To determine the health-related quality of life (HRQOL) of overweight and obese multi-ethnic children compared with normal-weight children; and to investigate differences in HRQOL between self- and parent-proxy reports and ethnic groups.

Prospective cross-sectional study.

Out-patient clinic where children and their parents filled out a validated HRQOL questionnaire (KIDSCREEN-52) and height, weight, waist circumference and fat percentage were measured.

Overweight and obese children, aged 8–18 years (mean BMI Z-score 3·2 (sd 0·6)), from the obesity out-patient clinic.

Three hundred and eight self- and 213 parent-proxy reported questionnaires were completed. Global HRQOL and the Physical Wellbeing, Moods & Emotions and Self-Perception subscales were markedly reduced in our multi-ethnic obese cohort, relative to the Dutch reference values. Parent proxies reported significantly lower on the global HRQOL and the Physical Wellbeing, Moods & Emotions and Bullying subscales. In Caucasian children, multivariate analyses showed that BMI was associated with the quality-of-life subscales Moods & Emotions, Self-Perception and Bullying.

HRQOL was markedly reduced in our multi-ethnic overweight and obese out-patient clinic cohort, with significantly lower parent-proxy scores compared with self-reported scores. We believe intervention programmes aiming to improve HRQOL should be directed to both parents and children, while ethnic-specific programmes to enhance HRQOL seem of less importance.

To compare the performance of mid-upper arm circumference (MUAC) against weight-for-height Z-score (WHZ) for predicting inpatient deaths in children under 5 years of age.

Diagnostic test accuracy study.

Paediatric emergency department of a tertiary care hospital catering to semi-urban and rural population in Delhi, India.

Hospitalized children (n 1663) aged 6 months to 5 years, for whom discharge outcome was available, were consecutively recruited over 14 months. MUAC (cm), weight (kg) height (cm), clinical details and the outcome were recorded. MUAC (index test) was compared with WHZ based on the WHO growth standards (reference test) for predicting the outcome.

One hundred and twenty-four (7 %) children died during hospital stay. Both MUAC < 11·5 cm (adjusted OR (95 % CI): 3·7 (2·43, 5·60), P<0·001) and WHZ<−3 (2·0 (1·37, 2·99), P<0·001) served as independent predictors of inpatient mortality. However, MUAC was a significantly better predictor of mortality compared with WHZ in terms of area under the receiver-operating characteristic curve (MUAC=0·698, WHZ=0·541, P<0·001). MUAC<11·5 cm had the best trade-off of sensitivity and specificity for predicting inpatient mortality. A combination of WHZ<−3 and/or MUAC<11·5 cm did not significantly improve the predictive value over that of MUAC/WHZ, assessed individually.

MUAC<11·5 cm is a better predictor of mortality in hospitalized under-5 children, as compared with WHZ<−3. It should be measured in all emergency settings to identify the children at higher risk of death.

To study the quantity and quality of water intake from beverages among pre-schoolers and investigate associations with gender and socio-economic status (SES).

Kindergarten-based cross-sectional survey within the large-scale European ToyBox-study. A standardized protocol was used and parents/caregivers filled in sociodemographic data and a semi-quantitative FFQ.

Kindergartens in six European countries (Belgium, Bulgaria, Germany, Greece, Poland and Spain).

European pre-schoolers (aged 3·5–5·5 years) and their parents/caregivers (n 7051).

Mean water intake was 1051 ml/d; plain water, 547 ml/d; plain milk, 241 ml/d; other fruit juice, 104 ml/d; pure fruit juice, 59 ml/d; soft drinks, 55 ml/d; tea, 45 ml/d; sugared and chocolate milk, 37 ml/d; smoothies, 15 ml/d; and light soft drinks, 6 ml/d. Boys had a higher water intake than girls due to a higher consumption of plain water, but more importantly to the consumption of beverages of less quality. Lower-SES pre-schoolers scored better on quantity than high-SES pre-schoolers, but as a consequence of consumption of sugared beverages. Nevertheless, the associations differed by country.

The water intake from beverages did not meet the European Food Safety Authority standard of 1280 ml/d; especially in Western European countries water intake from beverages was low. The most important water sources were plain water, milk and fruit juices. Interventions aiming at a proper and sufficient water intake should focus on both quantity and quality. Messages about water and water sources should be clear for everyone and interventions should be sufficiently tailored.

Although existing evidence links breakfast frequency to better dietary quality, little is known specifically in regard to the benefits associated with eating breakfast together with one’s family. The present study describes the prevalence and experience of having family meals at breakfast among rural families and examines associations between meal frequency and adolescent diet quality.

Data were drawn from Project BreakFAST, a group-randomized trial aimed at increasing school breakfast participation in rural Minnesota high schools, USA. Linear mixed models were used to examine associations between student reports of family breakfast frequency and Healthy Eating Index 2010 (HEI-2010) scores while accounting for clustering within schools, demographics and household food security.

Adolescent students from sixteen schools completed online surveys, height and weight measurements, and dietary recalls at baseline in 2012–2014.

The sample included 827 adolescents (55·1 % girls) in grades 9–10 who reported eating breakfast on at most three days per school week.

On average, adolescents reported eating breakfast with their family 1·3 (sd 1·9) times in the past week. Family breakfast meals occurred most frequently in the homes of adolescents who reported a race other than white (P=0·002) or Hispanic ethnicity (P=0·02). Family breakfast frequency was directly associated with adolescent involvement in preparing breakfast meals (P<0·001) and positive attitudes (P≤0·01) about mealtime importance, interactions and structure. Family breakfast frequency was unrelated to most diet quality markers.

Family meals may be one important context of opportunity for promoting healthy food patterns at breakfast. Additional research is needed to better inform and evaluate strategies.

The aim of the present study was to validate figural drawing scales depicting extremely lean to extremely obese subjects to obtain proxies for BMI and waist circumference in postal surveys.

Reported figural scales and anthropometric data from a large population-based postal survey were validated with measured anthropometric data from the same individuals by means of receiver-operating characteristic curves and a BMI prediction model.

Adult participants in a Scandinavian cohort study first recruited in 1990 and followed up twice since.

Individuals aged 38–66 years with complete data for BMI (n 1580) and waist circumference (n 1017).

Median BMI and waist circumference increased exponentially with increasing figural scales. Receiver-operating characteristic curve analyses showed a high predictive ability to identify individuals with BMI > 25·0 kg/m2 in both sexes. The optimal figural scales for identifying overweight or obese individuals with a correct detection rate were 4 and 5 in women, and 5 and 6 in men, respectively. The prediction model explained 74 % of the variance among women and 62 % among men. Predicted BMI differed only marginally from objectively measured BMI.

Figural drawing scales explained a large part of the anthropometric variance in this population and showed a high predictive ability for identifying overweight/obese subjects. These figural scales can be used with confidence as proxies of BMI and waist circumference in settings where objective measures are not feasible.

To estimate the folate status of New Zealand women of childbearing age following the introduction, in 2010, of a new voluntary folic acid fortification of bread programme.

The 2011 Folate and Women’s Health Survey was a cross-sectional survey of women aged 18–44 years carried out in 2011. The survey used a stratified random sampling technique with the Electoral Roll as the sampling frame. Women were asked about consumption of folic-acid-fortified breads and breakfast cereals in a telephone interview. During a clinic visit, blood was collected for serum and erythrocyte folate measurement by microbiological assay.

A North Island (Wellington) and South Island (Dunedin) city centre in New Zealand.

Two hundred and eighty-eight women, of whom 278 completed a clinic visit.

Geometric mean serum and erythrocyte folate concentrations were 30 nmol/l and 996 nmol/l, respectively. Folate status was 30–40 % higher compared with women of childbearing age sampled as part of a national survey in 2008/09, prior to the introduction of the voluntary folic acid bread fortification programme. In the 2011 Folate and Women’s Health Survey, reported consumption of fortified bread and fortified breakfast cereal in the past week was associated with 25 % (P=0·01) and 15 % (P=0·04) higher serum folate concentrations, respectively.

Serum and erythrocyte folate concentrations have increased in New Zealand women of childbearing age since the number of folic-acid-fortified breads was increased voluntarily in 2010. Consumption of fortified breads and breakfast cereals was associated with a higher folate status.

The purpose of the present study was to examine the effects of food label nutrition colouring schemes in interaction with food category healthiness on consumers’ perceptions of food healthiness. Three streams of colour theory (colour attention, colour association and colour approach-avoidance) in interaction with heuristic processing theory provide consonant predictions and explanations for the underlying psychological processes.

A 2 (food category healthiness: healthy v. unhealthy)×3 (food label nutrient colouring schemes: healthy=green, unhealthy=red (HGUR) v. healthy=red, unhealthy=green (HRUG) v. no colour (control)) between-subjects design was used.

The research setting was a randomised-controlled experiment using varying formats of food packages and nutritional information colouring.

Respondents (n 196) sourced from a national consumer panel, USA.

The findings suggest that, for healthy foods, the nutritional colouring schemes reduced perceived healthiness, irrespective of which nutrients were coloured red or green (healthinesscontrol=4·86; healthinessHGUR=4·10; healthinessHRUG=3·70). In contrast, for unhealthy foods, there was no significant difference in perceptions of food healthiness when comparing different colouring schemes against the control.

The results make an important qualification to the common belief that colour coding can enhance the correct interpretation of nutrition information and suggest that this incentive may not necessarily support healthier food choices in all situations.

To examine associations between geographic measures of retail food outlets and perceived availability of healthy foods.

Cross-sectional.

A predominantly rural, eight-county region of South Carolina, USA.

Data from 705 household shoppers were analysed using ordinary least-squares regression to examine relationships between geographic measures (presence and distance) of food outlets obtained via a geographic information system and perceived availability of healthy foods (fresh fruits and vegetables and low-fat foods).

The presence of a supermarket within an 8·05 km (5-mile) buffer area was significantly associated with perceived availability of healthy foods (β=1·09, P=0·025) when controlling for all other food outlet types. However, no other derived geographic presence measures were significant predictors of perceived availability of healthy foods. Distances to the nearest supermarket (β=−0·16, P=0·003), dollar and variety store (β=−0·15, P=0·005) and fast-food restaurant (β=0·11, P=0·015) were all significantly associated with perceptions of healthy food availability.

Our results suggest that distance to food outlets is a significant predictor of healthy food perceptions, although presence is sensitive to boundary size. Our study contributes to the understanding and improvement of techniques that characterize individuals’ food options in their community.

To explore salt content in bread and to evaluate the feasibility of reducing salt contained in ‘pan francés’ bread.

The study had two phases. Phase 1, an exploratory phase, involved the estimation of salt contained in bread as well as a triangle taste test to establish the amount of salt to be reduced in ‘pan francés’ bread without detection by consumers. In Phase 2, a quasi-experimental, pre–post intervention study assessed the effects of the introduction of low-salt bread on bakery sales.

A municipal bakery in Miraflores, Lima, Peru.

Sixty-five clients of the bakery in Phase 1 of the study; sales to usual costumers in Phase 2.

On average, there was 1·25 g of salt per 100 g of bread. Sixty-five consumers were enrolled in the triangle taste test: fifty-four (83·1 %) females, mean age 58·9 (sd 13·7) years. Based on taste, bread samples prepared with salt reductions of 10 % (P=0·82) and 20 % (P=0·37) were not discernible from regular bread. The introduction of bread with 20 % of salt reduction, which contained 1 g of salt per 100 g of bread, did not change sales of ‘pan francés’ (P=0·70) or other types of bread (P=0·36). Results were consistent when using different statistical techniques.

The introduction of bread with a 20 % reduction in salt is feasible without affecting taste or bakery sales. Results suggest that these interventions are easily implementable, with the potential to contribute to larger sodium reduction strategies impacting the population’s cardiovascular health.

Carotid intima-media thickness (IMT) is a validated surrogate marker of preclinical atherosclerosis and is predictive of cardiovascular morbidity and mortality. Research on the association between IMT and diet, however, is lacking, especially in low-income countries or low-BMI populations.

Cross-sectional analysis. Dietary intakes were measured using a validated, thirty-nine-item FFQ at baseline cohort recruitment. IMT measurements were obtained from 2010–2011.

Rural Bangladesh.

Participants (n 1149) randomly selected from the Health Effects of Arsenic Longitudinal Study, an ongoing, population-based, prospective cohort study established in 2000. Average age at IMT measurement was 45·5 years.

Principal component analysis of reported food items yielded a ‘balanced’ diet, an ‘animal protein’ diet and a ‘gourd and root vegetable’ diet. We observed a positive association between the gourd/root vegetable diet and IMT, as each 1 sd increase in pattern adherence was related to a difference of 7·74 (95 % CI 2·86, 12·62) μm in IMT (P<0·01), controlling for age, sex, total energy intake, smoking status, BMI, systolic blood pressure and diabetes mellitus diagnoses. The balanced pattern was associated with lower IMT (−4·95 (95 % CI −9·78, −0·11) μm for each 1sd increase of adherence; P=0·045).

A gourd/root vegetable diet in this Bangladeshi population positively correlated with carotid IMT, while a balanced diet was associated with decreased IMT.

As ageing is associated with changes in body composition, BMI may not be the appropriate obesity measure for older adults. To date, little is known about associations between obesity measures and health-related quality of life (HRQoL). Thus, we aimed to compare different obesity measures in their association with HRQoL and self-rated physical constitution (SRPC) in older adults.

Seven obesity measures (BMI, waist circumference (WC), waist-to-hip ratio, waist-to-height ratio, fat mass percentage based on bioelectrical impedance analysis, hypertriglyceridaemic waist (HTGW) and sarcopenic obesity) were assessed at baseline in 2009. HRQoL, using the EQ-5D questionnaire, and SRPC, using one single question, were collected at baseline and at the 3-year follow-up in 2012. Linear and logistic regression analyses were used to examine the associations between the obesity measures and both outcomes. Model comparisons were conducted by area under the receiver-operating characteristic curve, R2, Akaike and Schwarz Bayesian information criteria.

KORA-Age study in Southern Germany (2009–2012).

Older adults (n 883; aged ≥65 years).

Nearly all obesity measures were significantly inversely associated with both outcomes in cross-sectional analyses. Concerning HRQoL, the WC model explained most of the variance and had the best model adaption, followed by the BMI model. Regarding SRPC, the HTGW and BMI models were best as rated by model quality criteria, followed closely by the WC model. Longitudinal analyses showed no significant associations.

These results suggest that, with regard to HRQoL/SRPC, simple anthropometric measures are sufficient to determine obesity in older adults in medical practice.

To compare the nutrient profile of packaged supermarket food products available in Australia and New Zealand. Eligibility to carry health claims and relationship between nutrient profile score and nutritional content were also evaluated.

Nutritional composition data were collected in six major Australian and New Zealand supermarkets in 2012. Mean Food Standards Australia New Zealand Nutrient Profiling Scoring Criterion (NPSC) scores were calculated and the proportion of products eligible to display health claims was estimated. Regression analyses quantified associations between NPSC scores and energy density, saturated fat, sugar and sodium contents.

NPSC scores were derived for 23 596 packaged food products (mean score 7·0, range −17 to 53). Scores were lower (better nutrient profile) for foods in Australia compared with New Zealand (mean 6·6 v. 7·8). Overall, 45 % of foods were eligible to carry health claims based on NPSC thresholds: 47 % in Australia and 41 % in New Zealand. However, less than one-third of dairy (32 %), meat and meat products (28 %) and bread and bakery products (27·5 %) were eligible to carry health claims. Conversely, >75 % of convenience food products were eligible to carry health claims (82·5 %). Each two-unit higher NPSC score was associated with higher energy density (78 kJ/100 g), saturated fat (0·95 g/100 g), total sugar (1·5 g/100 g) and sodium (66 mg/100 g; all P values<0·001).

Fewer than half of all packaged foods available in Australia and New Zealand in 2012 met nutritional criteria to carry health claims. The few healthy choices available in key staple food categories is a concern. Improvements in nutritional quality of foods through product reformulation have significant potential to improve population diets.

Bread is a staple of the Mediterranean diet but contributes substantially to its salt content (19 % in Spain). The objective of the present study was to assess the feasibility and acceptability of salt reduction in partially baked breads, partly replacing salt (NaCl) with a potassium salt, with subsequent follow-up.

During 2013, nine breads already on the market (1·8 % NaCl flour basis) had 0·5 % of NaCl replaced with potassium citrate (27·7 % reduction in sodium) and were commercialized in Spain. Later, breads were baked in bake-off stores and sold ready-to-eat to consumers. This market test was evaluated by comparing the sales between standard- v. reduced-salt breads and the complaints related to flavour attributes. The wholesalers involved in the market test were then surveyed.

Spain.

The market test confirmed good acceptance of the reduced-salt breads, as 2013 sales were 3678 tonnes v. 2012 sales of 3577 tonnes for the same standard breads. No complaints were received. The wholesaler survey showed, in general, little awareness of salt reduction.

It is feasible that potassium citrate can reduce the salt content of bread without negatively affecting sales or complaints. This shows potential for introducing this type of bread on a larger scale.

To describe trends in country- and individual-level dual burden of malnutrition in children <5 years, and age-stratified (<2 years, ≥2 years) country-level trends, in thirty-six low- and middle-income countries (LMIC).

Using repeated cross-sectional nationally representative data, we calculated the prevalence of malnutrition (stunting, wasting, overweight) at each survey wave, annualized rates of prevalence change for each country over time, and trends before and after 2000, for all children <5 years and separately for those </≥2 years. We examined country- (ratio of stunting to overweight) and individual-level (coexistence of stunting and overweight) dual burden in children <5 years.

Demographic and Health Surveys from thirty-six LMIC between 1990 and 2012.

Children <5 years.

Overall malnutrition prevalence decreased in children <5 years, driven by stunting decreases. Stunting rates decreased in 78 % of countries, wasting rates decreased in 58 % of countries and overweight rates increased in 36 % of countries. Rates of change differed for children </≥2 years, with children <2 years experiencing decreases in stunting in fewer countries yet increases in overweight in more countries. Countries with nearly equal prevalences of stunting and overweight in children <5 years increased from 2000 to the final year. Within a country, 0·3–10·9 % of children <5 years were stunted and overweight, and 0·6–37·8 % of stunted children <5 years were overweight.

The dual burden exists in children <5 years on both country and individual levels, indicating a shift is needed in policies and programmes to address both sides of malnutrition. Children <2 years should be identified as a high-risk demographic.

Breakfast consumption has been consistently associated with health outcomes and cognitive functioning in schoolchildren. Evidence of direct links with educational outcomes remains equivocal. We aimed to examine the link between breakfast consumption in 9–11-year-old children and educational outcomes obtained 6–18 months later.

Data on individual-level free school meal entitlement and educational outcomes (Statutory Assessment Tests (SATs) at Key Stage 2) were obtained via the SAIL databank and linked to earlier data collected on breakfast consumption. Multilevel modelling assessed associations between breakfast consumption and SATs.

Trial of the Primary School Free Breakfast Initiative in Wales.

Year 5 and 6 students, n 3093 (baseline) and n 3055 (follow-up).

Significant associations were found between all dietary behaviours and better performance in SATs, adjusted for gender and individual- and school-level free school meal entitlement (OR=1·95; CI 1·58, 2·40 for breakfast, OR=1·08; CI 1·04, 1·13 for healthy breakfast items). No association was observed between number of unhealthy breakfast items consumed and educational performance. Association of breakfast consumption with educational performance was stronger where the measure of breakfast consumption was more proximal to SATs tests (OR=2·02 measured 6 months prior to SATs, OR=1·61 measured 18 months prior).

Significant positive associations between self-reported breakfast consumption and educational outcomes were observed. Future research should aim to explore the mechanisms by which breakfast consumption and educational outcomes are linked, and understand how to promote breakfast consumption among schoolchildren. Communicating findings of educational benefits to schools may help to enhance buy-in to efforts to improve health behaviours of pupils.

The main indicator adopted to track universal salt iodization has been the coverage of adequately iodized salt in households. Rapid test kits (RTK) have been included in household surveys to test the iodine content in salt. However, laboratory studies of their performance have concluded that RTK are reliable only to distinguish between the presence and absence of iodine in salt, but not to determine whether salt is adequately iodized. The aim of the current paper was to examine the performance of RTK under field conditions and to recommend their most appropriate use in household surveys.

Standard performance characteristics of the ability of RTK to detect the iodine content in salt at 0 mg/kg (salt with no iodine), 5 mg/kg (salt with any added iodine) and 15 mg/kg (‘adequately’ iodized salt) were calculated. Our analysis employed the agreement rate (AR) as a preferred metric of RTK performance.

Twenty-five data sets from eighteen population surveys which assessed household iodized salt by both the RTK and a quantitative method (i.e. titration or WYD Checker) were obtained from Asian (nineteen data sets), African (five) and European (one) countries.

In detecting iodine in salt at 0 mg/kg, the RTK had an AR>90 % in eight of twenty-three surveys, while eight surveys had an AR<80 %. When the RTK was used for detecting adequately iodized salt, the AR decreased significantly, with only one of fourteen surveys achieving an AR>90 %.

The RTK is not suited for assessment of adequately iodized salt coverage. Quantitative assessment, such as by titration or WYD Checker, is necessary for estimates of adequately iodized salt coverage.

Excessive free sugars consumption has a possible role in health issues, diet quality and obesity development. The present cross-sectional study aimed to identify the major food sources of free sugars in Greek children’s diet and investigate possible associations of dietary patterns with free sugars intake.

Anthropometric measurements and information on dietary and physical activity habits were obtained. Energy and free sugars intake coming from foods were estimated and principal components analysis was applied to identify dietary patterns.

The GRECO (Greek Childhood Obesity) study.

Nationwide sample of 3089 children (aged 10–12 years).

Adopting WHO criteria, 44·2 % of participants were categorized as having free sugars intake above 10 % of total energy intake. Mean contribution of free sugars to energy intake was 11·2 %, and the major food sources of free sugars differed from those of other childhood populations. Free sugars intake was not associated with overweight/obesity. Multiple linear regression analysis revealed that two lifestyle and dietary patterns, characterized by higher consumption of sweets, fast foods, fries, sugared drinks, frequently ordering/eating outside home and having meals in front of a screen (pattern 1) and higher consumption of whole fruits, 100 % fruit juices, vegetables, legumes and honey/jam (pattern 2), were positively associated with free sugars intake.

A large proportion of children exceeded the recommended cut-off and free sugars intake was associated with lifestyle patterns rather than single foods. Public health programmes aiming to reduce free sugars consumption should be tailored on promoting the correct dietary habits of specific childhood populations.

Nutrient and food standards exist for school lunches in English primary schools although packed lunches brought from home are not regulated. The aim of the present study was to determine nutritional and dietary differences by lunch type.

A cross-sectional survey was carried out in 2007 assessing diet using the Child and Diet Evaluation Tool (CADET), a validated 24 h estimated food diary. The data were analysed to determine nutritional and dietary intakes over the whole day by school meal type: school meals and packed lunches.

Fifty-four primary schools across England.

Children (n 2709) aged 6–8 years.

Children having a packed lunch consumed on average 11·0 g more total sugars (95 % CI 6·6, 15·3 g) and 101 mg more Na (95 % CI 29, 173 mg) over the whole day. Conversely, children having a school meal consumed, on average, 4·0 g more protein (95 % CI 2·3, 5·7 g), 0·9 g more fibre (NSP; 95 % CI 0·5, 1·3 g) and 0·4 mg more Zn (95 % CI 0·1, 0·6 mg). There was no difference in daily energy intake by lunch type. Children having a packed lunch were more likely to consume snacks and sweetened drinks; while children having a school meal were more likely to consume different types of vegetables and drink water over the whole day.

Compared with children having a school meal, children taking a packed lunch to school consumed a lower-quality diet over the whole day, including higher levels of sugar and Na and fewer vegetables. These findings support the introduction of policies that increase school meal uptake.

In many countries of the world millions of people are not registered at birth. However, in order to assess children’s nutritional status it is necessary to have an exact knowledge of their age. In the present paper we discuss the effects of insufficient or imprecise age data on estimates of undernutrition prevalence.

Birth registration rates and levels of stunting, underweight and wasting were retrieved from Multiple Indicator Cluster Surveys and Demographic and Health Surveys of thirty-seven sub-Saharan African countries, considering the subdivision in wealth quintiles. The composition of the cross-sectional sample used for nutritional evaluation was analysed using a permutation test. Logistic regression was applied to analyse the relationship between birth registration and undernutrition. The 95 % probability intervals and Student’s t test were used to evaluate the effect of age bias and error.

Heterogeneous sampling designs were detected among countries, with different percentages of children selected for anthropometry. Further, registered children were slightly more represented within samples used for nutritional analysis than in the total sample. A negative relationship between birth registration and undernutrition was recognized, with registered children showing a better nutritional status than unregistered ones, even within each wealth quintile. The over- or underestimation of undernutrition in the case of systematic over- or underestimation of age, respectively, the latter being more probable, was quantified up to 28 %. Age imprecision was shown to slightly overestimate undernutrition.

Selection bias towards registered children and underestimation of children’s age can lead to an underestimation of the prevalence of undernutrition.

Non-compliance with food record submission can induce bias in nutritional epidemiological analysis and make it difficult to draw inference from study findings. We examined the impact of demographic, lifestyle and psychosocial factors on such non-compliance during the first 3 years of participation in a multidisciplinary prospective paediatric study.

The Environmental Determinants of Diabetes in the Young (TEDDY) study collects a 3 d food record quarterly during the first year of life and semi-annually thereafter. High compliance with food record completion was defined as the participating families submitting one or more days of food record at every scheduled clinic visit.

Three centres in the USA (Colorado, Georgia/Florida and Washington) and three in Europe (Finland, Germany and Sweden).

Families who finished the first 3 years of TEDDY participation (n 8096).

High compliance was associated with having a single child, older maternal age, higher maternal education and father responding to study questionnaires. Families showing poor compliance were more likely to be living far from the study centres, from ethnic minority groups, living in a crowded household and not attending clinic visits regularly. Postpartum depression, maternal smoking behaviour and mother working outside the home were also independently associated with poor compliance.

These findings identified specific groups for targeted strategies to encourage completion of food records, thereby reducing potential bias in multidisciplinary collaborative research.