| Meta-analysis of the detection rate of eating disorders among global femalesObjective:Eating disorders(ED)is a kind of serious mental illness,women have 10%of the risk in the lifetime,but the risk is low in men.At the same time,eating disorders was associated with quite a number of physical and mental disease,and it's mortality was highest compared with any mental illness,equated with drug abuse.Although,there were numerous reports on academic periodicals at domestic and overseas for the prevalence of eating disorder,the application of screening tool and diagnosis methods and the source of population were varied.Meta-analyses of the detection rate of eating disorders(ED)among women based on healthy population groups have not been reported.To improve the evidence system of the detection rate of eating disorders,merge the detection rate of original studies,summarize the various assessment methods for eating disorders screening,and observe the impact of different research characteristics on the detection rate of DE.Methods:1 Literature retrieval and retrieval strategies.The published papers on the detection rate of ED and pertinent reference articles which were published before December 2018 were collected from databases including Chinese National Knowledge Infrastructure(CNKI),Chinese BioMedical Literature Base(CBM),Wanfang(Chinese),EMBASE and PubMed.The key words were‘disordered eating'or‘eating disorder'and‘prevalence'or‘detection rate'.2 Inclusion and exclusion criteria.Studies were included following the criteria:1)Baseline data was collected from cross-sectional,longitudinal study or experimental studies;2)The subjects of the study were healthy women;3)Eating disorders can be assessed by specific questionnaires or scales;4)The specific number of subjects was reported or can be obtained by extrapolation;5)sample size?60.Exclusion criteria included:1)Only male samples were reported;2)A sample of children(<11 years of age)was used;3)The study included people who were not healthy(overweight,obese,diabetes,etc.)or professional(athletes,military,dancers,etc.)and pregnant women;4)Unclear diagnostic or evaluation criteria.3 The risk of bias assessment.The derivative evaluation tool of the Newcastle-Ottawa Scale(NOS)was used to evaluate the risk bias of each article.The risk bias was divided into three grades,namely low,medium and high,or unclear,according to the results of the representatives of samples,sample size,the source of population,and the measure of results.4 Data extraction.Authors,year of publication,country or region,source of population and age,sample size,and assessment scales of eating disorders and cutoff were selected for each study.5 Statistical analysis.Due to the change of expected changes such an population,methods and the transformation of epidemiology,random effects model were used.The heterogeneity of research literature was evaluated by application of forest figure and I~2 test.Potential publication bias were evaluated through assessing the symmetry of the funnel figure using Egger test.The potential effect of factors of various study-level on the pooled detection rate were explored using subgroup and univariable random-effects meta-regression analyses.Stata 12.0 was used for statistical analysis.Results:1 Literature search results.A total of 77 papers involving 140,846 subjects were included,covering 35 countries and regions.2 Evaluation of literature quality.The literature quality was medium and high(average 9.9 points),and all studies were not excluded according to the quality.3 Statistical results.3.1 The measure of eating disorders.A total of six measurement and evaluation methods were included in the included literature:16 SCOFF scales,43 EAT-26 scales,10 EAT-related scale.Eating disorders was commonly classified when the cutoff was 2 for SCOFF,or 20 for EAT-26.3.2 The pooled detection rate of eating disorders was 20.47%(95%CI:16.43-24.51).In the subjects screened through SCOFF and EAT-26 scales,the detection rate was 19.28%(95%CI:14.63-23.93)and 19.64%(95%CI:16.25-23.03),respectively.Furthermore,the detection rate of ED among adolescents(21.65%;95%CI:14.98-28.32)was slightly higher than young adults(19.83%;95%CI:16.92-22.75).The detection rate of ED in developing countries was lower(15.76%;95%CI:15.34-16.17)than in developed countries(22.41%;95%CI:22.19-22.62).Age group(?=-0.045;95%CI:0.087-0.004;P=0.033)had a significant effect on the pooled detection rate,suggesting that the detection rate of eating disorders was higher in the younger.The study of the detection rate of eating disorders and relevant associations among Chinese female adolescents and young adultsObjective:With the rapid social and economic development and urbanization in China,media use and dietary practices have changed,that may associate with risk of eating disorders.This research aimed to compare the changes of detection rate of eating disorders and relevant associations in China between 2009 and 2011,and explore the associations among weight loss ideas and behaviors(WLID),overweight and obesity,media use,dietary practices and DE in 2011.Methods:1 Design.Participants were recruited from a cohort survey,namely,the China Health and Nutrition Survey(CHNS).A cross-sectional design was conducted in this study by using data from the 2009 and 2011 wave.2 Participants.This study focused on females aged 12–35 years who participated in the 2009 and 2011 wave survey and provided information on age,height and weight,eating disorder symptoms,media use(Internet access and use,television use)and dietary practices.3 Measures.Two separate CHNS questionnaires were created for children and adolescents(<18 years)and young adults(?18 years)to complete.3.1 Eating disorders.Eating disorder symptoms were assessed using the SCOFF questionnaire.3.2 Weight loss ideas and behaviors.Weight loss ideas and behaviors(WLIB)in the past four weeks were reported by participants by following the eating disorder symptoms questionnaire.3.3 Weight status.Weight status were classified as underweight,normal weight,overweight and obesity.3.4 Media use.Media use included television and Internet use.3.5 Dietary practices.Dietary practice variables were obtained from the three days of 24-hour recall of individual dietary intake and the frequency and location of meals.3.6 Covariates.Age,body mass index(BMI),and urbanization were potential confounding effects and conducted on the models of adjustments.3.7 Statistic analysis.All statistical analyses were conducted using SPSS software,version 21.0,for Windows.Analyses were stratified by age,and participants were grouped into adolescents and young adults.The detection rate of disordered eating or screen-detected eating disorders were estimated based on the 2011 CHNS data and compared with the 2009 wave among female adolescents and young adults.Comparisons of relevant associations of eating disorders between the 2011 and 2009 survey were also conducted.A chi-square(?~2)test and a t test were used for all differential comparisons.The associations between urbanization index,media use,and weight loss ideas or/and behaviors in 2011 were explored by adjusting and not adjusting for confounders.Logistic regression models were applied to assess the associations between media use,dietary practices,and eating disorders adjusting,without adjusting for age,BMI,and urbanization.All two-tailed P values<0.05 were considered statistically significant.Results:According to the inclusion criteria,a total of 1,555 participants at baseline were included:322 female adolescents and 1,233 female young adults in CHNS 2011.Compared with 2009,the detection rate of ED for young adults significantly increased from 6.3%to 9.4%;a similar but nonsignificant trend was seen in adolescents,from 7.8%to 12.7%in 2011.With the development of urbanization,Internet use were significantly popularized,positively associated with WLID.There were nonsignificant associations with DE among adolescents,except for WLID and overweight-obesity.The percent of food eaten at home and dietary macronutrient intake of protein and fat had significant associations with DE among young adults.Conclusion:1 This analysis suggested that DE was common in women in various countries and regions.Adolescents might have higher risk of ED than young adults,and people in developed countries could be more likely to have DE than in developing countries.2 This study reported a significantly higher detection rate of eating disorders for female young adults and a similar but nonsignificant trend for female adolescents in China in 2011 compared with 2009:the detection rate increased from 7.8%to 12.7%for adolescents and from 6.3%to 9.4%for young adults,respectively.These results are still lower than the meta-analysis,which estimated the detection rate of eating disorders in the world using the SCOFF questionnaire,the pooled detection rate were 21.65%and 19.83%for female adolescents and young adults,respectively.3 With the development of urbanization,media use was significantly popularized,especially Internet use,and closely associated with WLIB but not associated with eating disorders.4 The risk of eating disorders will increase with the WLIB and overweight-obesity in female adolescents.5 Additionally,associations among media use,dietary practices and eating disorders symptoms are still extant.6 Status of eating disorders in adolescents is relatively high.Awareness and understanding of DE and its associated risk factors is essential for females. |
PDF Full Text Request