Article Text
PDF
Abstract
Objective Overweight and obesity constitute a new challenge in low-income and middle-countries. The obesity prevention programme, called ‘5-2-1-0’, promotes healthy eating habits, physical activity and limited screen time among young people.
This study aimed to assess adherence to the ‘5-2-1-0’ recommendations and to study multiple risky behaviours among adolescents in nine countries in sub-Saharan Africa.
Methods Meta-analyses with a random effect were used to calculate overall prevalence. Non-random patterns of the co-occurrence of the four risky behaviours were explored using observed/expected prevalence ratios. Data came from the Global School-based Health Survey and 18 314 adolescents were considered.
Results Among the participants, 12.7% (95% CI 7.5% to 19.0%) had overweight and 3.2% (95% CI 1.1% to 6.1%) had obesity. In almost all countries studied, girls were more affected by overweight and obesity than boys.
While only 0.2% (95% CI 0.1% to 0.4%) of the adolescents fully complied with the recommendations, 4.8% (95% CI 3.1% to 6.9%), 28.4% (95% CI 22.4% to 34.8%), 43.8% (95% CI 41.9% to 45.8%) and 17.0% (95% CI 11.8% to 23.0%), respectively, combined 1, 2, 3 and 4 risky behaviours among the four ‘5-2-1-0’ criteria. The most observed combination was found for co-occurrence of three risky behaviours: insufficient fruit/vegetables consumption, physical activity and non-zero consumption of sugar-sweetened beverages.
Conclusion In conclusion, the insufficient adherence to ‘5-2-1-0’ recommendations and the high prevalence of the co-occurrence of risky behaviours underscore the need to strengthen health interventions and programmes to prevent obesity among adolescents in sub-Saharan Africa.
- Dietary patterns
- Malnutrition
- Preventive counselling
- Weight management
Data availability statement
Data are available in a public, open access repository. Data are available on the WHO website.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Adherence to ‘5-2-1-0’ guidelines decreases children’s risk of obesity and related morbidities in later life.
Adherence to ‘5-2-1-0’ guidelines in sub-Saharan Africa is unknown.
WHAT THIS STUDY ADDS
More female adolescents were at risk of having overweight and obesity than males in sub-Saharan Africa.
Adherence to the ‘5-2-1-0’ recommendations among adolescents was low and risky behaviours were present simultaneously.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Health interventions need to be strengthened among African adolescents considering the ‘5-2-1-0’ guidelines.
The co-occurrence of risky behaviours needs to be considered in prevention programmes.
Introduction
In recent years, prevalence rates of overweight and obesity have sharply increased in low-income and middle-income countries, constituting a new challenge in terms of prevention strategies,1 a problem that coincides with existing problems of undernutrition. About 15% of adolescents in Africa have overweight or obesity.2 3 As childhood obesity has been significantly associated with cardiovascular and metabolic diseases in adulthood,4 representing a major public health problem that must be addressed to avoid multiple comorbidities in adulthood.5 It also has been shown that prevention programmes give better results in younger people, particularly when they are based on individual behaviour but also on the community aspect, and therefore, the environment.6
Result of genetic, socioeconomic but also cultural factors, obesity can also be partly explained by poor lifestyle habits such as poor diet or a low level of physical activity.7 Therefore, it is important to understand and analyse the risky behaviours associated with obesity to reduce the existing disparities among adolescents. For instance, previous studies have shown that the prevalence of obesity is lower in adolescents with lower screen time.8 Other studies have shown that high consumption of sugar-sweetened beverages leads to overweight or obesity.9 10 Furthermore, a healthy diet and sufficient physical activity have also been shown to be particularly protective factors against overweight or obesity among adolescents.11
Since its creation in 2006 by the Maine Youth Overweight Collaborative in the USA,12 the obesity prevention programme called ‘Let’s go! 5-2-1-0’ has been adopted as the main message in several prevention campaigns.13 This programme is made up of the following four criteria: 5—consume at least five fruits and vegetables per day; 2—limit recreational screen time to a maximum of 2 hours per day; 1—do at least 1 hour of physical activity per day; 0—drink zero sugar-sweetened beverages. These recommendations are also part of the recommendations of the WHO in the fight against obesity.3 In 2019, during the 60 s World Health Assembly, during high-level meetings of the United Nations General Assembly on health-related issues, these recommendations were included in the third objective. These objectives were aimed at disseminating advocacy campaigns and developing communication supports. The objectives promoted a good consumption of fruits and vegetables, a reduction in sedentary behaviour and awareness of the risks of excessive sugars intakes.14
The promotion of physical activity was already part of the first global action plan created by the WHO in 2018 which planned to reduce physical inactivity by 15% by 2030.15 In its latest publication of the plan ‘Ending Childhood Obesity’, the WHO has also defined six areas of action to fight obesity, including the promotion of healthy diets and regular physical activity, including in schools.16
These recommendations must also be monitored to avoid other health problems such as psychological distress,17 cardiovascular diseases, etc. For example, adolescents are the most sedentary paediatric population, spending more than half of their time engaging in sedentary activities after school.18 Another beneficial indicator for mental health, a good consumption of fruits and vegetables, as suggested by criterion 5 of the ‘5-2-1-0’ programme, is also associated with a so-called cardio protective effect, thus limiting the risks of cardiovascular diseases.19 The screen time, associated with a strong sedentary lifestyle, is also associated with a decrease in the perceived quality of life among adolescents.20
As the first generation to have access to high computer technology,21 African adolescents are increasingly exposed to significant screen time, and therefore, to an increased risk of health problems. Likewise, higher consumption of sugar-sweetened beverages has also been observed in recent years in certain regions of Africa.22 This higher consumption is partly linked to an expansion of the urbanisation.23
Several studies have shown that these risky behaviours are often present simultaneously24 and increase the risk of overweight and obesity.25 26 However, despite the existence of these results, very few studies have comprehensively studied this public health question taking into account data from several African countries.27 28
The objectives of this study were to assess the proportion of individuals who met the ‘5-2-1-0’ criteria and the prevalence of co-occurrence of risky behaviours among adolescents from nine countries in sub-Saharan Africa (Benin, Ghana, Liberia, Mauritania, Mauritius, Mozambique, Namibia, Seychelles and Tanzania).
Materials and methods
Data
The study used the Global School-based Student Health Survey (GSHS) database, set up by the WHO and the Centers for Disease Control and Prevention. This survey aims to establish programmes, health policies and preventive interventions to be used and evaluated, and health promotion in the school environment.29 30
Since 2003, the GSHS survey has been carried out in eighteen countries in sub-Saharan Africa. Among these countries, only 9 (50.0%) gathered all necessary information to study the ‘5-2-1-0’ recommendations (online supplemental table A1) and only 6 gathered all necessary information for a study including both overweight and obesity. In countries where several studies carried out over different years were available; the most recent version was then chosen.
Supplemental material
Data from 2010 and 2017 for the following surveys were used: Benin (2016), Ghana (‘Junior’ and ‘Senior’, 2012), Liberia (2017), Mauritania (2010), Mauritius (2017), Mozambique (2015), Namibia (2013), Seychelles (2015), Tanzania (2014). Two data tables were available for Ghana, one corresponding to ‘Junior’ type establishments (the youngest) and the other to ‘senior’ type establishments.
In each country, secondary schools were selected with a probability proportional to the number of enrolments in each establishment. Then the classes were selected at random. All students of the selected classes were eligible to participate. For each participant, the design variables (primary sampling unit and ‘strata’ variables) and associated weights, calculated by the WHO, were provided and used in the analysis.
For each country, the GSHS survey has been approved by a government administration (most often the Ministry of Health or Education) and an institutional review board or ethics committee. Participants and parents, or legal guardians, had to give a verbal or written consent.
Measures
Self-administered questionnaires were distributed to participating adolescents in each country and several questions on sociodemographic characteristics, lifestyle habits, family and social environment were asked.
Sociodemographic data were age, sex and socioeconomic status. The socioeconomic status was obtained according to the level of hunger the individual reported, as has been done in other studies,31–33 using the following question: ‘Over the last 30 days, how often did you go hungry because there was not enough food in your home?’. Socioeconomic status was considered high if there was never or rarely a feeling of hunger and low if the answer to the question was sometimes, most of the time or always, as already done in previous studies.33
Height and weight of adolescents were self-reported in the questionnaire. Overweight and obesity were determined using WHO’s growth reference curves. An adolescent was considered as having overweight according to the following criterion: ‘>+1 SD from the median for body mass index (BMI) by age and sex’ and had obesity if the criterion ‘>+2 SD from the median for BMI by age and sex’ was observed. The calculations for the reference curves were made by the WHO software AnthroPlus34 and only the associated categories (overweight or not, obesity or not) were given.
To obtain the data associated with the ‘5-2-1-0’ criteria, the same questions were used in each of the countries studied, which allowed making cross-country comparisons. Fruit and vegetable consumption were estimated through the following questions: ‘During the past 30 days, how many times per day did you usually eat fruit?’ and ‘During the past 30 days, how many times per day did you usually eat vegetables?’. The variable related to criterion 5 was thus separated into two categories: ‘at least five fruits and vegetables per day’ and ‘less than five fruits and vegetables per day’. Screen time was obtained by answering the following question: ‘How much time do you spend during a typical or usual day sitting and watching television, computer games, or doing other sitting activities?’. Each country took as an example in the question, in addition to screen time, activities such as chess, card games, etc. Two categories were provided: less than 2 hours per day/more than 2 hours per day.
For criterion 1, related to physical activity, the question ‘During the past 7 days, on how many days were you physically active for a total of at least sixty minutes per day?’ was asked. Only respondents indicating ‘7 days’ were considered as having done sufficient physical activity. For criterion 0, related to the consumption of sugar-sweetened beverages, two categories were also created (no consumption/consumption) based on the following question: ‘During the past thirty days, how many times per day did you usually drink carbonated soft drinks?’. If an adolescent answered ‘I did not drink carbonated soft drinks during the past 30 days’, the category considered was ‘no consumption’. Otherwise, if he or she answered ‘less than one time per day’ or ‘one time per day’ or ‘three times per day’, the category considered was ‘consumption’.
For this study, a ‘5-2-1-0’ score from 0 to 4 was calculated to assess the number of cumulative risky behaviours in adolescents. An adolescent obtained a maximal score of 4 if he or she did not meet any recommendation corresponding with the presence of four risky behaviours. On the contrary, an adolescent respecting the four recommendations had a final score of 0 since he did not accumulate any risky behaviour. Each possible combination according to the total number of risky behaviours could then be examined.
Statistical analysis
All the overall prevalence rates, including sociodemographic characteristics, overweight and obesity or criteria 5-2-1-0, for the selected countries were obtained by a meta-analysis with a random effect using the ‘meta’ and ‘metaphor’ packaged, V.4.19.0 and V.3.0.2, respectively, in R software (V.4.0.3).
To determine the prevalence by country of the various sociodemographic characteristics, outcomes and criteria of the ‘5-2-1-0’ model, the weights and other design factors were considered using the ‘survey’ package, V.4.0. Differences of this prevalence according to gender were tested for each country using a χ2 test.
To determine the co-occurrence of these four risky behaviours, a comparison of the observed/expected ratio for each possible score combination was made. The expected prevalence rates were obtained by multiplying the probabilities of each risky behaviour defined in the combination according to its distribution in the population studied. Thus, it was possible to identify the combinations with higher prevalence (ratio>1) or lower (ratio<1) than the expected prevalence.
Results
Sociodemographic characteristics
This study included 26 769 adolescents. The mean age was 15.3 years (SD: ±0.06). Overall, 49.9% of the adolescents were boys, 51.3% were fifteen or younger and 38.3% had a low socioeconomic status (table 1).
Survey characteristics of GSHS according to country
More specifically, the mean age was higher in Benin, Ghana (data from ‘Senior’) and Liberia (16.5 (SD: ±0.11), 17.2 (SD: ±0.10) and 16.5 (SD: ±0.13) years, respectively). On the contrary, the youngest adolescents (from ‘Junior’ type establishments) in Ghana, Mauritius, Seychelles and Tanzania were representative of younger adolescents with a mean age of less than 15 years. All the countries had a fairly balanced distribution of boys and girls, except for Benin where the proportion of adolescent boys was higher (73.0%). Finally, more than 60% of the adolescents from Ghana and Liberia had a low socioeconomic status, while this proportion was lower in Namibia, Benin, Mauritania, Mauritius, Seychelles, Mozambique and Tanzania.
Relevance of overweightness and obesity
In the 6 countries for which information was available, the prevalence of overweight and obesity was 12.7% (95% CI 7.5% to 19.0%) and 3.2% (95% CI 1.1% to 6.1%), respectively (table 2). Mauritania, Mauritius and Seychelles were particularly affected by overweight, with more than one in five adolescents being overweight. Mauritius and the Seychelles were also the most affected countries by obesity with around 1 in 10 adolescents.
Prevalence of overweight and obesity among adolescents in seven countries of sub-Saharan Africa
Among boys, respectively 9.3% (95% CI 3.8% to 16.7%) and 2.5% (95% CI 0.5% to 5.8%) of adolescents had overweight and obesity. These prevalence rates were higher among girls (overweight: 16.2% (95% CI 11.2% to 21.9%) and obesity: 3.8% (95% CI 1.8% to 6.6%)). In all countries excepted Mauritius, boys were less affected by overweight and obesity than girls.
‘5-2-1-0’ recommendations adherence
Across all countries, more than 4 in 5 adolescents (88.4%) had inadequate fruits and vegetable intake, that is, less than 5 fruits and vegetables per day (table 3). This proportion varied from 81.4% in Seychelles to 93.9% in Mozambique. Less than 1 in 3 adolescents (31.1%) spent more than 2 hours in front of the screens with considerable variation across countries (from 18.7% in Ghana Junior to 51.0% in Seychelles). Physical inactivity was high in all countries, with 4 out of 5 (82.0%) adolescents having less than 1 hour of physical activity per day. Benin was the country with the lowest prevalence (70.4%) while Ghana (‘Junior’) had the highest proportion of ‘inactive’ adolescents (89.9%). Finally, 73.3% of the adolescents had consumed sugar-sweetened beverages during the past 30 days. This prevalence was much lower in Ghana where 58.5% of adolescents in senior institutions consumed sugary drinks and the highest in Seychelles (91.7%).
Non-adherence with the ‘5-2-1-0’ recommendations among adolescents in nine countries of sub-Saharan Africa
In addition, a significant difference regarding the insufficient consumption of fruits and vegetables between adolescent boys and girls was found in Namibia and Tanzania (table 3). In Namibia, girls were more concerned by this insufficient consumption while in Tanzania, the reverse was observed. Regarding the excessive screen time, only the Seychelles showed a significant difference between genders, with girls being more concerned. Boys were also more likely to practice sufficient physical activity than girls, which was observed in all countries. Finally, adolescent boys were less likely to consume sugar-sweetened beverages than girls, and this was observed in all countries except Mauritania and Mauritius.
Co-occurrence of non-adherence to the ‘5-2-1-0’ recommendations
Overall, 17.0% (95% CI 11.8% to 23.0%) of the adolescents combined the four risky behaviours among the ‘5-2-1-0’ criteria and thus did not consume enough fruits and vegetables, had a too high screen time, insufficient physical activity and consumed sugar-sweetened beverages. About 43.8% (95% CI 41.9% to 45.8%) of the adolescents engaged in 3 risky behaviours among the 4 criteria, while 28.4% (95% CI 22.4% to 34.8%) engaged in 2 and 4.8% (95% CI 3.1% to 6.9%) engaged in 1 risky behaviour. Only 0.2% (95% CI 0.1% to 0.4%) of the adolescents met the 4 ‘5-2-1-0’ criteria, and therefore, did not engage in any risky behaviour.
Again, girls were less likely to adhere to the recommendations. Indeed, while only 0.1%, 4.0% and 26.2% of the adolescent girls had 0, 1 and 2 risky behaviours accumulated, respectively, these proportions were 0.3%, 5.6% and 30.6% among adolescent boys (table 4). On the contrary, girls were more likely to accumulate 3 or 4 risky behaviours, compared with boys (45.5% and 18.7% among girls vs 42.2% and 15.2% among boys).
Observed and expected values for combinations of 5210 risky behaviours among adolescents in 9 sub-Saharan African countries
For both girls and boys, the most common combination, affecting more than 1 in 3 adolescents (respectively, 37.7 and 33.3%), was the combination of three risky behaviours: low consumption of fruits and vegetables, and insufficient physical activity, and high consumption of sugar-sweetened beverages.
By looking more precisely at the differences between the observed and expected prevalence rate’s in girls and boys, we first notice that the observed number of adolescents respecting all the recommendations (ie, having a score of 0) was lower than expected for either boys or girls. On the contrary, the observed prevalence associated with the highest score (ie, four risky behaviours) were higher than those expected for both girls and boys. The highest observed/expected ratio, respectively, equal to 2.30 for boys and 2.03 for girls, was found for a total score of 2 risky behaviours: high screen time and sugar-sweetened beverages consumption. On the contrary, the lowest observed/expected ratio, equal to respectively 0.45 for boys and 0.47 for girls, was associated with a total score of 2, with the following risky behaviours: high screen time and insufficient physical activity.
Discussion
The main purpose of this study was to assess the prevalence of overweight and obesity, to investigate the adherence to ‘5-2-1-0’ guidelines and to evaluate the co-occurrence of the risky behaviours related to these recommendations among adolescents in Africa. After an initial selection of GSHS data available on the WHO website, nine countries were selected for performing these analyses.
Among the adolescents included in these studies, which were carried out between 2010 and 2017, obesity particularly affected young people in Mauritius and Seychelles where 1 in 10 adolescents had obesity. One in four adolescents had overweight in these two countries.
The high prevalence of obesity in Seychelles in this study is in line with previous findings. Indeed, Seychelles is the country in sub-Saharan Africa that is most affected by overweight and obesity problems.35 Aware of this important public health issue, Seychelles has recently created programmes to fight against the poor eating habits of the population, in part linked to the high consumption of take-out meals.35 Mauritius, another island country in sub-Saharan Africa, was also strongly affected by overweight problems, both among the children and the adults.36 The country has, therefore, developed a national action plan for nutrition37 and other preventive programmes to control the high prevalence of overweight and obesity.38
In Mauritania, which is less affected by obesity, a high prevalence of overweight was found among young girls, with around 3 in 10 adolescent girls being overweight. As a direct consequence of the country’s Leblouh tradition, young girls are sometimes force-fed to meet cultural criteria and take on the roles of wife and mother as quickly as possible.39 In the other countries studied (Benin, Ghana and Namibia), the prevalence of obesity was quite low, but overweight nevertheless affected between 6.6% and 8.4% of adolescents. In all of the countries studied, girls were significantly more affected by overweight or obesity than boys, which is consistent with results from other African countries.40 Hypothesis mentioned by the authors were differences in physiology, lifestyle habits such as physical activity,41 or different sex chromosomes.42 It is, therefore, essential to take these differences into account in health interventions aiming at fighting obesity.
Looking at adherence to the ‘5-2-1-0’ recommendations, more than four in five adolescents did not consume at least five fruits and vegetables per day, which is consistent with various studies targeting adolescents from other African countries43 44 but other regions in the world as well.22 45 Some countries, such as Mozambique, the Seychelles and Mauritius, were also particularly affected by high consumption of sugar-sweetened beverages among adolescents, with more than four in five adolescents regularly consuming a sugary drink. This high consumption has also been observed in other studies.22 46 In order to combat this scourge, some countries such as South Africa have recently implemented a tax on sugary drinks leading to a drop in consumption, especially among individuals with lower incomes.47
Screen time was the only criterion in the ‘5-2-1-0’ recommendations that was fairly well respected, with approximately less than one in two adolescents having too much screen time. These criteria are also respected in some regions of the world, such as China, for example,48 but on the contrary is worrying in other countries such as the USA, where adolescents spend on average more than 7 hours a day in front of the screens.49 Finally, lack of physical activity among adolescents was found in all countries studied with a significant difference between girls and boys, in favour of boys. Four in five adolescent girls were active less than 1 hour per day, which was deemed insufficient. Since this gender difference was also shown in other studies,50 adolescent girls are an important target for promoting physical activity. In our study, this gender difference was also found for the three other criteria, and therefore, needs to be taken into consideration.
By analysing the co-occurrence of these risky behaviours, only 0.2% of adolescents met the ‘5-2-1-0’ criteria at the same time, that is, a good consumption of fruits and vegetables, sufficient physical activity, limited screen time and a good consumption of sugar-sweetened beverages. On the contrary, three in five adolescents accumulated three or four risky behaviours. The most observed combination of these risky behaviours was: low consumption of fruits and vegetables, insufficient physical activity and high consumption of sugar-sweetened beverages, and affected more than one in three adolescents. This study then showed that few adolescents respected the ‘5-2-1-0’ recommendations, suggesting that more intensive efforts in terms of nutrition, screen time and physical activity are needed.
These efforts are not only to be made on behaviours but also on the environment. Indeed, these risky behaviours are obviously dependent on the environment in which the adolescent grows up. It is then necessary to make the necessary modifications in order to change the behavioural habits of this population. These actions are notably the subject of the ‘Taking Action on Childhood Obesity’ campaign led by the WHO, suggesting, for example: to improve the environments in which children live, play and learn, to implement policies to support healthy food environments, for mothers, infants and children, to increase policy priority to ensure safe and accessible environments for physical activity and to strengthen the measurement of food and physical activity environments and policy implementation.51
Results should be interpreted in light of some limitations. First, the main limitation is the cross-sectional design of the study, which hampers the possibility to establish causal links. Second, our study relies on self-reported measures, which may be susceptible to various error biases, such as desirability or memory bias. Moreover, some variables were not directly available. For example, the screen time question included ‘other sitting activities’ and is, therefore, not restricted to screen time only. The answer given by the adolescents to the screen time question also needs careful interpretation since access to a smartphone or a computer device may vary across participants. In addition, physical activity was measured over the last 7 days, and therefore, does not allow more general monitoring of the situation. The same limitation also applies to the measurement of sugar-sweetened beverage consumption, for which adolescents who rarely consume this type of drink may have hesitated between ‘less than one time per day’ and ‘I did not drink carbonated soft drinks during the past 30 days’ propositions. Finally, caution is needed in extrapolating these results to national levels since the data were only collected among schooled adolescents.
Despite these limitations, our study is the first to focus on adherence to the ‘5-2-1-0’ recommendations in nine countries of sub-Saharan Africa. By studying the prevalence of each of these criteria as well as obesity and overweight, our study provides new insights into the health condition of adolescents in these countries. Additionally, the study was conducted on a large sample with high response rates.
Conclusion
This study was the first to assess the prevalence rates of overweight and obesity among adolescents simultaneously in nine countries of sub-Saharan Africa. In addition to the high prevalence of obesity and overweight, this study showed that the adherence to the ‘5-2-1-0’ recommendations was insufficient, especially among adolescent girls, and provided new insights in the co-occurrence of several risky behaviours. These results emphasise the need to strengthen health interventions and programmes, such as the ‘5-2-1-0’ recommendations, to promote healthy behaviours and prevent obesity and overweight among African adolescents.
Data availability statement
Data are available in a public, open access repository. Data are available on the WHO website.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by CDC. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We are grateful to the WHO and its partners for the free availability and data collection. We also thank the students who made this study possible.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors Conceptualisation, FH, SS and FK-S; methodology, FH, SS and FK-S; software, FH; validation, SS, AD, IS, PD, JVdH and FK-S; formal analysis, FH and SS; data curation, FH; writing—original draft preparation, FH; writing—review and editing, SS, AD, IS, PD, JVdH and FK-S; visualisation, FH, SS, AD, IS, PD, JVdH and FK-S; supervision, SS and FK-S; guarantor, FK-S. All authors have read and agreed to the published version of the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.