour change

The committee was aware that cultural context may influence whether or not people respond to public health messages. Information is usually a necessary precursor to behaviour change, but information alone is not always enough. Members agreed that the best outcome from information provision is a change in behaviour. But they also felt there was some value in using information to alter attitudes for example, towards tanning, because this may eventually lead to behaviour change.

Report 1 on Communicating the benefits and risks of ultraviolet light to the general population: a qualitative documentary analysis of UK newspapers and magazines (print and online) highlighted a generally positive portrayal of sun tanning in the media. For example, images of sunbathing are usually accompanied by references to a 'healthy tan' and the value of 'escaping to the sun'. The committee recognised that it is a challenge to change people's knowledge, attitudes and behaviour towards tanning.

The degree to which people believe they can change their level of risk plays a role in their decision‑making process. The time‑lag between sunlight exposure and the development of skin cancer and wrinkling also play a part. Members agreed that there is a need to help people more accurately determine how they can achieve vitamin D synthesis while not damaging their skin.

The committee recognised the importance of making children aware of the risks and benefits of sunlight. It also discussed the importance of helping children establish life‑long health‑promoting behaviours when they are most susceptible to habit‑forming advice.

The risk‑benefit ratio of sunlight exposure will vary depending on how dark or light someone's skin is. The committee was particularly concerned about the risks and benefits for darker skinned people because much of the evidence and existing advice is focused on those with lighter skin (as outlined in expert paper 5).

Evidence

The evidence base underpinning the content of safe sunlight exposure messages was not systematically reviewed for this guideline because the content of these messages was beyond the remit of the guideline. The advice from NHS Choices was the nearest that could be achieved to a consensus, following a trawl of existing authoritative sources.

A large volume of evidence suggests sunlight may provide protection against chronic diseases such as cancer, heart disease and diabetes. However, the relationship is associative rather than causative that is, it has not been proven (SACN is currently considering the relationship between vitamin D and a range of health outcomes). Sunlight is also associated with improved mental wellbeing. (But this is to do with the visible rather than the UV component.)

The balance of published evidence suggests that skin with darker pigmentation needs longer sunlight exposure than lighter skin to produce equivalent levels of vitamin D. But further research is needed. In the meantime, the committee was clear that people of all skin types should not risk burning their skin.

The evidence on the effectiveness of strategies to communicate complex messages was very limited.

The review of effectiveness identified a number of interventions that have changed behaviours in the sun, or reduced the incidence of sunburn. But none of the studies focused on delivering a complex message that conveyed both the risks and benefits. The committee also noted that the interventions in the review tended to have small sample sizes, small effect sizes and measured only short‑term outcomes.

Most studies identified in the evidence reviews were based in countries with a very different climate from the UK (for example, Australia and the US). The committee also felt that it would be difficult, for example, to transfer evidence from Australia to the UK context because Australian campaigns have been in place for longer and are better funded than in the UK. The committee was also aware that studies on people at risk of low vitamin D status would need to be judged in light of whether the study took place in a country that fortifies food with vitamin D. (Because this would result in the population having higher baseline levels of vitamin D.)

There is growing interest in the use of new technology, including phone and tablet apps, to deliver behaviour change interventions. But the committee noted a lack of formal evaluations of effectiveness. In addition, although currently there is no evidence to show text messages are cost effective, members were aware that this may change. They suggested that any such change could be captured in an update of this guideline.

Photoageing interventions were not found to be cost effective at the time of publication, so they were not recommended for NHS settings. But the committee acknowledged that this did not mean they were not effective.

The committee did not look at evidence on the risks and benefits of artificial sources of UV rays because it was beyond the remit of the guideline. The absence of any recommendations on these sources should not be taken as a judgement on whether they are beneficial, cost effective or pose any risks.

The committee recognised that it is not easy to understand how to use information from the UV index to assess the risks and benefits people face from sunlight. They agreed that the information it provides is only useful if combined with someone's own skin type and behaviour.

Health inequalities

The recommendations stress the need for tailored individual advice to back up the universal messages. The committee noted that universal interventions could result in adverse effects for some groups and so increase health inequalities. For example, universal messages about protecting the skin from sunlight exposure may inadvertently lead to a reduction in the amount of skin exposed to sunlight among groups at risk of low vitamin D status.

Many people have photosensitive skins, for various reasons, which means that sunlight exposure has particular implications for their health. The committee did not discuss the particular needs of these groups for this guideline.

Health economics

The committee noted that there was limited and inconsistent evidence from the review of cost‑effectiveness. It was not possible to include the health conditions caused by low vitamin D status in the economic model because of insufficient effectiveness evidence. So the model focused on the risks of sunlight exposure.

The economic evidence review did not identify any studies applicable to the UK so a bespoke economic model was developed, based on the effectiveness evidence. The interventions included: an information programme for schoolchildren; photoageing; tailored messaging; text messages; and a mass media campaign. The comparator used was 'no intervention' because it was not possible to establish current practice. The outcome measures modelled were: sunburn, basal cell carcinoma, squamous cell carcinoma and malignant melanoma. The incremental cost‑effectiveness ratio (ICER) of the information programme for schoolchildren, photoageing and text messages were: £312,744, £316,968 and £65,945 per quality‑adjusted life‑year (QALY) gained, respectively. Tailored messages had an estimated ICER of £14,249 per QALY gained. The mass media campaign was cheaper and more effective than no intervention because it avoided future expenditure on treatment and the cost saving outweighed the cost of the intervention. The committee noted that the uncertainties were explored in sensitivity analyses.

A lack of sunlight exposure is associated with vitamin D deficiency. The lack of evidence on interventions aimed at delivering a complex message covering both the risks and benefits of sunlight exposure meant that the economic model could not assess the cost effectiveness of any such intervention. As a consequence, conditions associated with vitamin D deficiency are not included in the model. This is because it is not possible to quantify the impact of any of the interventions on the prevalence of vitamin D deficiency.

The committee heard evidence on the links between sunlight exposure and cataracts. But members acknowledged that the effects could not be modelled because of a lack of suitable data.

The committee discussed differences between the economic model used for this guideline and the economic model used for NICE's guideline on skin cancer prevention. The model for this guideline used the effectiveness evidence to calculate the relative risks of sunburn. In addition, it used epidemiological evidence to link the use of any kind of protection with the incidence of sunburn. This was important because several interventions showed significant reductions in the incidence of sunburn and these reductions were captured in the economic model.

It was difficult to link behavioural changes to health outcomes in the economic model because of a lack of relevant evidence. The committee discussed uncertainties about the duration of effects and how often an intervention needed to be repeated to maintain the size of effect. It also discussed whether assumptions used in the economic model to link study outcomes with health outcomes and healthier behaviours were reasonable, given the lack of evidence. However, the associated uncertainties were sufficiently explored in the sensitivity analyses.

Assuming a cost-effective threshold of £20,000 per QALY, tailored messages should cost a maximum of £5.89 per person and a mass media campaign should cost no more than £2.15 per person. Generally interventions must be cheap to be cost effective. For example, messages delivered as part of practitioners' routine practice could be cost effective.

Members noted that the information for the economic evaluation was drawn from single studies for each type of intervention.

Economic modelling

Overall, tailored messages and mass media campaigns were cost effective. Information programmes for schoolchildren, photoageing and text messaging interventions were not cost effective.

Cost‑effective estimates for the different interventions were wide ranging. The incremental cost‑effectiveness ratio (ICER) of tailored messages was £14,249 per quality of life year gained (QALY).

The mass media campaign is less costly and more effective. The ICERs of information programmes for schoolchildren, photoageing and tailored interventions ranged from £65,945 to £316,968 per QALY gained.

All input values used in the model were subject to a degree of uncertainty. Uncertainties associated with the assumptions made were explored in a range of deterministic sensitivity analyses. The one‑way sensitivity analysis revealed that the key drivers of cost‑effectiveness were the cost of implementing the intervention and its effectiveness.

The specific scenarios considered and the full results can be found in the economic modelling report.

Evidence reviews

The guideline recommendations are based on the best available evidence. Listed below are the evidence statements that provide the best available evidence and are directly linked to the recommendations. The evidence base underpinning the sun exposure messages has not been systematically reviewed for this guideline. The section on supporting information for practitioners is based on sun exposure messages from a range of authoritative sources.

Gaps in the evidence

The Public Health Advisory Committee identified a number of gaps in the evidence related to the programmes under examination based on an assessment of the evidence and expert comment. These gaps are set out below.

1. There is a lack of good quality evidence on the effectiveness of different approaches to communicating, disseminating and presenting risk information. (Source: Review 1)

2. There is a lack of good quality evidence on the effectiveness of risk communication among different subpopulations. (Source: Review 1)

3. There is a lack of evidence on how health and social care practitioners and policy makers should convey messages about the risks and benefits of sunlight exposure, particularly in the UK. (Source: Review 2)

4. There is a lack of evidence on how messages about the risks and benefits of sunlight exposure can be effectively tailored for different groups. In particular, there is a lack of evidence on tailoring messages for: people who are non‑English speaking or whose first language is not English, people from different religious or cultural backgrounds, and people with dark skin, or people who have low or no exposure to the sun. (Source: Reviews 2 and 3)

5. There is a lack of epidemiological evidence linking sunlight exposure to the incidence of cataracts. (Source: Economic modelling report 1)

6. There is a lack of evidence on interventions aimed at increasing sun‑exposure among groups at risk of low vitamin D status. (Source: Review 3)