Objective To examine patterns of energy drink consumption by children and young people, attitudes towards these drinks, and any associations with health or other outcomes.
Design Rapid evidence assessment and narrative synthesis.
Data sources 9 electronic bibliographic databases, reference lists of relevant studies and searches of the internet.
Results A total of 410 studies were located, with 46 meeting the inclusion criteria. The majority employed a cross-sectional design, involved participants aged 11–18 years, and were conducted in North America or Europe. Consumption of energy drinks by children and young people was found to be patterned by gender, with boys consuming more than girls, and also by activity levels, with the highest consumption observed in the most and least sedentary individuals. Several studies identified a strong, positive association between the use of energy drinks and higher odds of health-damaging behaviours, as well as physical health symptoms such as headaches, stomach aches, hyperactivity and insomnia. There was some evidence of a dose–response effect. 2 experimental studies involving small numbers of junior athletes demonstrated a positive impact on limited aspects of sports performance. 3 themes emerged from the qualitative studies: reasons for use; influences on use; and perceived efficacy and impact. Taste and energy-seeking were identified as key drivers, and branding and marketing were highlighted as major influences on young people's consumption choices. Awareness of possible negative effects was low.
Conclusions There is growing evidence that consumption of energy drinks is associated with a range of adverse outcomes and risk behaviours in terms of children's health and well-being. However, taste, brand loyalty and perceived positive effects combine to ensure their popularity with young consumers. More research is needed to explore the short-term and long-term impacts in all spheres, including health, behaviour and education.
Trial registration number CRD42014010192.
- child health
- systematic review
- energy drinks
This is an Open Access article distributed in accordance with the terms of the Creative Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix, adapt and build upon this work, for commercial use, provided the original work is properly cited. See: http://creativecommons.org/licenses/by/4.0/
Statistics from Altmetric.com
Strengths and limitations of this study
This is the first independent review of the scientific literature relating solely to the consumption of energy drinks by children and young people.
Key strengths include the comprehensiveness of the searches, the systematic study selection process and rigorous synthesis methods used.
The inclusion of qualitative research exploring children and young people's views, alongside quantitative studies on health and other effects, helps to enhance the relevance of the findings for the design and evaluation of future policy and practice interventions.
The strength of the conclusions is limited by the quality of the individual studies, which varied due to factors such as the sample sizes, cross-sectional designs and reliance on self-report data.
Few studies examined educational or social outcomes, highlighting a need for further research that examines the short-term and long-term impact of energy drinks in relation to a wider range of outcomes.
Energy drinks are non-alcoholic beverages that typically contain high levels of caffeine (>150 mg/L) and sugar in combination with other ingredients known to have stimulant properties. They are marketed explicitly as a way to relieve fatigue and improve mental alertness, in contrast with sports or isotonic drinks which are intended to help athletes rehydrate after exercise. There are implicit claims that energy drinks promote a more active and healthy lifestyle, in spite of the British Soft Drinks Association (BSDA) pledging that they ‘will not be marketed as sports beverages which deliver a rehydration benefit’.1 Between 2006 and 2014, consumption of energy drinks in the UK increased by 155%, from 235 to 600 million L.2 This equated to a per capita consumption of 9.4 L and a total value of £1.48 billion. Despite the growing energy drinks market and media reports of serious adverse events associated with their consumption, research into their use and effects has been sparse. In 2011, the European Food Safety Authority (EFSA) commissioned a study to gather consumption data for energy drinks in 16 countries of the European Union.3 They found that young people aged 10–18 years had the highest reported consumption prevalence (68%), compared with adults over 18 years (30%) and children under 10 years (18%). On average, young people in the UK were found to consume more energy drinks than their counterparts in other EU countries (3.1 L/month in comparison with 2.1 L).
The scientific literature focuses largely on the effects of energy drink consumption in adults, who may experience temporary benefits such as increased cognitive performance, enhanced mood, more physical energy and promotion of wakefulness.4–7 However, evidence is emerging on the harmful physiological and psychological effects of these drinks, and it is possible that prolonged use may affect physical and mental well-being.8 With children and young people, anecdotal evidence suggests that those who regularly consume energy drinks can become dependent on them and even moderate consumption may be detrimental.9–11 Based on the known effects of caffeine, consumption of energy drinks may lead to: caffeine intoxication and withdrawal; sleep disruption and insomnia; and disruptive, hyperactive and risky behaviour.12–14 There are also likely to be longer term health implications associated with excessive sugar intake, such as dental erosion, obesity and type 2 diabetes.15–18
Based on the importance and impact of energy drink consumption outlined above, the objectives of this study were to review the existing literature in order to: (1) examine evidence of associations (if any) between children and young people's consumption of energy drinks and their health and well-being, social, behavioural or educational outcomes; (2) determine whether the magnitude and direction of these associations vary according to the quantity or frequency of energy drinks consumed and (3) explore children and young people's attitudes towards energy drinks and, in particular, what factors motivate them to consume or to abstain from consuming these drinks.
We undertook a time-limited review of the literature following the guidance on rapid evidence assessments (REAs).19 REAs aim to be rigorous and explicit in method, but make concessions to breadth or depth by limiting particular aspects of the review process; for example, less exhaustive use of ‘grey’ sources and a preference for electronically available texts. Rapid or pragmatic reviews have been shown to produce similar results, more quickly and at a lower cost when compared with systematic reviews, suggesting that the approach employed here was useful and valid.20 ,21
Searches of the following major bibliographic databases were undertaken: ASSIA, CINAHL, Cochrane Library, DARE, EMBASE, ERIC, MEDLINE, PsycINFO and Web of Science. We also conducted searches of OpenGrey and the internet using Google to locate grey literature. Specific search strategies were employed for each database. See box 1 for lists of the key terms used. The results of each search were exported to an independent database using EndNote 7 software. These databases were subsequently merged into a single unique database and duplicates were removed automatically.
List 1: topic
‘Energy drink’ OR ‘stimulant drink’ OR ‘energy shot’ OR ‘energy strip’ OR ‘energy mint’
List 2: population
Child* OR adolesc* OR teen* OR (young AND (person OR people)) OR youth
List 3: outcomes
‘Health effect’ OR ‘adverse effect’ OR ‘positive effect’ OR wellbeing OR ‘physical energy’ OR wakeful* OR alert* OR ‘mental boost’ OR performance OR sleep OR insomnia OR mood OR depress* OR anxi* OR (caffeine AND (intoxication OR withdrawal)) OR dehydrat* OR headache OR nausea OR pain OR stress OR weight OR BMI OR ‘body mass index’ OR ‘metabolic rate’ OR ‘blood sugar’ OR ‘blood pressure’ OR ‘heart rate’ OR cardiovascular OR (dental AND (health OR erosion OR caries)) OR ((disruptive OR risky OR hazardous OR anti-social OR criminal) AND behavio*) OR ADHD OR ADD OR ‘attention hyperactivity deficit disorder’ OR drug OR alcohol OR smok* OR ‘screen time’ OR ‘physical activity’ OR exercise OR sport OR sedentary OR sex OR ‘self-harm’ OR violence OR injury OR sociability OR ‘peer effects’ OR learning OR memory OR attention OR attainment OR achievement OR ((absence OR exclusion) NEAR school)
We adopted an inclusive approach to locating original and review articles published between January 2000 and April 2016 that examined the consumption of energy drinks by children and young people aged 0–18 years (or up to 20 years if still in secondary education). This period was chosen to reflect the time when major brands became widely available; for example, Red Bull was introduced in the USA in 1997 and Monster was created in 2002. The inclusion of studies was not restricted according to outcome type or study setting, although we excluded studies that only involved college or university students. Animal studies, articles not published in English and studies focusing on individual ingredients (eg, caffeine or taurine) rather than energy drinks as a whole were excluded. We also excluded opinion pieces, editorials and descriptive papers without empirical findings.
Titles of studies identified from the searches were scanned by two researchers to make an initial assessment of relevance. In cases where there was any doubt, abstracts were retrieved in order to make a further judgement. Where possible, we obtained the full text of all references included after the title/abstract screening stage and PDF files were added to the bibliographic database. Articles deemed potentially relevant were reviewed independently by two researchers based on the inclusion criteria. Any disagreements (26 in total) were resolved by discussion between the researchers, with referral to a third member of the team where necessary. Information regarding the eligibility of a reference—relevant or not relevant—was recorded in the database to monitor the screening process. The process is summarised in the study selection flow chart (figure 1).
Formal appraisal of eligible studies was undertaken using the Quality Assessment Tool for Quantitative Studies developed by the Effective Public Health Practice Project (EPHPP) and the Critical Appraisal Skills Programme checklist for qualitative studies.22 ,23 Both checklists have been widely used in previous systematic reviews and allow for rapid evaluation of study quality. Each paper was independently appraised by two researchers and disagreements (four in total) were resolved through discussion to reach an overall judgement concerning the quality of the available evidence (strong, moderate or weak). Studies rated as weak were excluded from the review. The six previous reviews located through our searches were either judged to be of insufficient quality to provide reliable evidence or they included studies involving adults or non-human animals, making it difficult to extract results relating specifically to children and young people.12 ,13 ,24–27 However, they contributed to the identification of relevant primary studies.
Eligible studies were divided between members of the research team. Data were then extracted onto a template by the lead researcher for each study. Variables to be extracted included: bibliographic information; country of origin; study setting; methods; participant characteristics; outcomes; time frame; and results of the critical appraisal. This information was stored in a Microsoft Access database. As the object was to explore all possible health, behavioural, educational and social impacts of energy drink consumption, data were extracted on any outcomes and measures used in the studies.
A quantitative synthesis proved to be inappropriate due to the heterogeneity of study designs, contexts and outcomes. Data from all studies that met the inclusion and quality criteria have therefore been descriptively summarised and narratively synthesised. Narrative synthesis relies primarily on the use of text rather than statistics to ‘tell the story’ of the findings from the included studies.28 This method is often used to increase the chances of the findings of a review being used in policy and practice. In the present review, the main narratives concerned the reported effects of energy drink consumption and the experiences and attitudes of young consumers, clearly related to the stated objectives.
Study characteristics and quality
Forty-two quantitative studies and four qualitative or mixed method studies met our inclusion criteria. The quantitative studies included 31 cross-sectional surveys, four longitudinal studies, four papers reporting retrospective analyses of national or regional poison centre data and three experimental studies (tables 1⇓⇓–4). Most (n=38) explored the use of energy drinks by young people aged between 11 and 18 years. Studies were largely conducted in North America (n=22) or Europe (n=12), yet all of the qualitative studies were from Australia or New Zealand (table 5). Other study contexts included the Middle East (n=5) and South America (n=1). Equal numbers of studies were rated as being of strong or moderate quality (n=23 each).
Effects of energy drink consumption
Consumption patterns and associations with health-related behaviours
Cross-sectional survey data suggest that the use of energy drinks is patterned by gender, with several studies indicating that boys were more likely to report consumption than girls, and in greater quantities.29–46 Larson et al33 found a significant association between regular consumption and lower frequency of breakfast for girls only, while Bryant Ludden and Wolfson36 found that girls were more likely than boys to report expectations around appetite suppression. Patterns of use according to age were less clear cut, with some studies showing that consumption levels increased with age32 ,38 ,40 ,46–50 and others demonstrating that the converse was true.29 ,43 ,48 ,51 Mixed racial and ethnic patterns were also identified. Martz et al found that black students were less likely to consume energy drinks than their white or Hispanic counterparts, whereas other studies have suggested that consumption levels are highest among black, Hispanic and/or Aboriginal students.44 ,48 ,52 Higher consumption levels were positively associated with being underweight or obese, being from a single parent family, receiving free school meals, having special educational needs and higher spending money.44 ,46 ,48–50 Young people with higher academic averages, higher sense of coherence, higher levels of parental monitoring and more educated parents were less likely to consume energy drinks.14 ,41 ,48 ,53 ,54
The main health-related behaviours found to be positively and consistently associated with energy drink consumption were use of alcohol and/or binge drinking, smoking or susceptibility to smoking and other substance use.14 ,30 ,33 ,34 ,38 ,39 ,41 ,47 ,55–57 Recent longitudinal studies have found that the use of energy drinks at baseline predicted either number of drinking days or frequency of alcohol consumption at follow-up (12 or 16 months).53 ,58 Furthermore, young people who consumed alcohol mixed with energy drinks were more likely to feel and be perceived as more intoxicated, and to have greater odds of driving violations and accidents, after controlling for all other factors.59 ,60 Consumption of energy drinks has been linked to sensation-seeking,29 ,38 ,39 ,53 ,57 self-destructive behaviour,39 problems with behavioural regulation and metacognitive skills,61 and poor lifestyle behaviours,49 including regularly eating junk food or fast food.46 ,52 Al-Hazzaa et al62 found that energy drink use was associated with increased sedentary behaviour and with higher levels of physical activity. This finding was reinforced by Larson et al33 and Park et al,52 who found that energy drink use was correlated with hours spent watching TV or playing video games, and Azagba et al48 and Nowak and Jasionowski,43 who found that consumption was more likely among young people who participated in sports, suggesting a link with diverse leisure activities.
Detrimental health and other effects
Using a representative sample of Finns aged 12–18 years, Huhtinen et al31 found that daily use of energy drinks was strongly associated with four health symptoms: headache, sleeping problems, irritation and tiredness/fatigue. Those who used energy drinks several times a day were 4.5 times as likely to experience headaches and 3.5 times as likely to experience sleeping problems, in comparison with those who did not consume these drinks. However, a more recent study found that the relationship between energy drinks and health symptoms was partly mediated through going to bed late.63 A similar survey of 10–12 years in Iceland found that prevalence of physical symptoms such as headaches, stomach aches and sleeping problems generally increased with greater energy drink use for boys and girls, although the frequency of these symptoms was less common among boys on all occasions (p<0.001).42 Other cross-sectional studies have demonstrated positive correlations between energy drink consumption and sleeping problems,46 ,61 self-reported medical treatment for an injury,57 odds of sustaining a recent or lifetime traumatic brain injury41 and hyperactivity/inattention symptoms.45
The link between adverse health outcomes and ingestion of energy drinks is supported by routinely collected poison centre data from Australia and the USA.64–66 However, these are based on self-report data and the numbers are relatively small. For example, 62 children (mean age 36 months) who had accidentally ingested energy drinks were reported to the New South Wales Poisons Information Centre between 2004 and 2010.65 Fourteen had symptoms probably related to energy drink consumption—most commonly hyperactivity—and nine required assessment in hospital. In the USA, 4854 calls (0.2%) received by the National Poison Data System in 2010–2011 were for energy drink exposure cases.64 Almost half (46%) were under 6 years, but older children reported the largest proportion of moderate or major effects, such as cardiac rhythm disturbances, hypertension and hyperthermia.
Impact on sports performance
Of the three experimental studies identified through the review, two measured the effects of a commercially available energy drink on aspects of sports performance in junior athletes. Pre-exercise ingestion of an energy drink significantly improved handgrip force, running pace at high intensity and number of sprints during a simulated match among tennis players,67 and enhanced jump performance, mean leg muscle power output, perception of muscle power and perceived endurance among basketball players.68 However, it did not have any influence on mean running pace, peak running speed or ball velocity in the first study, or the precision of basketball shots, number of free throws per second or distances covered by the players in the second study. During the simulated tennis match, sweat rate was slightly higher in the experimental group, producing significantly higher dehydration (p<0.05).67
Attitudes towards energy drink consumption
Three major themes emerged through our analysis of the qualitative or mixed method studies and relevant survey results: reasons for use; influences on use and perceived efficacy and impact.
Reasons for use
Taste was consistently reported as the primary driver motivating the purchase and consumption of energy drinks, with energy-seeking emerging as an important but secondary driver.69–72 Participants involved in sports, particularly boys, reported using energy drinks as stimulants to enhance their sports performance. Others described using energy drinks as an alternative to soft drinks but only when they had enough money, as they were reported to be more expensive. Typical responses included: “Wakes you up, makes you feel alert and it tastes nice”; “It makes me go hyper” and “I drink it before soccer and I don't lose energy as fast”.72 Jones71 explored perceptions of alcohol-energy drinks (AEDs) among 12–17 years and suggested that young people liked them because they increased the ‘fun’ at parties and acted as a ‘pick me up’. They also found that the packaging of AEDs (to look like soft drinks) was a factor, particularly for younger teenagers and girls.
Influences on use
Advertising and brand loyalty emerged as major influences on young people's use of energy drinks, with participants reporting seeing them advertised on TV, the internet, through games promotions, via sports sponsorship and in shops. In a focus group study involving three age groups (16–21, 22–28 and 29–35 years), industry marketing was seen to target specific drinks at men or women, using sexualised imagery and humour.69 Participants in the youngest age group appeared to be more conscious than those in the older groups of the social image they were portraying in their choices, as well as being more sensitive to peer influences when making purchasing decisions. Social situations—and spending time with friends—provided a common context for energy drink consumption across the studies. Parents also played a key role in influencing participants' use of energy drinks, either by disapproving and prohibiting or encouraging and endorsing their use.70 It was generally agreed that energy drinks were easily accessible, from convenience stores or supermarkets, provided by parents, shared by siblings or friends, or obtained free at sponsored events.
Perceived efficacy and impact
Energy drinks were perceived to confer various beneficial effects on young people's bodies and their sports performance. Participants in one study described short-term health benefits, prevention of illness, improved immunity and a desire to rectify a poor diet.72 Others described scenarios where adults used energy drinks to effectively alleviate tiredness related to work, travel or family commitments, leading Costa et al70 to suggest that these drinks were ‘normalised and perceived as necessary to meet the demands of a busy lifestyle’ (p. 187). Few participants across the studies raised any negative or harmful effects, suggesting young people were either unaware of, or chose to ignore, the possible risks. Negative consequences associated with using AEDs were perceived to relate to the inclusion of a stimulant and depressant in one drink, and difficulties sleeping after consumption.71 In the study by Bunting et al,69 more negative connotations became apparent in the older, adult age groups (22–28 and 29–35 years), who displayed greater scepticism as to whether energy drinks were safe. However, concerns regarding sugar content emerged across all groups and moderation was stressed due to the perceived risks of overconsumption, as opposed to general consumption of energy drinks. The youngest age group (16–21 years) believed these drinks were safe as they would not be on sale if caffeine levels were too high. Findings from other studies highlight limited knowledge of the ingredients of energy drinks, particularly among younger participants.70 ,71
Summary of principal findings
This review set out to examine evidence of any associations between children and young people's health and well-being, social, behavioural and educational outcomes, and their consumption of energy drinks. It also sought to explore consumer experiences and attitudes towards these drinks.
The evidence demonstrates that the use of energy drinks by children and young people is associated with a number of adverse outcomes and health-damaging behaviours. A total of 410 studies were located, with 46 meeting our inclusion criteria. Two randomised controlled trials demonstrated that pre-exercise ingestion of an energy drink had a positive impact on some aspects of sports performance. However, both studies involved small numbers of elite junior athletes and the results should therefore be treated with some caution. Several cross-sectional studies indicated that energy drink use by children and young people was strongly and positively associated with higher rates of smoking, alcohol and other substance use, as well as being linked to physical health symptoms such as headaches, stomach aches, hyperactivity and insomnia. Two studies provided some evidence of a dose–response effect, although none of the studies was able to determine causality. Use was found to be patterned by gender, with boys consuming more energy drinks than girls, and also by age, although there was some disagreement between studies on the direction of the association. Interestingly, the highest consumption levels have been observed in sedentary individuals and in physically active individuals, suggesting a link with sport as well as screen-based leisure activities. Previous qualitative studies have reported perceived beneficial effects on young people's bodies and sports performance, with little mention of any negative effects and limited knowledge of energy drink ingredients among participants. Taste and energy-seeking were identified as key drivers for consumption. Advertising and brand loyalty have been highlighted as major influences on young people's attitudes towards energy drinks, and peers, parents and siblings were also found to play an important role.
Comparison with other studies
This is the first comprehensive review of the scientific literature to focus exclusively on evidence relating to the consumption of energy drinks by children and young people. Previous reviews have tended to examine the benefits and risks associated with specific energy drink components, such as sugar or caffeine. These studies provide important insights but fail to account for the fact that the presence of other substances such as guarana, ginseng and taurine in variable quantities may generate uncertain interactions and exacerbate any risks.73 In addition, there is an established biochemical interaction between energy drink contents and alcohol, resulting in physical and psychological side effects and increased risk-taking behaviour.74 ,75 The aforementioned EFSA study found that more than half of young energy drink consumers (53%) reported co-consumption with alcohol.3 Qualitative studies suggest that young adults use energy drinks to continue partying and drinking alcohol over a longer period, and that they may experience negative effects ranging from difficulty sleeping to being admitted to hospital.76 ,77 Further evidence is provided by the numerous clinical case reports relating to young people receiving emergency treatment for overconsumption of energy drinks with or without alcohol in recent years (for examples, see:78–82).
Consumption of sugar-sweetened beverages by children and young people has been shown to result in greater weight gain, increased body mass index and higher incidence of type 2 diabetes.18 ,83 A review on the suitability of caffeinated drinks for children found that high caffeine intakes (>5 mg/kg body weight per day) were associated with an increased risk of anxiety and withdrawal symptoms.84 However, evidence from adult studies and expert panels was used to suggest that relatively small amounts of caffeine may benefit cognitive function and sports performance, as well as contributing to daily fluid intakes. Furthermore, the author was a paid member of the Tea Advisory Panel and the work was funded by the UK Tea Council. Several other studies located through our review were funded by the soft drinks industry or conducted by researchers with acknowledged conflicts of interest. Previous independent reviews on energy drinks highlight a number of implications for children's health and well-being, although they also draw on expert opinion and adult studies.12 ,13 ,24–27 A report by the Committee on Nutrition and the Council on Sports Medicine and Fitness in the US raises concerns about the unintentional (through the use of energy drinks for rehydration) and intentional (through the use of energy drinks to combat fatigue) ingestion of potentially large amounts of caffeine and other stimulant substances.12 They suggest that paediatricians have a role to play in educating children and parents on the differences between sports and energy drinks, and to counsel that routine ingestion of sugar-sweetened beverages should be avoided or restricted.
Strengths and limitations of the study
The strengths of our review include the comprehensiveness of our searches, the systematic study selection process and rigorous synthesis methods used. The review was undertaken by a multidisciplinary team of independent academic researchers. The inclusion of qualitative research exploring children and young people's views on energy drinks, alongside quantitative studies on health and other effects associated with their consumption, helps to enhance the relevance of the findings for the design and evaluation of future policy and practice interventions.
As with any literature review, the strength of our conclusions is limited by the quality of the individual studies, which varied. The small sample sizes in the experimental and retrospective studies, reliance on self-report data in many of the observational studies and small number of qualitative studies located are all limitations of the review. Very few of the included studies examined educational or social outcomes, highlighting a need for further research that examines the short-term and long-term impact of energy drinks in relation to a wider range of outcomes. Almost half of the studies were conducted in North America and most involved high school students rather than younger children. Owing to time and resource constraints, we excluded non-English language publications and may not have identified all unpublished studies.
Conclusion and policy implications
This review adds to the growing evidence base on the health effects associated with energy drink consumption and suggests that there may be more negative than positive implications for children and young people. However, factors such as taste, brand loyalty and perceived beneficial effects help to enhance their popularity among young consumers. Consideration of the patterns and reasons for energy drink consumption identified in this review may help future interventions to ensure appropriate behaviours are targeted and are relevant to the population. Gender was identified as an important factor, in combination with gendered marketing and perceived links to sports performance, particularly among boys. The challenge for policies and interventions that seek to address this issue is to recognise the complexities of children and young people's consumption choices. Although health education targeting individuals is unlikely to achieve a substantial impact, definitive information about the safety of energy drink consumption should be provided by healthcare and other professionals, who may in turn need guidance, for example, from the UK National Institute of Health and Care Excellence (NICE). More research is needed to explore the longer-term health impacts, given that childhood and adolescence are critical yet understudied periods in the development of health-related behaviours. The potential effects of heavy and long-term energy drink consumption on child development, behaviour and educational outcomes also warrant further study.
Contributors SV, AAL, MC and DMR designed the study and obtained funding. SV wrote the review protocol and conducted the searches. SV, DMR, MC and AAL screened titles and full papers, assessed study quality, extracted data and undertook the narrative synthesis. SJC provided administrative and technical support. All authors contributed to the drafting of the paper and approved the final submitted version. The authors of this manuscript take responsibility for the integrity of the data and the accuracy of the data analysis.
Funding The review was funded as part of a larger study by The Children's Foundation (registered charity no. 1000013). AAL and MC are members of Fuse (UKCRC Centre for Translational Research in Public Health), and SV is an associate member. Funding for Fuse comes from the British Heart Foundation, Cancer Research UK, Economic and Social Research Council, Medical Research Council, and the National Institute for Health Research, under the auspices of the UK Clinical Research Collaboration, and is gratefully acknowledged (MRC grant number MR/K02325X/1).
Disclaimer This paper reports independent research commissioned and funded by The Children's Foundation Child Health Research Programme 2013–2014. The funders had no role in the study design, data collection and analysis, or preparation of any publications or dissemination materials. The views expressed in this publication are those of the authors and not necessarily those of The Children's Foundation or any other funder.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.