Weight misperception amongst youth of a developing country: Pakistan -a cross-sectional study
- Muhammad Danish Saleem1Email author,
- Gulrayz Ahmed1,
- Juwaria Mulla1,
- Syed Sami Haider†1 and
- Mustafa Abbas†1
© Saleem et al.; licensee BioMed Central Ltd. 2013
Received: 27 January 2013
Accepted: 31 July 2013
Published: 2 August 2013
Weight misperception is the discordance between an individual’s actual weight status and the perception of his/her weight. It is a common problem in the youth population as enumerated by many international studies. However data from Pakistan in this area is deficient.
A multi-center cross-sectional survey was carried out in undergraduate university students of Karachi between the ages of 15–24. Participants were questioned regarding their perception of being thin, normal or fat and it was compared with their Body Mass Index (BMI). Measurements of height and weight were taken for this purpose and BMI was categorized using Asian cut offs. Weight misperception was identified when the self-perceived weight (average, fat, thin) did not match the calculated BMI distribution. Chi square tests and logistic regression tests were applied to show associations of misperception and types of misperception (overestimation, underestimation) with independent variables like age, gender, type of university and faculties. P-value of <0.05 was taken as statistically significant.
42.4% of the total participants i.e. 43.3% males and 41% females misperceived their weight. Amongst those who misperceived 38.2% had overestimated and 61.8% had underestimated their weight. Greatest misperception of was observed in the overweight category (91%), specifically amongst overweight males (95%). Females of the underweight category overestimated their weight and males of the overweight category underestimated their weight. Amongst the total participants, females overestimated 8 times more than males (OR 8.054, 95% CI 5.34-12.13). Misperception increased with the age of the participants (OR 1.114, 95% CI 1.041-1.191). Odds of misperception were greater in students of private sector universities as compared to public (OR 1.861, 95% CI: 1.29-2.67). Odds of misperception were less in students of medical sciences (OR 0.693, 95% CI 0.491-0.977), engineering (OR 0.586, 95% CI 0.364-0.941) and business administration (OR 0.439, 95% CI 0.290-0.662) as compared to general faculty universities.
There was marked discrepancy between the calculated BMI and the self-perceived weight in the youth of Karachi. Better awareness campaigns need to be implemented to reverse these trends.
KeywordsWeight misperception BMI Pakistan Undergraduate students Youth Overweight Underweight Overestimation Underestimation
Weight misperception is the discordance between an individual’s actual weight status and the perception of his/her weight . Weight misperception may be broadly divided into under-estimating or over-estimating ones weight. Underestimation or underweight misperception is considering oneself of a lower weight than they actually are, as determined by Body Mass Index (BMI) . Overestimation or overweight misperception is considering oneself of being a higher weight than they actually are, as determined by BMI .
The youth (age group of 15–24), according to the 2007 United Nations Development Program statistics form a substantial 21.8% (36 million) of the total population of Pakistan . In spite of this high percentage, very few epidemiological studies have been conducted on this age group in our country.
Obesity is emerging as a public health problem. Traditionally thought of as a disease of adulthood, it is finding its way into adolescents and the youth population. Rehman et al. reported an overwhelming 18% overweight distribution amongst tenth grade adolescents of Karachi . Obesity or being overweight increases the risk of adverse cardiovascular outcomes, type 2 diabetes mellitus, cancers and even death . On the other end of the spectrum, being underweight poses serious problems, especially in a developing country like Pakistan. A report on the nutritional status of Pakistan states a high prevalence of the population as underweight, majority of which are women of reproductive ages . Being underweight affects both sexes alike, however the effects are more harmful in the female youth and young mothers with implications such as anemia, infertility or complications in pregnancy endangering both the life of the mother and of their babies .
Research studies have demonstrated that self-awareness of being overweight/ obese is an essential factor in attempting to lose weight . Consequently, weight loss is associated with decreased health risk and can substantially reduce the incidence of hypertension and cardiovascular diseases . Intake of proper food and diet with regular eating hours and awareness in the general population about the harms of being underweight and undernourished can help improve the complications associated with undernourishment .
The key to health and weight management is proper weight perception. Numerous studies show that overweight and underweight misperception is a dangerous albeit common problem in the youth around the globe. Data from the 2007 Minnesota student survey of 87,418 high school students showed that a quarter of their students had improper weight perception which was more prevalent in males, with girls more likely to overestimate and boys more likely to underestimate . The CDC’s National youth risk behavior survey of 2007 consisting of 12,853, 9th to 12th grade students in U.S public and private schools showed that 23% of overweight girls and 40% of overweight boys had misperceived . A Turkish study in high school students aged 14 to 21 years from Adana, a southern city of Turkey, during 1999 also reported high weight misperception . This study on 2,352 students incorporated the Turkish version of youth risk behavior survey questionnaire and reported that girls misperceived themselves to be overweight and obese as compared to boys and displayed a greater percentage of weight loosing behaviors .
The youth, at the brink of joining professional life are a country’s major asset and any health problems identified and resolved at this stage would benefit the adults of future years. Weight misperception has previously been identified in the adults of Pakistan. A survey conducted on 493 adults in Karachi, Pakistan demonstrated a poor agreement between self-perception and actual BMI with a 73% and 50% misperception amongst the obese and overweight participants respectively . However no comprehensive survey demonstrating weight perception in the youth of Pakistan has been conducted.
For our sample we targeted the youth population of Karachi studying in various undergraduate universities. The purpose of our study was to obtain weight misperceptions amongst the youth population by comparing their actual weight categories based upon calculated BMI with their self-perception and to explore some relationships of this misperception with gender, type of university (public or private) and faculties (general, engineering, medical sciences and business administration).
We conducted a multi-center cross-sectional study in different undergraduate universities of Karachi, Pakistan, during the months of January and February 2012. Sixteen hundred individuals from six universities were interviewed. Universities were selected on the basis of systematic sampling from the list of 33 public and private sector universities of Karachi, recognized by the Higher Education Commission of Pakistan. Universities were arranged in alphabetical order. The first university was selected randomly, while the remainder five universities were chosen with a gap of five universities each. For the purpose of data collection, the sample was distributed amongst the individual universities by proportion of their estimated student population. Simple random sampling using the student roll numbers was used to select the study participants.
Both private and government undergraduate universities were selected to ensure participation of students from all socioeconomic backgrounds and localities of Karachi, producing a better representation of the youth population. Even though these universities do cater to a meager amount of international students, our study involved data collection only from Pakistani nationals. Karachi is a mega-city and the economic capital of the country. It hails inhabitants from all cultures and migrants from all cities of the country. Thus participants selected form an ideal representation of all the students of Pakistan. All participants, male and female, present in the university during the data collection period that fell in the age range of 15 - 24  years were included. Any student above and below these ages were excluded. We approached a total of 1,600 students, out of which 1,400 agreed to participate in the survey (response rate 87.5%).
Data collection tool
We designed a short structured, close-ended questionnaire in English, which is the medium of instruction in colleges and universities of Pakistan. It is comprised of the variables and demographics. The students were approached by the investigators, the procedure of the study was explained to them and informed signed consent was obtained. The questionnaire was administered in the form of an interview. The interview was followed by the height and weight measurements of the participants. Participants were interviewed prior to the procedure of physical measurements to reduce bias in reporting misperception. A meeting of the investigators was held prior to the data collection procedure to maintain uniformity in its administration.
The weight (in kilograms) and height (in centimeters) of the participants was measured by the investigators. BMI was calculated from height and weight measurements and then categorized into underweight (<18.5), normal (18.5-22.9), overweight (23–24.9) and obese (≥ 25.0) based on the new Asian cutoffs .
Self-perception of body weight was determined by the question: Do you consider yourself as a) average b) thin c) fat. This question was structured in plain English so that all the participants could answer correctly.
Participant demographics namely age, gender, faculty (general, engineering, medical sciences and business administration) and type of university (public or private) were considered as the other independent variables. Ethnicity was not reported as all the participants belonged to the South Asian ethnicity.
The responses were then compared to the calculated BMI to determine misperception. For underweight participants, misperception was considered when the response to the question was average or fat. For normal participants, misperception was considered when the response was thin or fat. For overweight/obese participants, misperception was considered when the response was average or thin. The types of misperception (overestimation or underestimation) were also determined.
All data were entered and analyzed on SPSS v.18. Mean and standard deviation was calculated for all the continuous variables like age and BMI. Frequencies and percentages were reported for all the categorical variables such as gender, BMI distribution, type of university and faculties. T-test was used to measure differences in means for continuous variables between males and females. Chi square test was applied to determine association for the categorical variables with gender, misperception and types of misperception. Two logistic regression models were applied to see the effect of demographic variables for both misperception and types of misperception (overestimation vs underestimation). Gender was not found to be statistically significant with misperception in the regression model. Meanwhile gender was the only variable found to be statically significant with types of misperception in the regression model. P-value of less than 0.05 was taken as statistically significant.
Data from female participants was collected by the female investigators keeping in view the social and cultural setup of our country. The questionnaires were kept anonymous to ensure confidentiality and were numbered serially. Informed signed consent was obtained from all the participants and the study protocol was reviewed and approved under the Dow University Ethical Review Board.
Gender based distribution of independent variables
N = 856
N = 544
1. University type
General (Art+ Science)
Business and Administration
Three hundred and eighty five (25.6%) participants considered themselves thin, 933 (66.6%) considered themselves average and 109 (7.8%) considered themselves fat. Most females (72.8%) self-perceived themselves to be average. Compared to females, more males considered themselves to be either thin or fat. The association between self-perception and gender was found to be significant (p-value = 0.001). (Refer to Table 1).
Association of misperception with independent variables
N = 856
N = 544
3. University type
General (Art+ Science)
Associations between misperception and age, type of university and faculty (results from logistic regression)
2. Type of university
Types of misperception
Misperception was classified as either overestimation or underestimation. Amongst those who misperceived, 227 (38.2%) had overestimated their weights and 367 (61.8%) had underestimated their weight status. Majority of the total overestimation was observed amongst the participants of the underweight category (n = 214, 94.3%), while majority of the total underestimation was observed from the participants of the normal (43.6%) and overweight category (38.1%). The association between the types of misperception and gender was found to be significant (p-value < 0.001). Underestimation was more commonly observed among males (296 of the 371 misperceiving males). In contrast overestimation was more commonly observed among females (152 of the 223 misperceiving females). Logistic regression for types of misperception showed that males had an eight times more chance of underestimation as compared to females (OR 8.054, 95% CI 5.34-12.13).
Distribution of overestimation and underestimation of weight according to BMI and Gender
N = 227
N = 367
Weight misperception was a highly prevalent finding in our study with as much as 42.4% overall misperception seen in the total youth population. We witnessed a considerable fraction of individuals in the overweight and obese categories erroneously reporting their weight status; hence showing a high level of misperception among overweight and obese participants. This is of concern as literature has shown that correct perception of one’s weight is strongly associated with positive efforts to lose weight and maintain a healthy life style. Skinner et al.  and Jones et al.  reported that those who misperceived their weight had less weight concerns, less control over distress and emotional overeating, more often indulged in improper dietary intake (including fast food and soda beverages) along with inappropriate snack timings and had more sedentary and stationary lifestyles. Edwards et al. reported that overweight youngsters who precisely perceived their weight were more likely to lose weight and indulge in exercises, all the while consciously consuming fewer calories .
Our results regarding weight misperception are similar to those seen in studies from both developed and developing countries. High proportions of misperception regarding weight in the youth populations were reported from USA , Turkey , China , Mexico , Spain  and Malaysia  where high percentages of underweight individuals misperceiving themselves as normal was observed.
Our study showed that misperception increased with the increasing age of participants (OR 1.114, 95% CI 1.041-1.191). Similar result has also been shown by a study conducted in Spanish adolescents which reported a higher mean age for those who misperceived as compared to the others . Chinese  and Mexican  studies have also reported underestimation to be associated with older age.
The chances of misperception were greater in students of private sector universities as compared to public sector. It was also noted that students of the general faculty misperceived more as compared to engineering, medical sciences and business administration faculties. Students from the public sector science faculties had the least misperception of weight as compared to the other undergraduate students in our sample.
The two most important findings in our study were the high prevalence of misperception amongst underweight females as being normal or overweight (overestimation) and the misperception amongst overweight and obese males as being normal or underweight (underestimation).
Overestimation of weight in females
Overestimation of weight was a less common finding in our setting; nevertheless one group outnumbered the rest i.e. underweight females overestimating their weights. It would not be a surprise if this misperception was accompanied by unhealthy weight reduction practices. Talamayan et al. reported that a significant portion of normal weight adolescents misperceive themselves as overweight and engage in unhealthy weight control behaviors. Females (16.8%) outnumbered males (6.8%) in practicing at least one such unhealthy behavior (use of diet pills, laxatives, and fasting) in the past 30 days . Another study conducted in students of Adana reported 35.5% participants as wanting to lose weight, with weight losing intentions and interventions (diet, provocative vomiting and 24-hours starving) more common amongst girls . Indeed anorexia has been noted as an important problem in female students of Pakistan , even though a relationship between anorexia and weight misperception still needs to be determined in Pakistani students.
Females tend to be more sensitive towards their weight status and are likely to take social, peer and family pressures more seriously than their male counter-parts. A study conducted on female college students showed that many females identify media as a source of motivation and pressure, encouraging them to be of certain weight . A study conducted amongst university students of Karachi had similar results with higher percentages of underweight women perceiving themselves as normal . This study reported that vast majorities of females were influenced by images of thin models and consequently strived to achieve this ideal body image even when they were majorly underweight . Wardle et al. showed that about half of the female students from 22 different countries considered themselves overweight and were trying to lose weight despite the fact that a majority of them had normal BMI values .
Underestimation of weight in males
In our study majority of overweight and obese males considered themselves as having a normal or average weight. The preference of males for a heavier body structure while perceiving themselves to be average when most were overweight can have many reasons. The abundance of ‘muscular’ models portrayed in the media  coupled with a relative absence of particularly thin male models could be one of the causes. A study amongst youth of the United States showed that males tend to perceive themselves as normal whilst they were actually overweight . Similar studies in residents of Guangdong province of China showed that males were more likely to underestimate weight than females (25.8% vs. 8.5%) . Customarily, increased body weight has been an emblem of health and wealth in a variety of populations . This is especially true for the Indian subcontinent where people still consider weight to be associated with wealth, prosperity and good health. This may lead them to misperceive their weight as normal which can put them at a higher risk for multiple diseases .
Bhanji et al. reported that few factors associated with underestimation of weight were male gender, being happy with ones’ weight and not knowing one’s ideal weight . Indeed inadequate knowledge of ideal body weight is also a major contributing factor in weight perception, especially in our social setup, where school heath and health education is not stressed upon. A study from New Delhi, India reported nearly half of the overweight or obese students misclassifying their weight status , statistics very similar to our own for the same socioeconomic and cultural background.
Observation of misperception at the ends of the spectrum is rather alarming due to adverse psychological and health related outcomes as severe as suicide, reported by Chinese  and Korean studies . Therefore it is important for health professionals to identify both underweight and overweight/obese, educate them about the health risks of their weight and advise them appropriate strategies for gaining or losing weight. The weight loss or gain regime should be closely monitored as drastic weight loss or gain might have adverse effects. More importantly, the interventions made to correct the weight should be carefully designed to protect against body image disorders, eating disorders and emotional distress. Moreover wide scale campaigns need to be undertaken to prevent unhealthy weight trends observed by initiating healthy eating habits and advocating adequate exercise in children , adolescents , the youth  and adults . School health should also be improved with emphasis on educating the students about ideal weight and healthy weight maintenance.
Strengths and limitations
There exists no data on such an age group in our country. Pakistan being the 6th most populous nation of the world, can give a good perspective of the developing world. The most significant achievement of this study was the objective calculation of BMI of the participants.
There were some limitations of our study as well. The sample cannot be a true reflection of the entire youth population of Pakistan, especially the rural population where many do not have access to university education. The terminology ‘fat’ which was part of the question assessing self-perception may be a limiting factor in our study. ‘Fat’ is a subjective word and may be difficult for people to accept. A major limitation with using BMI scale in measuring up male body was that it does not consider increased muscle mass which could have been a contributing factor of raised BMI in some males. Moreover it could be possible that that these muscular males who ‘misperceived’ themselves as normal weight would in fact not be misperceiving, rather the BMI scale was wrongly classifying them as overweight.
There was marked discrepancy between the calculated BMI and the self-perceived weight in the youth of Karachi. There was overestimation of weight by underweight females and underestimation by overweight males. Better awareness campaigns need to be implemented to reverse these trends.
Body mass index.
We would like to acknowledge the help provided to us by Miss Mehwish Hussain (Senior Lecturer, Research Department DUHS) in the statistical analysis and results writing part of the manuscript. We would also like to thank Prof. Nazeer Khan (Director Research Department DUHS) for his guidance throughout. We would also like to thank the administrations of Baqai Medical University, Dow University of Health Sciences, Institute of Business administration, Karachi Institute of Engineering and Technology, Preston University and University of Karachi for their cooperation.
- Duncan DT, Wolin KY, Scharoun-Lee M, Ding EL, Warner ET, Bennett GG: Does perception equal reality? Weight misperceptionin relation to weight-related attitudes and behaviors among overweight and obese US adults. Int J BehavNutrPhys Act. 2011, 8: 20-Google Scholar
- Park E: Overestimation and underestimation: adolescents’ weight perception in comparison to BMI-based weight status and how it varies across socio-demographic factors. J Sch Health. 2011, 81 (2): 57-64. 10.1111/j.1746-1561.2010.00561.x.View ArticlePubMedGoogle Scholar
- United Nations Development Program [Internet]: Pakistan: UNDP and Youth, c2013 [cited 2013 January 13]. URL available at http://undp.org.pk/undp-and-the-youth.html
- Rehman T, Rizvi Z, Siddiqui U, Ahmad S, Sophie A, Siddiqui M, et al: Obesity and adolescence of Pakistan. J Pak Med Assoc. 2003, 53: 315-319.Google Scholar
- WHO: Obesity: preventing and managing the global epidemic: report of a WHO consultation. World Health Organ Tech Rep Ser. 2000, 894: i-xii,:1–253-Google Scholar
- Qureshi SK, Nazli H, Soomro GY: Nutritional Status of Pakistan. 2001, Islamabad, Pakistan: Institute of Nutritional Economics, Report No.: 8. Available at URL: http://www.pide.org.pk/Mimap/Report08.pdf Google Scholar
- Resse MATB: Underweight: a heavy concern. Today’s Dietitian. 2008, 10 (1): 56-Google Scholar
- Lemon SC, Rosal MC, Zapka J, Borg A, Andersen V: Contributions of weight perceptions to weight loss attempts: differences by body mass index and gender. Body Image. 2009, 6 (2): 90-96. 10.1016/j.bodyim.2008.11.004.View ArticlePubMedPubMed CentralGoogle Scholar
- Elmer PJ, Obarzanek E, Vollmer WM, Simons-Morton D, Stevens VJ, Young DR, et al: Effects of comprehensive lifestyle modification on diet, weight, physical fitness, and blood pressure control: 18-month results of a randomized trial. Ann Intern Med. 2006, 144 (7): 485-495. 10.7326/0003-4819-144-7-200604040-00007.View ArticlePubMedGoogle Scholar
- Edwards NM, Pettingell S: Where perception meets reality: self-perception of weight in overweight adolescents. Pediatrics. 2010, 125 (3): e452-e458. 10.1542/peds.2009-0185.View ArticlePubMedGoogle Scholar
- Kurdak H, Bozdemir N, Saatci E, Ozturk P, Ozcan S, Akpinar E: Self-perceived body weight status and weight-control behaviors of high school students in a southern city of Turkey. CollAntropol. 2010, 34 (4): 1295-1302.Google Scholar
- Bhanji S, Khuwaja AK, Siddiqui F, Azam I, Kazmi K: Underestimation of weight and its associated factors among overweight and obese adults in Pakistan: a cross sectional study. BMC Publ Health. 2011, 11: 363-10.1186/1471-2458-11-363.View ArticleGoogle Scholar
- WHO Expert Consultation: Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004, 363 (9403): 157-163.View ArticleGoogle Scholar
- Skinner AC, Weinberger M, Mulvaney S, Schlundt D, Rothman RL: Accuracy of perceptions of overweight and relation to self-care behaviors among adolescents with type 2 diabetes and their parents. Diabetes Care. 2008, 31 (2): 227-229.View ArticlePubMedGoogle Scholar
- Jones M, Grilo CM, Masheb RM, White MA: Psychological and behavioral correlates of excess weight: misperception of obese status among persons with Class II obesity. Int J Eat Disord. 2010, 43 (7): 628-632. 10.1002/eat.20746.View ArticlePubMedPubMed CentralGoogle Scholar
- Xu BY, Zhang YH, Ma WJ, Xu YJ, Song XL, Nie SP, et al: Prevalence regarding weight misperception and related influencing factors among residents in Guangdong province. Zhonghua Liu Xing Bing XueZaZhi. 2011, 32 (10): 964-968.Google Scholar
- Andrade FC, Raffaelli M, Teran-Garcia M, Jerman JA, Garcia CA: Weight status misperception among Mexican young adults. Body Image. 2012, 9 (1): 184-188. 10.1016/j.bodyim.2011.10.006.View ArticleGoogle Scholar
- Jáuregui-Lobera I, Bolaños-Ríos P, Santiago-Fernández MJ, Garrido-Casals O, Sánchez E: Perception of weight and psychological variables in a sample of Spanish adolescents. Diabetes MetabSyndrObes. 2011, 4: 245-251.Google Scholar
- Khor GL, Zalilah MS, Phan YY, Ang M, Maznah B, Norimah AK: Perceptions of body image among Malaysian male and female adolescents. Singapore Med J. 2009, 50 (3): 303-311.PubMedGoogle Scholar
- Talamayan KS, Springer AE, Kelder SH, Gorospe EC, Joye KA: Prevalence of overweight misperception and weight control behaviors among normal weight adolescents in the United States. Scientific World Journal. 2006, 6: 365-373.View ArticlePubMedGoogle Scholar
- Babar N, Alam M, Ali SS, Ansari A, Atiq M, Awais A, et al: Anorexic behavior and attitudes among female medical and nursing students at a private university hospital. J Pak Med Assoc. 2002, 52 (6): 272-276.PubMedGoogle Scholar
- Malinauskas BM, Raedeke TD, Aeby VG, Smith JL, Dallas MB: Dieting practices weight perceptions and body composition: a comparison of normal weight overweight and obese college females. Nutr J. 2006, 5: 11-10.1186/1475-2891-5-11.View ArticlePubMedPubMed CentralGoogle Scholar
- Khan AN, Khalid S, Khan HI, Jabeen M: Impact of today’s media on university student’s body image in Pakistan: a conservative, developing country’s perspective. BMC Publ Health. 2011, 11: 379-10.1186/1471-2458-11-379.View ArticleGoogle Scholar
- Wardle J, Haase AM, Steptoe A: Body image and weight control in young adults: international comparisons in university students from 22 countries. Int J Obes (Lond). 2006, 30: 644-651. 10.1038/sj.ijo.0803050.View ArticleGoogle Scholar
- Lorenzen LA, Grieve FG, Thomas A: Exposure to muscular male models decreases men’s body satisfaction. Sex Roles. 2004, 51 (11–12): 743-748.View ArticleGoogle Scholar
- Howard NJ, Hugo GJ, Taylor AW, Wilson DH: Our perception of weight: socioeconomic and sociocultural explanations. Obes Res ClinPrac. 2008, 2 (2): 125-131.View ArticleGoogle Scholar
- Jafar TH, Chaturvedi N, Pappas G: Prevalence of overweight and obesity and their association with hypertension and diabetes mellitus in an Indo-Asian population. CMAJ. 2006, 175 (9): 1071-1077.View ArticlePubMedPubMed CentralGoogle Scholar
- Stigler MH, Arora M, Dhavan P, Shrivastav R, Reddy KS, Perry CL: Weight-related concerns and weight-control behaviors among overweight adolescents in Delhi, India: a cross-sectional study. Int J BehavNutrPhys Act. 2011, 8: 9-Google Scholar
- Lo WS, Ho SY, Mak KK, Lai HK, Lai YK, Lam TH: Weight misperception and psychosocial health in normal weight Chinese adolescents. Int J PediatrObes. 2011, 6 (2–2): e381-e389. Epub 2010 Sep 20Google Scholar
- Kim DS, Cho Y, Cho SI, Lim IS: Body weight perception, unhealthy weight control behaviors, and suicidal ideation among Korean adolescents. J Sch Health. 2009, 79 (12): 585-592. 10.1111/j.1746-1561.2009.00452.x.View ArticlePubMedGoogle Scholar
- Han JC, Lawlor DA, Kimm SY: Childhood obesity. Lancet. 2010, 375 (9727): 1737-1748. 10.1016/S0140-6736(10)60171-7.View ArticlePubMedPubMed CentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2458/13/707/prepub