Obesity results from the interaction of genetic and environmental factors. To search for sequence variants that affect variation in two common measures of obesity, weight and body mass index (BMI), both of which are highly heritable, we performed a genome-wide association (GWA) study with 305,846 SNPs typed in 25,344 Icelandic, 2,998 Dutch, 1,890 European Americans and 1,160 African American subjects and combined the results with previously published results from the Diabetes Genetics Initiative (DGI) on 3,024 Scandinavians. We selected 43 variants in 19 regions for follow-up in 5,586 Danish individuals and compared the results to a genome-wide study on obesity-related traits from the GIANT consortium. In total, 29 variants, some correlated, in 11 chromosomal regions reached a genome-wide significance threshold of P < 1.6 x 10(-7). This includes previously identified variants close to or in the FTO, MC4R, BDNF and SH2B1 genes, in addition to variants at seven loci not previously connected with obesity.

The traditional epidemiologic modes of data collection, including paper-and-pencil questionnaires and interviews, have several limitations, such as decreasing response rates over the last decades and high costs in large study populations. The use of Web-based questionnaires may be an attractive alternative but is still scarce in epidemiologic research because of major concerns about selective nonresponse and reliability of the data obtained. The authors discuss advantages and disadvantages of Web-based questionnaires and current developments in this area. In addition, they focus on some practical issues and safety concerns involved in the application of Web-based questionnaires in epidemiologic research. They conclude that many problems related to the use of Web-based questionnaires have been solved or will most likely be solved in the near future and that this mode of data collection offers serious benefits. However, questionnaire design issues may have a major impact on response and completion rates and on reliability of the data. Theoretically, Web-based questionnaires could be considered an alternative or complementary mode in the range of epidemiologic methods of data collection. Practice and comparisons with the traditional survey techniques should reveal whether they can fulfill their expectations.

Obesity is one of the biggest challenges facing global reproductive health. Women in the UK and USA are today more likely to be obese or overweight at booking than normal weight, and many low-and middle-income countries (LMICs) seem destined to follow suit (Poston et al. Lancet Diabetes Endocrinol 2016;4:1025-36). Understanding how, and to what extent, maternal body mass index (BMI) and weight gain during pregnancy contribute to adverse outcomes for mothers and their offspring is therefore vital to informing future health policy.In an individual participant data meta-analysis of over 265 000 births, Santos et al. (BJOG 2019;126:984-95) confirm strong correlations between pre-pregnancy maternal BMI and the risks of gestational hypertension, preeclampsia and gestational diabetes. Over one-third of such complications in the study population were considered attributable to maternal overweight and obesity. The risk of large size for gestational age (LGA) at birth increased similarly across all categories of pre-pregnancy BMI and gestational weight gain, although these data should be interpreted in the context of a continuing debate regarding the customisation of fetal growth charts. It remains uncertain how maternal height and weight influence fetal growth potential, and whether LGA babies born to mothers who are obese or mothers with excessive weight gain carry the same short-and long-term health risks as LGA babies born to mothers who are normal weight. Preterm birth was also more common among women who are obese and past literature has suggested that this association is strongest for extremely preterm delivery (Cnattingius et al. JAMA 2013;309:2362-70), whether spontaneous or iatrogenic.Whereas women who are obese or have high weight gain are consistently shown to be at greatest risk, there is clear evidence of a continuum of risk across the full BMI range, which is emphasised by the authors' use of population attributable risk (PAR). Notably, the overall burden of pregnancy complications is similar in overweight and obese groups (PAR 11.4 and 12.5%, respectively). This calls into question traditional models of care targeting women with a booking BMI above 30 kg/m 2 or even higher thresholds. Minimising gestational weight gain in these women ameliorates but does not remove the excess risk, and ultimately may have less impact on outcomes at a population level than previously hoped.The authors acknowledge that the data were derived from cohorts who were largely white; however, comparable findings have been reported in LMICs with varied ethnic populations (Rahman et al. Obes Rev 2015;16:758-70).Being healthy entails more than just not being obese, and the study also draws important attention to the risks of small size for gestational age and preterm birth, particularly amongst underweight women with inadequate weight gain during pregnancy. These findings strengthen the argument for novel public health approaches to optimise maternal health with a shift in focus towards pre-conception and interpregnancy intervent...

Mental retardation (AIR) is a serious and lifelong disability that places heavy demands on society and the health system. Since the first publication on this topic',theprevalenceofMR has been thoroughly studied for different purposes. Most prevalence studies are designed for the planning of services and establish an 'ascertained' prevalence rate, which is the number of cases officially recorded by the authorities'.The 'true'prevalence rate is the total number of mentally retarded people in a population, whether or not they require services,and is defined by the prevalence of M R at birth and the mortality rate.For mild mental retardation (MWR, I& 50-70) ' the true prevalence rate is more difficult to estimate than for severe mental retardation (SMR, IQ <50).Very often ascertained prevalence rates are mistaken for true prevalence rates. Moreover, the estimates of both rates are influenced by the design of the study, the assessment criteria used, and the method applied for the identification of case^^-^.As a result, differences in prevalence rates might partly reflect the true variation over populations and partly reveal discrepancies between studies and in the interpretation of the prevalence measure used. For instance, different frequencies are yielded by uses of the organic, psychological and social WHO criteria, which are reflected in impairment, disability and handicaps. The prevalence rates observed range from two to 85 per 1000. According to the WHO, the true prevalence rate of total MR in industrialised countries comes close to 3%".O, but in the United States controversy exists over whether the rate is 1% versus 3%5.7.'0.1', whereas the Scandinavian countries claim that the 1 % figure is their true pre~alence~.'~.'~. 'The aim ofthis annotation is to establish valid estimates of the true prevalence rates for SMR and AIMR in children of school age and to elucidate the variation in prevalence rates. Therefore the methodology of prevalence studies performed since 1960 was critically evaluated and a distinction was made between ascertained and true prevalence estimates. Methods of selectionA computerised literature search was conducted on MED-LINE regarding publications from 1981 to 1995, using the keywords'MR' and 'occurrence'. Only a few original articles and reviews were found; most papers were traced through references listed in these reviewss.'.'" l i and by browsing through relevant journals.Thirteen papers were unobtainable and 31 studies were excluded in accordance with the following criteria: 1) the study was restricted to institutionalised cases; 2) the population size was not given; 3) the age group studied exceeded the age range 5-19 years (school age) and no age structure was specified; 4) the I& levels studied were not specified; 5) hIR could not be distinguished from other disabilities. If rates for a specific population were described more than once, only the most recent publication was selected.As a result,43 original articles were included in this review. These are summarised in Table 1,l8 wi...

IMPORTANCE Both low and high gestational weight gain have been associated with adverse maternal and infant outcomes, but optimal gestational weight gain remains uncertain and not well defined for all prepregnancy weight ranges. OBJECTIVES To examine the association of ranges of gestational weight gain with risk of adverse maternal and infant outcomes and estimate optimal gestational weight gain ranges across prepregnancy body mass index categories. DESIGN, SETTING, AND PARTICIPANTS Individual participant-level meta-analysis using data from 196 670 participants within 25 cohort studies from Europe and North America (main study sample). Optimal gestational weight gain ranges were estimated for each prepregnancy body mass index (BMI) category by selecting the range of gestational weight gain that was associated with lower risk for any adverse outcome. Individual participant-level data from 3505 participants within 4 separate hospital-based cohorts were used as a validation sample. Data were collected between 1989 and 2015. The final date of follow-up was December 2015. EXPOSURES Gestational weight gain. MAIN OUTCOMES AND MEASURES The main outcome termed any adverse outcome was defined as the presence of 1 or more of the following outcomes: preeclampsia, gestational hypertension, gestational diabetes, cesarean delivery, preterm birth, and small or large size for gestational age at birth. RESULTS Of the 196 670 women (median age, 30.0 years [quartile 1 and 3, 27.0 and 33.0 years] and 40 937 were white) included in the main sample, 7809 (4.0%) were categorized at baseline as underweight (BMI <18.5); 133 788 (68.0%), normal weight (BMI, 18.5-24.9); 38 828 (19.7%), overweight (BMI, 25.0-29.9); 11 992 (6.1%), obesity grade 1 (BMI, 30.0-34.9); 3284 (1.7%), obesity grade 2 (BMI, 35.0-39.9); and 969 (0.5%), obesity grade 3 (BMI, Ն40.0). Overall, any adverse outcome occurred in 37.2% (n = 73 161) of women, ranging from 34.7% (2706 of 7809) among women categorized as underweight to 61.1% (592 of 969) among women categorized as obesity grade 3. Optimal gestational weight gain ranges were 14.0 kg to less than 16.0 kg for women categorized as underweight; 10.0 kg to less than 18.0 kg for normal weight; 2.0 kg to less than 16.0 kg for overweight; 2.0 kg to less than 6.0 kg for obesity grade 1; weight loss or gain of 0 kg to less than 4.0 kg for obesity grade 2; and weight gain of 0 kg to less than 6.0 kg for obesity grade 3. These gestational weight gain ranges were associated with low to moderate discrimination between those with and those without adverse outcomes (range for area under the receiver operating characteristic curve, 0.55-0.76). Results for discriminative performance in the validation sample were similar to the corresponding results in the main study sample (range for area under the receiver operating characteristic curve, 0.51-0.79). CONCLUSIONS AND RELEVANCE In this meta-analysis of pooled individual participant data from 25 cohort studies, the risk for adverse maternal and infant outcomes varied by gestational wei...