Accelerated cognitive decline increases the risk of dementia. Slowing down the rate of cognitive decline leads to the preservation of cognitive functioning in the elderly, who can live independently for a longer time. Alcohol consumption may influence the rate of cognitive decline. The aim of the present study was to evaluate the associations between the total consumption of alcoholic beverages and different types of alcoholic beverages and cognitive decline at middle age. In 2613 men and women of the Doetinchem Cohort Study, aged 43–70 years at baseline (1995–2002), cognitive function (global cognitive function and the domains memory, speed and flexibility) was assessed twice, with a 5-year time interval. In linear regression analyses, the consumption of different types of alcoholic beverages was analysed in relation to cognitive decline, adjusting for confounders. We observed that, in women, the total consumption of alcoholic beverages was inversely associated with the decline in global cognitive function over a 5-year period (P for trend = 0·02), while no association was observed in men. Regarding the consumption of different types of alcoholic beverages in men and women together, red wine consumption was inversely associated with the decline in global cognitive function (P for trend < 0·01) as well as memory (P for trend < 0·01) and flexibility (P for trend = 0·03). Smallest declines were observed at a consumption of about 1·5 glasses of red wine per d. No other types of alcoholic beverages were associated with cognitive decline. In conclusion, only (moderate) red wine consumption was consistently associated with less strong cognitive decline. Therefore, it is most likely that non-alcoholic substances in red wine are responsible for any cognition-preserving effects.

Whole-grain intake has been reported to be associated with a lower risk of several lifestyle-related diseases such as type 2 diabetes, CVD and some types of cancers. As measurement errors in self-reported whole-grain intake assessments can be substantial, dietary biomarkers are relevant to be used as complementary tools for dietary intake assessment. Alkylresorcinols (AR) are phenolic lipids found almost exclusively in whole-grain wheat and rye products among the commonly consumed foods and are considered as valid biomarkers of the intake of these products. In the present study, we analysed the plasma concentrations of five AR homologues in 2845 participants from ten European countries from a nested case–control study in the European Prospective Investigation into Cancer and Nutrition. High concentrations of plasma total AR were found in participants from Scandinavia and Central Europe and lower concentrations in those from the Mediterranean countries. The geometric mean plasma total AR concentrations were between 35 and 41 nmol/l in samples drawn from fasting participants in the Central European and Scandinavian countries and below 23 nmol/l in those of participants from the Mediterranean countries. The whole-grain source (wheat or rye) could be determined using the ratio of two of the homologues. The main source was wheat in Greece, Italy, the Netherlands and the UK, whereas rye was also consumed in considerable amounts in Germany, Denmark and Sweden. The present study demonstrates a considerable variation in the plasma concentrations of total AR and concentrations of AR homologues across ten European countries, reflecting both quantitative and qualitative differences in the intake of whole-grain wheat and rye.

The protective effect of dietary carotenoid intake on the risk of breast cancer is inconclusive. Moreover, data on dietary carotenoids in relation to breast cancer in non-Western populations are scarce. The aim of the present study was to examine the association between dietary carotenoid intake and the risk of breast cancer among Chinese women. A total of 561 cases and 561 controls who were frequency matched by age (5-year interval) and residence were recruited in the present case–control study. Dietary intake information was collected by a face-to-face interview using a validated FFQ. The OR and 95 % CI were assessed by multivariate logistic regression after adjusting for various potential confounders. An inverse association was observed between the consumption of α-carotene, β-carotene, β-cryptoxanthin and lutein/zeaxanthin and the risk of breast cancer. The multivariate-adjusted OR for the highest quartile of intake compared with the lowest quartile of intake were 0·61 (95 % CI 0·43, 0·88) for α-carotene, 0·54 (95 % CI 0·38, 0·78) for β-carotene, 0·38 (95 % CI 0·26, 0·52) for β-cryptoxanthin and 0·49 (95 % CI 0·34, 0·71) for lutein/zeaxanthin. Lycopene intake was not found to be associated with the risk of breast cancer, with the adjusted OR of 0·89 (95 % CI 0·61, 1·30). These inverse associations were more evident among pre-menopausal women and women who were exposed to second-hand smoke. The protective effect of specific carotenoid intake was observed for all subtypes of hormone receptor status of breast cancer. The present study indicated that a greater intake of specific carotenoids was associated with a decreased risk of breast cancer among Chinese women residing in Guangdong.

In the present study, we investigated whether maternal exposure to a cafeteria diet affects the metabolism and body composition of offspring and whether such an exposure has a cumulative effect during the lifetime of the offspring. Female rats were fed a control (CON) or a cafeteria (CAF) diet from their own weaning to the weaning of their offspring. At 21 d of age, male offspring were divided into four groups by diet during gestation and after weaning (CON-CON, CON-CAF, CAF-CON and CAF-CAF). Blood was collected from dams (after weaning) and pups (at 30 and 120 d of age) by decapitation. CAF dams had significantly greater body weight and adipose tissue weight and higher concentrations of total cholesterol, insulin and leptin than CON dams (Student's t test). The energy intake of CAF rats was higher than that of CON rats regardless of the maternal diet (two-way ANOVA). Litters had similar body weights at weaning and at 30 d of age, but at 120 d, CON-CAF rats were heavier. At both ages, CAF rats had greater adipose tissue weight than CON rats regardless of the maternal diet, and the concentrations of TAG and cholesterol were similar between the two groups, as were blood glucose concentrations at 30 d of age. However, at 120 d of age, CAF rats were hyperglycaemic, hyperinsulinaemic and hyperleptinaemic regardless of the maternal diet. These findings suggest that maternal obesity does not modulate the metabolism of male offspring independently, modifying body weight only when associated with the intake of a cafeteria diet by the offspring.

Previous studies have suggested that the intake of trans-fatty acids (TFA) plays a role in the development of obesity. The proportions of adipose tissue fatty acids not synthesised endogenously in humans, such as TFA, usually correlate well with the dietary intake. Hence, the use of these biomarkers may provide a more accurate measure of habitual TFA intake than that obtained with dietary questionnaires. The objective of the present study was to investigate the associations between the proportions of specific TFA in adipose tissue and subsequent changes in weight and waist circumference (WC). The relative content of fatty acids in adipose tissue biopsies from a random sample of 996 men and women aged 50–64 years drawn from a Danish cohort study was determined by GC. Baseline data on weight, WC and potential confounders were available together with information on weight and WC 5 years after enrolment. The exposure measures were total trans-octadecenoic acids (18 : 1t), 18 : 1 Δ6-10t, vaccenic acid (18 : 1 Δ11t) and rumenic acid (18 : 2 Δ9c, 11t). Data were analysed using multiple regression with cubic spline modelling. The median proportion of total adipose tissue 18 : 1t was 1·52 % (90 % central range 0·98, 2·19) in men and 1·47 % (1·01, 2·19) in women. No significant associations were observed between the proportions of total 18 : 1t, 18 : 1 Δ6-10t, vaccenic acid or rumenic acid and changes in weight or WC. The present study suggests that the proportions of specific TFA in adipose tissue are not associated with subsequent changes in weight or WC within the exposure range observed in this population.

The objectives of the present study were to examine the associations between macronutrient intake and insulin sensitivity (IS) and insulin secretion (ISct), taking into consideration moderate-to-vigorous physical activity (MVPA), fitness and sedentary behaviour. Caucasian youth (n 630) aged 8–10 years at recruitment, with at least one obese biological parent, were studied (QUebec Adipose and Lifestyle InvesTigation in Youth cohort). IS was measured using the homeostasis model assessment (HOMA) of insulin resistance and Matsuda IS index. ISct was measured using HOMA2 %-β, the ratio of the AUC of insulin:glucose over the first 30 min (AUC I/Gt= 30min) of the oral glucose tolerance test and AUC I/Gt= 120min over 2 h. Fitness was measured using VO2peak, percentage of fat mass by dual-energy X-ray absorptiometry, and 7 d MVPA using accelerometry; screen time (ST) by average daily hours of self-reported television, video game or computer use. Dietary composition was measured using three non-consecutive dietary recalls. Non-parametric smoothing splines were used to model non-linear associations; all models were adjusted for age, sex, season, pubertal stage, MVPA, fitness, ST and adiposity. The percentage of total daily energy from dietary protein, fat, saturated fat and carbohydrate and the consumption of dietary vitamin D, sugar-sweetened beverages, fibre and portions of fruits and vegetables were taken into consideration. No dietary component was associated with any measure of IS after adjusting for MVPA, fitness, ST and adiposity. For every 1 % increase in daily protein intake (%), AUC I/Gt= 30min decreased by 1·1 % (P= 0·033). Otherwise, dietary composition was not associated with ISct. While long-term excess of energy intake has been shown to lead to overweight and obesity, dietary macronutrient composition is not independently correlated with IS or ISct in youth.

Bioactive milk proteins may be important in protecting preterm infants from developing inflammation and necrotising enterocolitis (NEC). A preterm pig model was used to investigate the protective effects of enteral bovine lactoferrin (bLF) against NEC development and inflammation. Caesarean-delivered preterm pigs were fed parenteral and minimal enteral nutrition for the first 2 d followed by 2 d of total enteral nutrition before euthanasia. Pigs were stratified into two groups and fed with either a control formula (CON, n 15) or a 10 g/l of bLF-enriched formula (LF, n 13). NEC incidence, gut functions and inflammatory cytokines were analysed. NEC incidence and nutrient absorption were similar between the two groups. In pigs that developed NEC, disease outcome was more severe in the colon accompanied by increased intestinal permeability in LF pigs. In contrary, the LF pigs had a lowered IL-1β level in the proximal small intestine. Dose-dependent effects of bLF on cell proliferation, intracellular signalling and cytokine secretion were tested in porcine intestinal epithelial cells (PsIc1) in vitro. Low doses (0·1–1 g/l) increased cell proliferation via extracellular signal-regulated kinase (ERK), limited IL-8 secretion and prevented NF-κB and hypoxia-inducible factor-1α (HIF-1α) activation, suggesting anti-inflammatory effects. In contrast, at a higher dose (10 g/l), bLF exerted adverse effects by reducing cell proliferation, stimulating IL-8 release, inhibiting ERK activation and up-regulating NF-κB and HIF-1α activation. Overall, at a dose of 10 g/l, bLF exacerbated disease severity in pigs that developed NEC, while the in vitro studies indicated the positive effects of bLF at low doses (0·1–1 g/l). Supplementation of infant formulas with bLF should therefore be optimised carefully.

Studies of diet and disease risk in India and among other Asian-Indian populations are hindered by the need for a comprehensive dietary assessment tool to capture data on the wide variety of food and nutrient intakes across different regions and ethnic groups. The nutritional component of the India Health Study, a multicentre pilot cohort study, included 3908 men and women, aged 35–69 years, residing in three regions of India (New Delhi in the north, Mumbai in the west and Trivandrum in the south). We developed a computer-based, interviewer-administered dietary assessment software known as the ‘NINA-DISH (New Interactive Nutrition Assistant – Diet in India Study of Health)’, which consisted of four sections: (1) a diet history questionnaire with defined questions on frequency and portion size; (2) an open-ended section for each mealtime; (3) a food-preparer questionnaire; (4) a 24 h dietary recall. Using the preferred meal-based approach, frequency of intake and portion size were recorded and linked to a nutrient database that we developed and modified from a set of existing international databases containing data on Indian foods and recipes. The NINA-DISH software was designed to be easily adaptable and was well accepted by the interviewers and participants in the field. A predominant three-meal eating pattern emerged; however, patterns in the number of foods reported and the primary contributors to macro- and micronutrient intakes differed by region and demographic factors. The newly developed NINA-DISH software provides a much-needed tool for measuring diet and nutrient profiles across the diverse populations of India with the potential for application in other South Asian populations living throughout the world.

We hypothesised that hypothalamic feeding-related neuropeptides are differentially expressed in obese-prone and lean-prone rats and trigger overeating-induced obesity. To test this hypothesis, in the present study, we measured energy balance and hypothalamic neuropeptide Y (NPY) and pro-opiomelanocortin (POMC) mRNA expressions in male JCR:LA-cp rats. We compared, in independent cohorts, free-feeding obese-prone (Obese-FF) and lean-prone (Lean-FF) rats at pre-weaning (10 d old), weaning (21–25 d old) and early adulthood (8–12 weeks). A group of Obese-pair-feeding (PF) rats pair-fed to the Lean-FF rats was included in the adult cohort. The body weights of 10-d-old Obese-FF and Lean-FF pups were not significantly different. However, when the pups were shifted from dams' milk to solid food (weaning), the obese-prone rats exhibited more energy intake over the days than the lean-prone rats and higher body and fat pad weights and fasting plasma glucose, leptin, insulin and lipid levels. These differences were consistent with higher energy consumption and lower energy expenditure. In the young adult cohort, the differences between the Obese-FF and Lean-FF rats became more pronounced, yielding significant age effects on most of the parameters of the metabolic syndrome, which were reduced in the Obese-PF rats. The obese-prone rats displayed higher NPY expression than the lean-prone rats at pre-weaning and weaning, and the expression levels did not differ by age. In contrast, POMC expression exhibited significant age-by-genotype differences. At pre-weaning, there was no genotype difference in POMC expression, but in the weanling cohort, obese-prone pups exhibited lower POMC expression than the lean-prone rats. This genotype difference became more pronounced at adulthood. Overall, the development of hyperphagia-induced obesity in obese-prone JCR rats is related to POMC expression down-regulation in the presence of established NPY overexpression.

Fat-soluble vitamins A, E and K have been shown to play roles in immunity and inflammation, but studies on child allergic disease have been few and inconsistent. The aim of the present study was to examine the relationship between maternal intake of vitamins A, E and K in mid-pregnancy and child asthma and allergic rhinitis. We used data on 44 594 mother–child pairs from the Danish National Birth Cohort. Maternal intake of fat-soluble vitamins was calculated based on the information from a validated FFQ completed in mid-pregnancy. At 18 months, interviews with the mothers were conducted to evaluate doctor-diagnosed child asthma. At age 7 years, we assessed child asthma and allergic rhinitis using questions from the International Study of Asthma and Allergies in Childhood questionnaire and by national registries on hospital contacts and medication use. Current asthma was defined as asthma diagnosis and wheeze in the past 12 months by maternal report. We calculated multivariable risk ratios and 95 % CI by comparing the highest v. lowest quintile (Q) of maternal vitamin A, E and K intake in relation to child allergic disease outcomes. Maternal total vitamin K intake was directly associated with ever admitted asthma (Q5 v. Q1: 1·23, 95 % CI 1·01, 1·50) and current asthma at 7 years (Q5 v. Q1: 1·30, 95 % CI 0·99, 1·70). Weak inverse associations were present for maternal vitamin A and E intake during pregnancy with child allergic rhinitis. Maternal vitamin K intake during pregnancy may increase the risk of child asthma, and should be explored further on a mechanistic level. Conversely, maternal vitamin A and E intake may protect against child allergic rhinitis.

Energy intake under-reporting (UR) is a concern in nutritional epidemiological studies, as it may distort the relationships between dietary habits and health. Although UR is known to be associated with certain characteristics, few studies have investigated them in France. Therefore, the goal of the present study was to assess the prevalence and characteristics of UR in French adults. UR was defined according to Goldberg's classification. A sample of 1567 adults was drawn from the nationally representative French dietary survey (Individuelle Nationale des Consommations Alimentaires 2 2006–7). Food intake (7 d record), dietary habits, socio-economic status, region of residence, sedentary behaviour and weight perception variables were assessed. Multivariate logistic regression was used to investigate the associations between UR and a number of covariates. The overall prevalence of UR was 22·5 %, similar in men and women. In both sexes, UR was positively associated with overweight and protein intake and inversely associated with age. In women, UR was associated with eating lunch in the office, poor perception of diet quality and sedentary behaviour and was inversely associated with educational level, residence in the Paris region, cereal product intake and eating lunch in a friend's or family member's home. In men, UR was positively associated with a history of slimming and inversely associated with dairy product intake and eating lunch at a staff canteen. In conclusion, UR is prevalent in French adults and is associated with several different characteristics. It is important to take account of UR when investigating diet–disease associations in adults.

The objective of the present study was to estimate ruminal feed N outflow in lactating cows using the omasal sampling, compartmental flux or in situ method. A total of five ruminally fistulated Finnish Ayrshire dairy cows were used in a 5 × 5 Latin square study with 21 d periods. Experimental silages of grass or red clover harvested at two stages of maturity in addition to a supplement of 9·0 kg concentrate/d were fed to the cows. In vivo omasal N flow was determined using the omasal sampling technique. Ruminal in situ N flow was calculated from N intake and degradability (38 μm nylon bags). The samples of ruminal contents and faeces were divided into seven particle-size fractions by wet sieving; the concentrations of indigestible neutral-detergent fibre and N were used to calculate N flow in the compartmental flux method. In vivo omasal N flow was greater for the red clover silage diets than for the grass silage diets. The N flow calculated using the compartmental flux technique and that calculated using the in situ technique were highly correlated, but both were less than and poorly correlated with the in vivo N flow. In both in situ and compartmental flux techniques, forage maturity increased the particle-associated N flow, with the increase being significantly greater for the red clover diets than for the grass silage diets. In conclusion, the compartmental flux and in situ methods described the N flow associated with the particle fractions rather than the total ruminal outflow of feed N.

Few longitudinal studies carried out in US adults have evaluated long-term dietary fat intakes and compared them with the national recommendations during the two-decade period when the prevalence of obesity and insulin resistance increased substantively. In the present study, we examined trends in the intakes of dietary fats and rich dietary sources of fats in the Framingham Heart Study Offspring Cohort over a 17-year period. The cohort was established in 1971–75 with follow-up examinations being conducted approximately every 4 years. Dietary data were collected using a semi-quantitative FFQ beginning in 1991 (exam 5). We included 2732 adults aged ≥ 25 years with complete dietary data in at least three examinations from 1991 to 2008. Descriptive statistics were generated using SAS version 9.3, and a repeated-measures model was used to examine trends in macronutrient and food intakes using R. Over the 17 years of follow-up, the percentage of energy derived from total fat and protein increased (27·3–29·8 % of energy and 16·8–18·0 % of energy, respectively) and that derived from carbohydrate decreased (51·0–46·8 % of energy; P-trend < 0·001). Increases in the percentage of energy derived from all fat subtypes were observed, except for that derived from trans-fats, which decreased over time (P-trend < 0·001). Trends were similar between the sexes, although women exhibited a greater increase in the percentage of energy derived from saturated fat and less reduction in the percentage of energy derived from trans-fats (P interaction < 0·05). Trends in fat intake were similar across the BMI categories. The number of weekly servings of cheese, eggs, ice cream desserts, nuts, butter and sausages/processed meats increased, whereas the intake of milk, margarine, poultry, confectioneries, chips and breads decreased (P-trend < 0·001). In this cohort of predominantly Caucasian older adults, the percentage of energy derived from dietary fats increased over time, but it remained within the national recommendations of less than 35 % of total energy, on average.

The present study investigated the impact of a Lactobacillus rhamnosus CGMCC1.3724 (LPR) supplementation on weight loss and maintenance in obese men and women over 24 weeks. In a double-blind, placebo-controlled, randomised trial, each subject consumed two capsules per d of either a placebo or a LPR formulation (1·6 × 108 colony-forming units of LPR/capsule with oligofructose and inulin). Each group was submitted to moderate energy restriction for the first 12 weeks followed by 12 weeks of weight maintenance. Body weight and composition were measured at baseline, at week 12 and at week 24. The intention-to-treat analysis showed that after the first 12 weeks and after 24 weeks, mean weight loss was not significantly different between the LPR and placebo groups when all the subjects were considered. However, a significant treatment × sex interaction was observed. The mean weight loss in women in the LPR group was significantly higher than that in women in the placebo group (P= 0·02) after the first 12 weeks, whereas it was similar in men in the two groups (P= 0·53). Women in the LPR group continued to lose body weight and fat mass during the weight-maintenance period, whereas opposite changes were observed in the placebo group. Changes in body weight and fat mass during the weight-maintenance period were similar in men in both the groups. LPR-induced weight loss in women was associated not only with significant reductions in fat mass and circulating leptin concentrations but also with the relative abundance of bacteria of the Lachnospiraceae family in faeces. The present study shows that the Lactobacillus rhamnosus CGMCC1.3724 formulation helps obese women to achieve sustainable weight loss.

Infants with slow weight gain cause concern in parents and professionals, but it is difficult to be certain whether such infants are genetically small or whether their energy intake is insufficient. The aim of the present study was to assess the impact of diet and feeding behaviours on slow weight gain early in infancy. The sample was 11 499 term infants from the Avon Longitudinal Study of Parents and Children (ALSPAC). A total of 507 cases of slow weight gain from birth to 8 weeks were identified and the remaining 10 992 infants were used as controls. It was found that infants who gained weight slowly between birth and 8 weeks were more likely to exhibit feeding problems such as weak sucking and slow feeding during this period. Feeding problems were substantially reduced during the recovery phase (8 weeks to 2 years) when these infants exhibited enhanced catch-up in weight. The proportion of mothers breast-feeding in the 4th week after birth was higher for slow weight gainers, but they were more likely to switch to formula at the start of recovery. During recovery, slow-weight gain infants had a slightly higher energy intake from formula and solids than controls. In conclusion, feeding problems seem to be the most important factors associated with the onset of early slow weight gain. Subsequently, a reduction of feeding problems and an increase in overall energy intake may contribute to their weight recovery. Health professionals should look for feeding problems in the first few weeks after birth and help mothers establish adequate feeding practices.

Few data exist on the aetiology of anaemia and Fe deficiency (ID) during early infancy in South Asia. The present study aimed to determine the contribution of ID, infections and feeding practices to anaemia in Bangladeshi infants aged 6–11 months. Baseline data from 1600 infants recruited into a cluster-randomised trial testing the effectiveness of micronutrient powder sales by frontline health workers on the prevalence of anaemia were used. Multivariate logistic regression was used to identify risk factors for anaemia and ID, and population attributable fractions (PAF) were computed to estimate the proportion of anaemia that might be prevented by the elimination of individual risk factors. It was found that 68 % of the infants were anaemic, 56 % were Fe deficient, and one-third had evidence of subclinical infections. The prevalence of anaemia and ID increased rapidly, until 8–9 months of age, while that of subclinical infections was constant. ID (adjusted OR (AOR) 2·6–5·0; P< 0·001) and subclinical infections (AOR 1·4–1·5; P< 0·01) were major risk factors for anaemia, in addition to age and male sex. Similarly, subclinical infections, age and male sex were significant risk factors for ID. Previous-day consumption of Fe-rich foods was very low and not associated with anaemia or ID. The PAF of anaemia attributable to ID was 67 % (95 % CI 62, 71) and that of subclinical infections was 16 % (95 % CI 11, 20). These results suggest that a multipronged strategy that combines improvements in dietary Fe intake alongside infection control strategies is needed to prevent anaemia during infancy in Bangladesh.

Red meat and dairy products contain high sialic acid (Sia) levels, but the metabolic fate and health impact in children remain unknown. The aims of the present study were to quantify the levels of urinary Sia N-acetylneuraminic acid (Neu5Ac), N-glycolylneuraminic acid (Neu5Gc) and ketodeoxynonulosonic acid (KDN) and to determine their relationship with dietary Sia intake. Spot urine samples were collected from 386 healthy children aged 3 (n 108), 4 (n 144) and 5 (n 134) years at 06.30–07.00, 11.30–12.00 and 16.30–17.00 hours. Food intake levels were recorded on the day of urine sample collection. Sia levels were quantified using LC–MS/MS with [13C3]Sia as an internal standard. We found that (1) total urinary Sia levels in healthy pre-school children ranged from 40 to 79 mmol Sia/mol creatinine; (2) urinary Sia levels were independent of age and consisted of conjugated Neu5Ac (approximately 70·8 %), free Neu5Ac (approximately 21·3 %), conjugated KDN (approximately 4·2 %) and free KDN (approximately 3·7 %); Neu5Gc was detected in the urine of only one 4-year-old girl; (3) total urinary Sia levels were highest in the morning and declined over time in 4- and 5-year-old children (P< 0·05), but not in 3-year-old children; (4) Sia intake levels at breakfast and lunch were approximately 2·5 and 0·16 mg Sia/kg body weight; and (5) there was no significant correlation between dietary Sia intake levels and urinary Sia levels. Urinary Sia levels varied with age and time of day, but did not correlate with Sia intake in 3- to 5-year-old children. The difference in urinary Sia levels in children of different age groups suggests that the metabolism and utilisation rates of dietary Sia are age dependent.

The present study aimed to explore previously unreported ethnic differences in infant feeding practices during the introduction of solid foods, accounting for maternal and birth factors, and to determine whether these feeding patterns are associated with BMI at 3 years of age. An observational study using Poisson regression was carried out to investigate the relationship between ethnicity and infant feeding practices and linear regression was used to investigate the relationship between feeding practices and BMI at 3 years of age in a subsample of 1327 infants in Bradford. It was found that compared with White British mothers, mothers of Other ethnicities were less likely to replace breast milk with formula milk before introducing solid foods (adjusted relative risk (RR) – Pakistani: 0·76 (95 % CI 0·64, 0·91), Other South Asian: 0·58 (95 % CI 0·39, 0·86), and Other ethnicities: 0·50 (95 % CI 0·34, 0·73)). Pakistani and Other South Asian mothers were less likely to introduce solid foods early ( < 17 weeks) (adjusted RR – Pakistani: 0·92 (95 % CI 0·87, 0·96) and Other South Asian: 0·87 (95 % CI 0·81, 0·93)). Other South Asian mothers and mothers of Other ethnicities were more likely to continue breast-feeding after introducing solid foods (adjusted RR – 1·72 (95 % CI 1·29, 2·29) and 2·12 (95 % CI 1·60, 2·81), respectively). Pakistani and Other South Asian infants were more likely to be fed sweetened foods (adjusted RR – 1·18 (95 % CI 1·13, 1·23) and 1·19 (95 % CI 1·10, 1·28), respectively) and Pakistani infants were more likely to consume sweetened drinks (adjusted RR 1·72 (95 % CI 1·15, 2·57)). No association between infant feeding practices and BMI at 3 years was observed. Although ethnic differences in infant feeding practices were found, there was no association with BMI at 3 years of age. Interventions targeting infant feeding practices need to consider ethnicity to identify which populations are failing to follow recommendations.