Article Text
PDF
Abstract
Objectives This study was conducted to determine the status of and factors associated with behavioural determinants (knowledge, attitude and practice) of sugary foods and beverages among adolescents in Kathmandu, Nepal.
Design Cross-sectional study.
Setting The study was done at educational institutions located in Nagarjun municipality, Kathmandu, Nepal.
Participants In this study, 768 adolescents from grades 8, 9 and 10 were selected using a multistage random sampling technique.
Primary and secondary outcome measures Knowledge regarding sugary foods and beverages was assessed using 9 questions, attitude was measured through 13 Likert scale statements and consumption was assessed using 9 questions through a semi-structured questionnaire. Body mass index (BMI) was calculated based on anthropometric measurements. Pearson’s χ2 test and logistic regression were used to assess factors associated with those behavioural determinants, along with their correlates with BMI. A p value of <0.05 was considered statistically significant.
Result Adequate level of knowledge was found among 84.11% (95% CI: 81.52% to 86.70%) of the adolescents. Awareness of the health risks and inclination to adopt healthier behaviours was seen among 60% of the adolescents (95% CI: 56.55% to 63.49%). The percentage of adolescents consuming sugary foods and beverage items was 84.50% (95% CI: 81.94% to 87.07%). The odds of having adequate knowledge among respondents was twice (adjusted OR (AOR)=2.05, 95% CI: 1.12 to 3.76) more likely for those who were living with their parents. Female adolescents (AOR=2.51, 95% CI: 1.61 to 3.89), whose mothers are homemakers (AOR=1.99, 95% CI: 1.04 to 3.58) and fathers are engaged in foreign employment (AOR=2.09, 95% CI: 1.04 to 4.21), were more likely to consume sugary items. Prevalence of overweight and obesity was seen among 6.38% (95% CI: 4.64% to 8.11%) of respondents. Consumption was seen to be significant to the model overweight/obesity versus normal (OR=11.95 (95% CI: 1.61 to 88.42)).
Conclusion Adequate knowledge alone was insufficient for influencing food selection and choices. Family-indulged interventions can be useful as familial factors seem to be impacting behavioural characteristics. Sugary foods and beverages consumption was linked to being overweight/obese, highlighting the importance of dietary discipline in reducing this risk.
- Adolescent
- Body Mass Index
- Behaviour
Data availability statement
Data are available upon reasonable request. All data analysed during this study are available from the corresponding author on reasonable request.
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
STRENGTHS AND LIMITATIONS OF THIS STUDY
This research focuses only on sugary foods and beverages filling a gap in research not previously explored in developing nations such as Nepal.
The sampling method accounts for the design effect for clustering and potential non-response rates to ensure proper statistical power.
The findings might not be generalisable at the national level but can be compared with big urban cities.
The study findings relied on self-reporting, allowing for under-reporting and recall bias.
Introduction
Added sugars are not essential to a healthy diet, even though they provide sensory effects that enhance meals and increase enjoyment. They can replace nutrient-dense meals and contribute to poor health outcomes by supplying calories without delivering other necessary nutrients, which is especially concerning for adolescents and young individuals.1 Excessive sugar consumption has been associated with various metabolic irregularities and harmful health effects, including a heightened risk of non-communicable diseases (NCDs) and nutritional deficiencies in individuals.2 Unhealthy diets are one of the most significant and heavily linked modifiable risk factors for NCDs.3
Evidence has shown that the lifestyle, dietary habits and behavioural patterns formed during adolescence can remain into adulthood and have a substantial impact on long-term health.4 5 Childhood and adolescent obesity raises the chance that an adult may continue to weigh more than what is healthy in the latter phases of life as well.4 5 Excessive consumption of sugar in various forms, most especially the sugary food items and beverages, has been one of the crucial factor for promoting overweight and obesity all around, including lower middle-income countries like Nepal.6 7 Adolescence is the phase of life between childhood and adulthood, from ages 10 to 19. It is a unique stage of human development and an important time for laying the foundations of good health.8 The nutrition and food habits during this phase are extremely important as they determine the health in the adult stage of life, and poor dietary habits are among the most prevalent risk factors for chronic illness in children and adolescents aged 12–17 years.9 Childhood exposure to sugary foods and beverages can influence an adult’s liking for sweet meals so an early-life nutrition is crucial for a child’s development as well as being a modifiable risk factor for NCDs.10–12 Since adolescence is a period that demands high nutrients with changes in dietary habits among adolescents, it makes the age group more favourable for getting exposed to the consumption of heavily processed and high-sugar products.13–15
Ultra-processed foods and non-alcoholic beverages with high amount of trans-fat, free sugar or salt are being heavily consumed by millions of children and adolescents worldwide, including in Nepal, leading to harmful effects on their health and development.16–18 Energy-dense foods and sugary drinks are significant contributors to weight gain, overweight and obesity by promoting excess energy intake.6 The prevalence of overweight has tripled in many countries, including Nepal, making it a major public health issue in the 21st century.19 Sugar-sweetened beverages have been found to be a unique dietary contributor to obesity.20 The shift toward increased consumption of sugary products is particularly noticeable among urban adolescents, especially in regions like the Kathmandu Valley.21 A typical Nepali consumes 30 g of sugar daily, with urban areas having substantially greater consumption rates, while the threshold safe limit for sugar consumption is 25 g/day.22 Sugary foods and beverages are a key dietary contributor to obesity, and projections from NCD Risk Factor Collaboration indicate that 6% of children and adolescents aged 5–19 globally, including in Nepal, are expected to be obese by 2030.23 A 2015 survey revealed that one-third of Nepalese adolescents regularly consumed carbonated drinks, and processed food consumption among the adolescents in Kathmandu valley has been rising abruptly.24 25 The consumption of such foods has increased by fivefolds in the past three decades because of their luring taste, convenience, low cost, range of options and taste.26 Multiple factors, including personal traits, physical and societal spheres, affect the dietary choices in Nepal.27 28 Behavioural determinants such as attitudes, beliefs and knowledge ultimately guide the practice of consuming sugary foods and beverages.29 30 The increasing availability and affordability of sugary and processed foods in urban areas of Nepal have exacerbated these trends. Therefore, this study aims to assess and understand the behavioural determinants such as awareness and consumption patterns of sugary foods and beverages among adolescents in Nepal, specifically within the context of these growing public health concerns.
Methods
Study design and setting
A cross-sectional study was conducted among school-going adolescents from August 2022 to February 2023. The study was conducted in 10 randomly selected schools (5 private and 5 public) from a pool of 58 schools in Nagarjun municipality, Kathmandu District. Nagarjun municipality, located in the northwestern part of Kathmandu Valley, is a rapidly urbanising area and the second most populated administrative unit in the district. The adolescent population in this municipality is 18 647, accounting for 16.15% of the total population.31
Sample size determination and sampling technique
The sample size was calculated using the single population proportion formula, 
. Considering 95% CI, 50% assumed proportion32 (p) and 5% margin of error (d). Assuming a 10% non-response rate and a two design effect, the calculated sample size was 768. A multistage random sampling method was used to select the participants. Initially, five random wards were selected from ten wards of Nagarjun municipality. There were 58 schools (13 public and 45 private) with a total of 4191 students in the sampling frame across these five wards. A complete list of all schools inside Nagarjun municipality was obtained from the education section of the municipality. Schools with less than total of 50 students across grades 8, 9 and 10 were excluded from the study. This was done to ensure that each selected schools could contribute an adequate number of participants for meaningful statistical analysis, aiming for robust results. Eventually, 34 schools remained in the sampling frame, consisting of 10 public schools and 24 private schools with a total of 3829 students. Finally, two schools (one public and one private) were randomly chosen from each of the five selected wards. All eligible students from grades 8, 9 and 10 in these schools were included in the study. Primary schools and adolescents who were absent on the days of data collection were excluded from the study.
Questionnaire design
A semi-structured questionnaire was developed for the study based on its objectives, following a thorough and vigorous literature review, referencing tools from similar studies done to assess knowledge, attitude and practices alongside factors associated with those variables.33–35 The development included consultations with experienced faculty members in public health. The tool was initially developed in English, translated into Nepali and back-translated to the English language to ensure that the originality of the questionnaire remains unchanged for accuracy. Linguistic and contextual validation were checked and recommended through expert consultations from academicians and managers working for more than a decade in the field of nutrition prior to pretesting. The questionnaire comprised several sections, including socio-demographic characteristics, parental characteristics, questions to assess knowledge regarding sugary foods and beverages and related health impacts (9 questions), questions to assess attitude towards sugary foods and beverages (13 questions/statements) and questions to evaluate the consumption of sugary foods and beverages (9 questions).
Evaluation of knowledge and practice was done by assigning a score of ‘1’ for positive/correct answers and ‘0’ for negative/incorrect answers. Eleven items related to sugary foods and beverages alongside ill effects accompanied by an inclination to adopt healthier behaviours were used to assess attitudes. The responses were based on a Likert scale with three possible levels, ranging from agree to disagree. The evaluation of attitudes was done by assigning a score of ‘2’ for positive/correct statements, ‘1’ for neutral responses and ‘0’ for negative/incorrect statements. Participants who scored at or above the median of the total score were considered to have adequate knowledge and a positive attitude. A score of 5 or higher was considered an adequate level of knowledge, while respondents’ scoring 19 or above were deemed to have a good attitude, based on a total attainable score of 22 and referenced to the median score from the Shapiro-Wilk test (p<0.05). A pilot study of the questionnaire was conducted in one non-sampled school, and the results indicated it was appropriate for the main study. A copy of the study questionnaire can be found in online supplemental file 1.
Supplemental material
Data collection and statistical analysis
Self-administered questionnaires were used for data collection from September 2022 to October 2022. Anthropometric measurements (height and weight) were obtained using calibrated digital bathroom weighing scales and a standard tape measure. The accuracy of the instruments was checked using the standard weight and height at the beginning of every data collection session. Statistical analysis was performed using SPSS V.16. The association between independent variables and the dependent variables (knowledge, attitude and practice) and outcome variables [body mass index (BMI)] was assessed through χ2 tests and logistic regression analysis. The Hosmer-Lemeshow test was performed to test the goodness-of-fit of the multivariate logistic regression model, and the model was found to be a good fit (p<0.05).
Public and patient involvement
None. There was no patient and/or public involvement in the design, or reporting, or dissemination plans of the study.
Ethical consideration
All the study-related activities were initiated after obtaining the relevant approval from the Institutional Review Committee of Manmohan Memorial Institute of Health Sciences. Prior permission and written consent were also taken before the data collection from the education section of the municipality and school administration. The principals of the respective institutions informed the parents of the adolescents and obtained their verbal consent. Written assent was secured from each participating adolescent before data collection, and they were clearly informed regarding the study prior to data collection. Confidentiality of information was assured and ensured throughout the study. The Strengthening the Reporting of Observational Studies in Epidemiology cross-sectional reporting guidelines36 was also used to ensure the quality of reporting and transparency.
Results
Demographic characteristics of the study participants are presented in table 1. The median (±IQR) age of adolescents was 15±1 years, ranging between 11 and 18 years of age. More than half of the participants were aged less than 14 (52.5%), male (51.4%) and Janajati (50.1%). Over one-third of the mothers (34.4%) had attained a secondary level of education, and 41.6% were homemakers.
Socio-demographic characteristics of study participants
More than four-fifths (81.4%) of the respondents considered sugary foods and beverages to be harmful to health. The prevalence of adequate knowledge regarding sugary foods and beverages and their ill health effects among adolescents was 84.11% (95% CI: 81.52% to 86.70%) as shown in online supplemental table A. Exactly three-fifths (95% CI: 56.55 to 63.49) of the respondents held a positive attitude, indicating awareness of the harmfulness and ill effects of sugary foods and beverages, accompanied by an inclination to adopt healthier behaviours. It is evident through online supplemental table B that nearly four-fifths of the adolescents stated that their parents disapprove of them consuming such foods and beverages. The rate of consumption of sugary foods and beverages among adolescents was 84.50% (95% CI: 81.94% to 87.07%). Among those consuming sugary foods and beverages, the highest proportion (68.1%) reported consuming confectionaries and bakeries like ice cream, ice pops, cakes and pastries, followed by cold sugary beverages (66.4%), including sugar-sweetened beverages, carbonated/aerated drinks, artificial fruit juices and energy drinks, as shown in online supplemental table C. More than four-fifths (81.8%) of respondents felt the need to control and limit the intake of sugary foods and beverages, while just slightly more than three-fifths (62.5%) of them were confident they could discontinue consumption if necessary (figure 1).
Supplemental material
Attitude towards sugary foods and beverages alongside its effects. This is the radar plot illustrating the responses from adolescents regarding their attitudes towards sugary foods and beverages across multiple dimensions. The data were collected through a Likert scale questionnaire. Each dimension was normalised and plotted along the axes of the radar chart, with higher values indicating stronger agreement or higher frequency. The shaded area represents the aggregate attitudes, where larger coverage on the radar plot corresponds to more positive attitudes or higher consumption of sugary foods and beverages. The concentric circles represent intervals of normalised scores from 0 (centre) to 1 (outer edge), indicating the relative strength of responses across dimensions.
A notable finding in online supplemental table G is the general pattern of BMI among the adolescents during the study. It was revealed that out of 768 adolescents, half (50.4%) of the participants had weight within the normal range. Additionally, 6.38% (95% CI: 4.64% to 8.11%) of the respondents were found to have a high BMI, categorised as either overweight or obese.
Table 2 represents the bivariate and multivariate logistic regression analyses of behavioural determinants and socio-demographic characteristics. The extended version of bivariate analysis involving all the variables is presented in online supplemental tables D–F. The odds of having adequate knowledge among respondents was twice (adjusted OR (AOR)=2.05, 95% CI: 1.12 to 3.76) more likely for those who were living with their parents compared with those who did not. Female adolescents were 2.51 times (AOR=2.51, 95% CI: 1.61 to 3.89) more likely to consume sugary foods and beverages than their male counterparts, after adjusting for remaining predictors. Adolescents whose mothers were homemakers had odds of consuming sugary foods and beverages 1.99 times higher (AOR=1.99, 95% CI: 1.04 to 3.58) than those whose mothers were employed in job services. Similarly, adolescents whose fathers were working aboard were twice (AOR=2.09, 95% CI: 1.04 to 4.21) more likely to consume sugary foods and beverages compared with those whose fathers were engaged in job services.
Factors affecting knowledge and practice regarding sugary food and beverages among adolescents
Multinomial logistic regression between behavioural determinants and BMI has been shown in table 3. BMI of adolescents with three categories—underweight, normal and overweight/obesity—were treated as the dependent variable, with behavioural determinants such as knowledge, attitude and practice as predictors/regressors. Consumption of sugary foods and beverages was found to be significant in both models: normal versus underweight and normal versus overweight/obese. The OR of consumption of sugary foods and beverages compared with no consumption is 1.58 (95% CI: 1.05 to 2.39) for underweight relative to normal BMI and 11.95 (95% CI: 1.61 to 88.42) for overweight/obese relative to normal BMI. This indicates that the adolescents who consume sugary foods and drinks are 1.58 times and 11.95 times more likely to be underweight relative to normal BMI and overweight/obese to normal BMI, respectively, compared with adolescents who do not consume sugary foods and beverages. The positive coefficients in both models suggest that higher consumption increases the odds of both underweight and overweight/obese.
Influence of behavioural determinants on body mass index
Discussion
Various studies have been conducted to examine behavioural determinants of processed foods previously; however, most studies have not exclusively focused on sugar-rich foods and beverages specifically, to the best of our knowledge. Our study suggests adolescents living with their parents had higher odds of having adequate knowledge regarding sugary foods and beverages and ill effects associated with them. Similarly, being female, children of homemaker mothers and fathers employed in foreign had higher odds of consuming sugary foods and beverages. The odds of being overweight/obese were higher among the adolescents consuming sugary foods and beverages.
An adequate level of knowledge regarding sugary foods, beverages and their harmful impacts was found among 84.1% of the study participants, which is higher than the figure found in a study conducted among adolescents in Pokhara Valley, where only two-thirds (66.5%) of the adolescents had adequate knowledge regarding harmful effects of processed foods.33 The results in this study were also higher compared with other studies conducted by Upreti et al and Subedi and Bhusal in Chitwan, where adequate knowledge among adolescents was found to be 55.9% and 65.6%, respectively.37 38 This finding contrasts with earlier studies conducted in neighbouring nations, including India,39 Vietnam and Thailand,40 with similar context, where a lower level of adequate knowledge was reported among adolescents. A possible explanation for this disparity could be the absence of an adequate curriculum focusing on nutrition and related issues in those settings. However, this finding aligns with another study conducted by Paudel and Shrestha,41 which also indicated a higher number of respondents with adequate knowledge regarding the health effects of processed foods, including sugary items, among a comparable population. This similarity is likely due to the parallel characteristics of the populations and study settings, both involving adolescents in Nepal. The consistent results suggest that awareness of the health risks associated with sugary foods and beverages may be influenced by shared educational systems, cultural factors or public health initiatives in similar contexts.
Knowledge level was not found to be associated with the age of respondents, similar to an Indian study conducted by Joseph.42 However, the odds of having adequate knowledge among respondents were twice as high for those living with their parents compared with those who were not, even after adjusting for other explanatory variables. Similar findings were seen in another study conducted in Kathmandu.43 This may be attributed to the fact that a supportive family environment and parental guidance significantly shape adolescents’ food choices. Parents may provide better access to information on healthy eating, which is further corroborated by this study’s finding that around four-fifths of the parents discouraged their children from consuming sugary foods and beverages. Three-fifths (60.0%) of the respondents held a positive attitude and were aware of the health risks accompanied inclination to adopt healthier behaviours. The rate is higher compared with those of other studies. A cross-sectional study conducted in Mangalore city of India reported a lower rate of positive attitude (48.3%) among participants.42 This inconsistency may be attributed to differences in the level of awareness, access to health information, cultural attitudes toward dietary habits and the role of family and community support in shaping behaviour. On the other hand, a study by Upreti et al in Chitwan, Nepal, found a much higher level of positive attitude (88.4%) among adolescents.37 Similarly, findings from a similar study on sugar-sweetened beverages conducted by Thanh Ha et al showed that more than two-thirds (68.5%) of the respondents had a positive attitude.44 The variability in these findings could be influenced by cultural, educational and socio-economic differences in the study populations. These findings highlight the need for context-specific strategies to promote healthier behaviours across different populations.
The overall level of consumption of sugary foods and beverages in the study was 84.5%, which is notably higher than figures reported in other studies, where confectionaries and sugary beverages were consumed by 58% and 48% of the respondents, respectively.33 However, when broken down into categories, this study also yielded consumption rates of 69% for confectionaries and 68% for sugary beverages, which are still slightly higher. A relatively lower frequency of chocolate consumption was observed in a study conducted among adolescents in India,42 and also a slightly higher age group in Nepal.43 It is possible that these results merely reflect a selection effect, as the referenced study undertook only male participants42 and higher age group43 who may have opted for other selections in processed foods rather than sweets. However, these results align with those of Paudel and Shrestha who higher proportion of study participants consuming chocolates (72.6%) and cold drinks (52%).41 This suggests that certain food categories, like chocolates and sugary beverages, remain popular across different adolescent groups. Similarly, a large proportion of adolescents were found to be consuming sugary beverages in Malaysia.45 These results highlight the widespread consumption of sugary foods and beverages among adolescents, indicating the need for targeted interventions to reduce intake.
Consumption was found to be associated with the sex of the adolescent, as well as the occupation of the mother and father. However, knowledge level was not found to be associated with consumption status alike some previous studies.41 42 This consistency across studies suggests that while adolescents may be informed about the health risks of sugary foods and beverages, their consumption behaviours are driven more by environmental and familial factors than by knowledge alone. In contrast, a study conducted among school adolescents in Selangor, Malaysia found no associations between the socio-demographic characteristics and the consumption of sugar-sweetened beverages.45 This difference highlights the importance of understanding the local context when addressing the consumption behaviours. Around one-third of the respondents in this study were consuming sugary foods and drinks primarily due to the influence of taste. This is consistent with findings from other studies, where the highest proportion (52%) of the adolescents consuming processed foods also cited taste as the major reason for consumption, as seen in a study conducted by Subedi and Bhusal among secondary-level students in Ratnanagar municipality.38 The similarity in these findings may be due to the appeal of sugary and processed foods among adolescents, who often prioritise taste over nutritional value. Likewise, a vast majority (91%) of the adolescents were also consuming processed foods due to taste in another study conducted among adolescents of Kathmandu Valley by Paudel and Shrestha.41 Additionally, a considerable number of adolescents in this study mentioned boredom with homemade foods (18.8%) and curiosity (34.4%) as major reasons (online supplemental figure A) for their consumption, reflecting patterns found in other similar studies conducted in Nepal.33 This consistency across studies suggests a similar mindset of adolescents favouring processed and sugary foods due to sensory appeal and curiosity about unfamiliar foods. Notably, the level of knowledge was not found to be associated with the practice of sugary foods and drinks which is consistent with the findings from another study conducted in schools of Kathmandu Valley by Paudel and Shrestha.41 This further emphasises that knowledge alone may not be sufficient to change behaviour, pointing to the need for more practical, behaviour-focused interventions that address both the emotional and sensory motivations behind food choices.
Out of 768 adolescents, 6.38% of the respondents were found to have high BMI considerably categorised into overweight and obese. These results are consistent with those of Acharya et al and Khatri et al whose findings were around or below 10%.28 46 Overweight and obesity among adolescents were significantly associated with the consumption of sugary foods and beverages which is consistent with the study by Singh et al conducted among adolescents in Kathmandu Valley, which demonstrated a significantly higher risk of being overweight/obese for those who consumed soft drink ≥1 time/day compared with <1 time/day over the past month of the study (RRR=5.44, 95% CI: 2.93 to 10.10).34 Similarly, a higher proportion of overweight or obesity was seen among adolescents who consumed more fast foods than the less frequent consumers, in the study conducted in Mangalore, India.42 These results suggest that habitual consumption of high-calorie, sugar-rich foods may lead to an increased risk of obesity among adolescents. However, contrary to these findings, no association was found between KAP (Knowledge, Attitude and Practice) regarding Sugar Sweetened Beverages intake and BMI among Malay adolescents in a study conducted by Teng et al.45 This discrepancy could be attributed to differences in the study design, as our study did not consider factors such as time period of consumption and frequency.
Conclusion
The findings revealed a good level of knowledge, moderate level of attitude and poor level of lifestyle choices among sugary foods and beverages intake of the participants. Despite awareness of the health effects of excessive sugar consumption, adolescents continued to favour high-sugar foods. To address these issues, long-term behavioural change interventions are necessary to decrease the consumption of sugary products and improve overall adolescent well-being. The government of Nepal should implement family-centred educational programmes that highlight parental influence on dietary habits, enforce policies prohibiting sugary food sales in schools and use health promotion strategies through various media channels. The schools can provide information sessions for parents on occasions like teacher–parents meet. Health promotion should be done through various audio-video materials and ringtones of the national telecommunication companies as done during COVID-19 pandemic.
Data availability statement
Data are available upon reasonable request. All data analysed during this study are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by Institutional Review Committee, Manmohan Memorial Institute of Health Sciences, Institute of Medicine, Tribhuvan University, Nepal. Reference number: MMIHS IRC 887; Ref 79/151. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
The authors would to like to thank all school authorities and students who participated in the study. The authors also acknowledge Office of the Municipal Office, Education Section and Health Section of Nagarjun municipality for the coordination offered during the study.
References
Footnotes
X @chetry_thapa, @poonampokhrel4
Contributors LG and MT contributed equally to this paper. LG and MT were responsible in conceptualising the study, designing the protocol and developing the research tools. MT was involved in the collection of data and reviewing the literature. MT and KP were involved in data curation alongside formal statistical analysis. PMS, PP, AB and MT were responsible for writing the original draft of the manuscript. LG and KP were responsible for critically reviewing and supervising the manuscript providing substantial input. MT is the guarantor of the manuscript. All authors contributed to data interpretation, reviewed successive drafts and approved the final 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.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
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.