To create, validate and assess the reliability of a checklist to measurecalcium intake in children.

Calcium intakes from a checklist and parent-assisted 24-h dietary recall werecompared. Checklist reliability was assessed separately.

After-school programmes in the United States.

Forty-two children (18 males, 24 females, age = 8.0 ± 0.9 years)participated in the validation analysis and 49 children (28 males, 21females, age = 7.5 ± 0.9 years) in the reliability analysis.

No differences in mean calcium intakes were found by method or gender. Thechecklist correlated well with recall among girls (r = 0.65, P = 0.01) but notboys (r = −0.33, P = 0.19). Agreement over time was above 80%for most foods.

The calcium checklist is useful for assessing calcium intake among groups of6–10-year-old children in settings that preclude parentalassistance. More research is needed to improve accuracy among boys.

To examine whether the self-reported fruit and vegetable consumption of Greek primary-school children is associated with the home availability of fruit and vegetables.

Cross-sectional study.

Five primary schools in the city of Pyrgos in south-west Greece.

One hundred and sixty-seven students in fifth and sixth grades.

Fruit and vegetable consumption was assessed using an FFQ. Home availability of fruit and vegetables was assessed using a modified version of a US home availability questionnaire. Participant BMI was assessed and parental education obtained by self-report. Hierarchical regression models that took account of the clustering of participants in schools were used to examine the relationship between consumption and availability after controlling for parental education and BMI.

Regression analyses showed that home availability of fruit was a significant predictor of consumption (β = 0·524, Z = 9·77, P < 0·001) in a model that explained 46 % of the overall variance. Home availability of vegetables was a significant predictor of vegetable intake (β = 0·378, Z = 6·23, P < 0·001) in a model that explained 29 % of the overall variance.

Home availability of fruit and vegetables was associated with consumption among Greek primary-school children. The association between home availability of fruit and intake was higher than the associations that have been reported in other samples. Results support focusing on the home availability of fruit and vegetables as a potential mediator of consumption that could be manipulated as a means of increasing fruit and vegetable intake among children.

Community iron supplementation programmes for pregnant women have lacked effectiveness, partly because of low compliance.

To determine factors that influence compliance among pregnant women in Senegal.

Two hundred and twenty-one pregnant women, recruited from six health centres in Dakar during their first prenatal visit, were randomly assigned to receive either a prescription to purchase iron/folic acid tablets (control, n = 112) to be taken daily, according to official policy, or to receive free tablets (treatment, n = 109). Compliance was assessed 20 weeks after enrolment through interviews and pill count. Women with low or high compliance (<70% or ≥70%) were asked to explain what influenced their adherence to supplementation.

Overall compliance was 69%; it was significantly higher in the treatment than in the control group (86% vs. 48%; P < 0.0001). Women with high compliance (58%) were motivated by: (1) the perception of improved health upon taking the tablets (treatment = 24%, control = 10%); (2) the insistence by midwives that they take the tablets; and (3) the mention that the tablets would improve health. Women with low compliance (42%) reported: (1) the experience of side-effects that they associated with the tablets (treatment = 13%, control = 14%); (2) misunderstanding that they needed to continue taking the tablets throughout pregnancy (treatment = 0%, control = 18%); and (3) forgetfulness.

Compliance with iron/folic acid supplementation in Senegal can be increased by providing women with clear instructions about tablet intake and educating them on the health benefits of the tablets.

Parents’ child-feeding behaviours have been implicated in children’s food choices and weight, but little is known about the social class distribution of parent’s child-feeding behaviours in the UK. The present study compares parents’ self-reported child-feeding behaviours in two socio-economically contrasting areas.

A cross-sectional survey using the Parental Feeding Style Questionnaire. Mean scores were calculated for five child-feeding behaviours: control over eating, emotional feeding, encouragement/prompting, instrumental feeding and restriction. Parents’ self-reported child-feeding behaviours were compared with their sociodemographic characteristics.

Three primary schools in two contrasting electoral wards of Sheffield, UK.

Two hundred and ten parents of children aged 4 to 11 years, recruited from a convenience sample.

Parents in the least deprived ward reported using all five types of child-feeding behaviour more frequently than parents in the most deprived ward. After adjusting for parent sex, parent age, single parent status, employment status and level of education, emotional feeding was the only behaviour showing any evidence of a difference between wards. The most frequently used behaviours were control, encouragement and restriction – behaviours that might be used to directly influence children’s food intake and weight.

Child-feeding behaviours differ between areas within a single city and within a largely white population, and this distribution is related to socio-economic and educational factors. Experimental and longitudinal studies are needed to further investigate the potential role of child-feeding behaviours in childhood overweight and obesity.

To assess the prevalence of zinc inadequacy based on dietary intakes and plasma zinc concentrations and, simultaneously, the prevalence of inadequate intakes of energy, protein, calcium and iron.

A cross-sectional study of a convenience sample of subsistence farming households in Sidama, Southern Ethiopia.

Dietary intakes were calculated from 1-day weighed food records and 40 repeats from 99 pregnant women in the third trimester using analysed values of major staple foods for zinc, iron, calcium and phytate. The distribution of observed intakes was adjusted for usual intakes and the prevalence of inadequacy estimated using the estimated average requirement (EAR) cutpoint method. Prevalence of inadequacy for zinc, protein and iron intakes were compared with those based on biochemical measures.

Prevalence of zinc inadequacy was very high: 99% for US FNB EAR and 100% for IZiNCG EAR compared to 72% based on low plasma zinc concentrations. Corresponding prevalence estimates for iron were much lower: 4% for inadequate intakes based on US FNB EAR vs. 8.7% for iron deficiency anaemia (haemoglobin < 110 g l−1; ferritin < 12 μg l−1) and 32.3% for low storage iron. Prevalence of inadequacy for protein was 100% for adjusted intakes and 91% for serum albumin < 32 g l−1. For calcium, 74% were at risk for inadequate intakes.

The high prevalence of inadequate intakes of zinc and protein was reasonably consistent with those based on biochemical measures. Such dietary deficits could be overcome by regular consumption of cellular animal protein. In contrast, both dietary and biochemical measures of iron inadequacy were low.

We investigated associations between consumption of fruits, vegetables, sweets and soft drinks and the psychological dimensions of eating in parents and their children. The role of the parent’s characteristics for their children’s food intake was also explored.

Food intake patterns were assessed by self-reported consumption of the respective foods. Eating behaviour was measured by the Dutch Eating Behaviour Questionnaire and self-esteem by Harter self-perception scale. The participants were 1441 families (mother, father and their 12-year-old child), and additionally 354 mothers and thirty fathers.

Among parents, reported intake of fruit and vegetables were associated with restrained eating, higher self-esteem, and higher education and age. Intake of sweets was related to more external and less restrained eating, and for mothers also emotional eating. Parent’s intake of soft drink was foremost related to a younger age, and also weakly associated with psychological characteristics. The food intake of parents was more important for the children’s food intake than any other characteristics. However, children’s intake of sweets showed clear-cut positive associations with external eating.

Psychological dimensions of eating behaviour are associated with patterns of food intake, in particular for consumption of sweets, and are most prominent in the parents. The children’s food intake mirrored their parents’ intake. Being sensitive to external food cues may increase unhealthy food consumption in our society, whereas more restrained eating may indicate proneness or intention to healthier food choices among parents. Emotional eating may imply a proneness to consume sweets for comfort, in particular among mothers.

The aim of this study was to examine whether there are associations between working conditions and the use of staff canteen or packed meals among Finnish employees.

Data were obtained from cross-sectional surveys on working conditions, conducted triennially (1997, 2000, 2003) since 1997.

In each survey, the subjects were 25–64-year-old employed Finnish employees: 3096 men and 3273 women.

Employees at large workplaces used canteens far more often than those at smaller workplaces. Working conditions played a different role in canteen use at small and large workplaces, as well as among the different sexes. At small workplaces, physically demanding jobs held by female employees and low job control encouraged employees to use the canteen. On the other hand, at large workplaces, low social support at work encouraged the use of canteens among men whereas high mental strain at work meant they used the canteen less. Among women, eating packed meals was not related to working conditions, but among men, low social support and high mental strain at work were associated with more frequent use of packed meals.

The use of a staff canteen is largely determined by the size of the workplace and by employee education. The underlying factor could be the availability of canteens, a question which must be confirmed in further studies, since well-planned mass catering at workplaces has major effects on public health, well-being and the nutrition education of employees.

To assess participants’ acquaintance with and willingness to try healthful food alternatives, and to test the psychometric properties of an adapted Dutch version of the Food Neophobia Scale (FNS) in order to study the role of food neophobia in this context.

A cross-sectional study incorporating two web-based questionnaires, including a retest of the FNS one week later. Measures included acquaintance with and willingness to try 15 healthful food alternatives, level of food neophobia, level of education, gender and age. Multiple linear regression analyses were used to study associations between demographics and level of food neophobia as well as associations between level of food neophobia and acquaintance with and willingness to try the healthful alternatives.

The study was conducted in The Netherlands using a representative Internet panel.

A total of 326 participants aged 18–50 years participated.

Internal consistency and test–retest reliability of the FNS version used were sufficient. On average participants were acquainted with 7.9 of the products and modestly willing to try the products. Lowly educated participants had significantly higher FNS scores than highly educated participants (β = −0.23, P < 0.01). FNS score was significantly associated with acquaintance with (β = −0.21, P < 0.001) and willingness to try the healthful alternatives (β = −0.26, P < 0.001).

Further research into the role of food neophobia is warranted when wanting to stimulate the integration of healthful alternative products in the daily diet, especially among persons with low education.

To establish the school eating habits of Mexican children, who are prone to obesity and later to high rates of adult chronic diseases.

Questionnaires for students and parents with staff questionnaires and interviews.

Randomly sampled schools in a socio-economically representative district of Mexico City.

Subjects were 1504 adolescents aged 10–19 years attending schools in Mexico City, 27 teachers and seven headmasters, sampled from both public and private schools and from the full range of socio-economic groups.

Foods brought from home were of a higher nutritional quality than those purchased at school, where purchases were dominated by crisps, soft drinks and other items with high energy density. Girls were more inclined to purchase inappropriately; those from poorer homes purchased less. Private-school students irrespective of socio-economic grade brought more food from home and purchased more expensive food at school. School policies allowed food and drink vendors to market any products within the schools, which benefited financially from these activities.

Current school food policies are conducive to amplifying the current epidemic of obesity and related adult chronic diseases, and need to change.

To assess the nutritional status of the Andean population of Puna and Quebrada of Humahuaca, Jujuy, using anthropometric measurements.

A cross-sectional nutritional survey was carried out in a representative sample (n = 1236) of individuals from these regions. Children aged 2–9 years, adolescents (10–17 years) and adults (≥18 years; pregnant and lactating women excluded) were considered. Height-for-age, weight-for-height and body mass index (BMI) were calculated in children and adolescents and compared with World Health Organization/National Center for Health Statistics/Centers for Disease Control and Prevention reference standards using Z-scores or percentiles, in order to assess the prevalence of stunting, wasting/thinness and excess weight. In adults, BMI, waist circumference (WC) and waist-to-hip (WHR) ratio were used to identify obesity and central adiposity.

Stunting (height-for-age Z-score <–2 standard deviations) and obesity (BMI ≥ 95th percentile) were found to be major nutritional problems in children and adolescents. Stunting was prevalent in 10.7% of children and 12.4% of adolescents; 8.2% of children and 3.5% of adolescents were obese. Adults were short (mean: 155.8 cm) and values of overweight (32.3%), obesity (18.3%) and central adiposity (mean WC: 86.5 cm) were high. Older adults and those with higher economic development showed higher prevalence of obesity and central adiposity.

The present population may be at the early stages of nutritional transition as symptoms of undernutrition and overnutrition coexist at the population level. These results suggest that rates of growth retardation may be decreasing owing to improved nutritional conditions; however, this could be accompanied by a sharp increase in the prevalence of other diet-related chronic diseases.

To validate a diet history questionnaire (DHQ) using a weighed food record(WFR) as the standard method in the estimation of food consumption andnutrient intake in a group of adults.

WFR: all foods consumed by subjects during 7 consecutive days were weighedand recorded by nutrition students. Two DHQ interviews were carried out ondays 1 (first diet history questionnaire, DHQ1) and 28 (second diet historyquestionnaire, DHQ2).

Costa Rica.

Sixty adults: 30 men and 30 women; 30 living in urban and 30 in ruralareas.

In comparison to the WFR, the DHQ1 gave statistically significant higherestimates of the mean intake of 19 nutrients for men and of three nutrientsfor women. The uncorrected correlation coefficients for nutrient intakeaccording to both methods ranged from 0.40 to 0.83 for males and from 0.22to 0.62 for females. Percentage of subjects classified in the same quartilesof nutrient intake according to each method ranged from 33.3% to 63.3% formales and from 23.3% to 53.3% for females. Misclassification in extremequartiles ranged from 0% to 13.3% for both sexes. The mean food groupconsumption, according to the DHQ1, when compared with the WFR, gavestatistically significant differences for three of the 18 food groups formen and for two groups in the case of women. The two applications of the DHQgave similar results.

Validation of a DHQ using a WFR as the standard method gave results thatcompare favourably with those reported by other authors. This study foundimportant differences in the response of men and women to the DHQ: amongmen, the estimates of mean nutrient intake from DHQ1 were significantlygreater than those of the WFR, while in the case of women, the mean nutrientintake estimates from both methods were not significantly different. Therewas a higher degree of correlation between the DHQ1 and the WFR meannutrient intakes among men than among women. The DHQ showed goodreproducibility.

To assess the proportion of children being stunted and underweight-for-age at 3, 9 and 15 months in Lambaréné, Gabon, using the WHO child growth standards released in 2006 as compared with the Centers for Disease Control and Prevention (CDC) 2000 and the National Center for Health Statistics (NCHS) 1978 child growth charts/references.

Prospective birth cohort in Lambaréné, Gabon.

Two hundred and eighty-nine children from birth to 15 months of age.

Weight and length were recorded at 3, 9 and 15 months. Corresponding Z scores for stunting and underweight-for-age were calculated for the three different standards/references. Children with a height-for-age or weight-for-age below –2 sd of the corresponding reference median (Z score ≤−2) were classified as stunted or underweight-for-age, respectively.

With the new WHO 2006 standards a higher proportion (4·0 %) of 3-month-old infants were underweight compared with the CDC (1·0 %) or the NCHS (0·7 %) child growth charts/references. In contrast to the NCHS references or the CDC charts, this proportion did not increase from 3 to 9 months or from 9 to 15 months. The proportion of children being stunted was highest (above 20 %) with the WHO 2006 standards at all three ages. Again, in contrast to the old standards, this proportion did not increase from 3 to 9 months or from 9 to 15 months.

The present results show considerably different growth faltering patterns for Gabonese children depending on the growth charts used to assess the prevalence of stunting and underweight. Shifting to the new WHO child growth standards may have important implications for child health programmes.

The aim of the present study was to examine the mediation effects of changes in psychosocial determinants of physical activity (attitude, social support, self-efficacy, perceived benefits and barriers) on changes in physical activity.

One-year intervention study with baseline and 1-year post measures of physical activity habits and psychosocial correlates.

Fifteen middle schools.

Boys and girls (n = 2840) aged 11–15 years completed the validated questionnaires during class hours.

The product-of-coefficients test was used to asses the mediating effects. Self-efficacy for physical activity at school was found to be the only significant mediator of physical activity change. Specifically, self-efficacy for physical activity at school partly mediated the effect of the intervention on total and school-related physical activity change in the intervention group with parental support (P < 0.05). None of the other potential mediators, attitudes, social support, perceived benefits and perceived barriers, seemed to have had a positive effect. Even a suppressor effect was found for attitudes. Given that the effects of self-efficacy and attitudes were of opposite direction, the total mediated/suppressed effects of the intervention were not statistically significant.

Positive changes in total and school-related physical activity in adolescents could be partly explained by increases in self-efficacy for physical activity at school through a physical activity intervention in middle schools with parental support. However, the suppressor effect of attitudes decreased this effect. As this is one of the first true mediation analyses in this age group, further research is needed to replicate the importance of these mediators.

The aim of this study was to assess the behavioural effects of a school-based healthful diet promotion intervention implemented in lower vocational schools.

A cluster-randomised pre-test–post-test experimental design was used, and data were collected by means of written questionnaires.

Students of 18 Dutch lower vocational schools.

The final sample consisted of 10 experimental schools with 13 teachers, 37 classes and 879 students, and eight control schools with 10 teachers, 31 classes and 734 students (total n = 1613).

Fixed regression analyses revealed beneficial effects on the behavioural measures relating to fruit intake (as assessed by a food frequency measure and fruit consumption during the previous day), and most behavioural measures related to high-fat snack intake (food frequency measure, and number of snacks and total fat intake from snacks during the previous day). The behavioural effects relating to breakfast habits were limited. Some positive behavioural effects occurred in the total target population, others mainly or only among students with more unfavourable intakes at baseline. Mixed regression analysis found comparable regression coefficients for the behavioural outcomes, but the effects related to fruit intake were no longer statistically significant.

Although we did not find the expected effects on all outcome indicators, the total pattern of results suggest that Krachtvoer offered a surplus value over existing curricula. However, some programme elements need to be revised to improve effectiveness and prevent negative effects, and more information is needed about the long-term effects of the programme.

Animal studies and a few clinical trials lend credibility to the hypothesis that not all types of fatty acids carry the same potential for weight gain. Only a few epidemiological studies concerning this issue are currently available and results are conflicting.

The purpose of the present ecological study was to test the existence of an association between obesity prevalence and the types of fat available in 168 countries.

Data on the prevalence of obesity (BMI ≥ 30 kg/m2) for women over 15 years of age were obtained from the WHO Global InfoBase. Food balance sheets for the years 1998 to 2002 were obtained from the FAOSTAT database. Five-year means for energy, total fat, MUFA, PUFA, SFA and ‘other fat’ per capita were calculated, with their standard deviations, for each country. Bivariate correlations and a multiple linear regression model were used to test for the association between prevalence of obesity and types of fat available in these countries.

Not surprisingly, dietary energy supply, SFA, PUFA and ‘other fat’ were positively associated with the prevalence of obesity. We also found, however, a strong negative association between MUFA availability and obesity prevalence (β = −0·68, P < 0·0001).

Populations with a lower prevalence of obesity seem to consume a greater amount of MUFA. Considering the partial correlations between variables, our results suggest that in countries with higher obesity prevalence, it is the shift from MUFA to PUFA that particularly appears to be associated with the risk of obesity.

To characterise the diet of First Nations in north-western Ontario, highlightfoods for a lifestyle intervention and develop a quantitative food-frequencyquestionnaire (QFFQ).

Cross-sectional survey using single 24 h dietary recalls.

Eight remote and semi-remote First Nations reserves in north-westernOntario.

129 First Nations (Oji-Cree and Ojibway) men and women aged between 18 and 80years.

The greatest contributors to energy were breads, pasta dishes and chips(contributing over 20 % to total energy intake). ‘Addedfats’ such as butter and margarine added to breads and vegetablesmade up the single largest source of total fat intake (8·4 %).The largest contributors to sugar were sugar itself, soda and othersweetened beverages (contributing over 45 % combined). The mean number ofservings consumed of fruits, vegetables and dairy products were much lowerthan recommended. The mean daily meat intake was more than twice thatrecommended. A 119-item QFFQ was developed including seven bread items, fivesoups or stews, 24 meat- or fish-based dishes, eight rice or pasta dishes,nine fruits and 14 vegetables. Frequency of consumption was assessed byeight categories ranging from ‘Never or less than one time in onemonth’ to ‘two or more times a day’.

We were able to highlight foods for intervention to improve dietary intakebased on the major sources of energy, fat and sugar and the low consumptionof fruit and vegetable items. The QFFQ is being used to evaluate a diet andlifestyle intervention in First Nations in north-western Ontario.

BMI and waist circumference (WC) are used to screen for cardio-metabolic risk; however it is unclear how well these indices perform in populations subject to childhood stunting.

To evaluate BMI and WC as indicators of cardio-metabolic risk and to determine optimal cut-off points among 1325 Guatemalan adults (44 % stunted: ≤150 cm women; ≤162 cm men).

Cardio-metabolic risk factors were systolic/diastolic blood pressure ≥130/≥85 mmHg, glucose ≥5·5 mmol/l, TAG ≥1·7 mmol/l, ratio of total cholesterol to HDL-cholesterol ≥5·0, and the presence of two or more and three or more of the preceding risk factors. Receiver operating characteristic (ROC) curve analysis was used.

Areas under the ROC curve were in the range of 0·59–0·77 for BMI and 0·59–0·78 for WC among men and 0·66–0·72 and 0·64–0·72 among women, respectively. Optimal cut-off points for BMI were 24·7–26·1 kg/m2 among men (24·5–26·1 kg/m2 stunted; 24·8–26·3 kg/m2 non-stunted) and 26·5–27·6 kg/m2 among women (26·3–27·8 kg/m2 stunted; 26·6–27·9 kg/m2 non-stunted). Optimal cut-off points for WC were 87·3–91·1 cm among men (85·3–89·4 cm stunted; 88·5–93·3 cm non-stunted) and 91·3–95·3 cm among women (90·9–94·4 cm stunted; 91·8–95·6 cm non-stunted).

Optimal cut-off points for BMI were slightly higher among women than men with no meaningful differences by stature. Optimal cut-off points for WC were several centimetres lower for stunted compared with non-stunted men, and both were substantially lower than the current recommendations among Western populations. Cut-off points derived from Western populations may not be appropriate for developing countries with a high prevalence of stunting.

To investigate the determinants of mild-to-moderate malnutrition in preschoolers.

Cross-sectional study conducted in October and November 1996, with a representative sample of 1740 children less than 5 years old from the city of Salvador, situated in the Brazilian Northeastern region. Socio-economic and dietary data were collected through a structured questionnaire. Anthropometric measures were performed in duplicate and data analysis was based upon the hierarchical model approach. Logistic regression analysis was used to estimate the prevalence ratio and to identify the determinants of mild-to-moderate deficits in weight-for-age and height-for-age Z-scores.

Family monthly income under US$67.00 per capita and family headed by a woman were the main basic determinants of mild-to-moderate weight-for-age and height-for-age deficits in the studied children. Household agglomeration, an underlying determinant, was associated with weight-for-age and height-for-age deficits. Among the immediate determinants, age above 6 months and dietary caloric availability in the lowest tertile (<930 kcal day−1) were also associated with weight-for-age deficits. In addition to these, hospitalisation in the 12 months preceding the interview was shown to be a predictor of mild-to-moderate weight-for-age and height-for-age deficits.

Adverse social and economic factors interact with family environmental factors to define food consumption and morbidity patterns that culminate in a high prevalence of mild-to-moderate malnutrition. The strengthening and restructuring of nutrition and healthcare actions, the definition of public policies that improve family income, and the adequate insertion of women in the labour market are possible strategies to reduce mild-to-moderate malnutrition and to sustain the decline already observed in severe malnutrition.

To assess the ability of anthropometric measurements to identify young women at risk of developing diabetes, hypertension and heart disease in the future and to compare cut-off points for common anthropometric measures established with receiver-operating characteristic (ROC) curves with those reported in the literature.

Cross-sectional study.

Eight hundred and two young Mexican women living in semi-urban poverty.

The ability of anthropometric measures of fatness and fat distribution (body mass index (BMI), summed skinfold thickness (SST), waist circumference (WC), waist-to-hip ratio (WHR), conicity index (CI), abdominal volume index (AVI)) to predict risk of future disease (pre-diabetes: fasting blood glucose 100–126 mg dl−1; pre-hypertension: systolic blood pressure 120–139 mmHg and/or diastolic blood pressure 80–89 mmHg; hypertriglyceridaemia: triglycerides ≥150 mg dl−1; or a combination of risk factors) was assessed using ROC curve analysis.

Twenty-three of the 802 women who were interviewed had incomplete data and 50 (6.4%) were eliminated from the analysis due to hypertension and/or diabetes. Mean age of the remaining 729 women was 29.6 ± 5.4 years and mean BMI was 27.7 ± 4.5 kg m−2. There were no significant differences in the area under the ROC curve for BMI, WC, AVI or SST for any of the four outcomes. However, these indices performed significantly better than WHR and CI (P < 0.05). The BMI cut-off points that maximised sensitivity and specificity for the four outcomes were in the range of 27.7–28.4 kg m−2, and for WC were 89.3–91.2 cm. To detect 90% of the cases of any metabolic alteration, the necessary BMI cut-off was 26.1 kg m−2. Younger women (<25 years) were at greater risk than older women for a given BMI increment (P < 0.05).

We found that BMI and WC cut-off points commonly used for the identification of risk of existing disease were also appropriate in this population for the identification of risk in the future among women without diabetes or hypertension. The early identification of at-risk individuals, prior to the onset of disease, is fundamental particularly in the context of a country with scarce resources that is rapidly undergoing nutrition transition.

To assess which socio-economic indicator best predicts overweight in the European Union: educational attainment, occupational class or household income.

The prevalence of overweight is strongly related to socio-economic position. The relative importance of different socio-economic dimensions is uncertain, and might vary between countries.

Cross-sectional self-report data of the European Community Household Panel were obtained from nine countries (n 52 855; age 25–64 years). Uni- and multivariate regression analyses were employed to predict overweight (BMI ≥ 25 kg/m2) in relationship to socio-economic indicators. Occupational class was measured using the new European Socioeconomic Classification.

Large socio-economic differences in overweight were observed in all countries, especially for women. For both sexes, a low educational attainment was the strongest predictor of overweight. After controlling for education, overweight was negatively related to household income in women, but positively in men. Similar patterns were found for occupational class. For women, but not for men, educational inequalities in overweight were generally greater in Southern European countries. A similar pattern of inequalities in overweight was observed for all ages between 25 and 64 years.

Across Europe, overweight was more strongly and more consistently related to educational attainment than to occupational class or household income. People with lower educational attainment should be a specific target group for programmes and policies that aim to prevent overweight.