To investigate associations between alcohol consumption patterns and diet quality, nutrient intakes and biochemical profile of women of childbearing age.

Nutrient intake data from 24 h diet recalls, alcohol consumption data and diet quality from the Dietary Habits Questionnaire, and biochemical analyses from the cross-sectional 2008/09 Adult Nutrition Survey in New Zealand.

New Zealand households.

New Zealand women aged 18–45 years (n 1124).

All analyses were completed using Stata and survey weights were used to allow for the complex survey design to produce population estimates. Multinomial logistic regression models were used to examine the associations between drinking patterns and the variables of interest, with ‘infrequent moderate/light’ drinkers being the reference category. The findings indicate that alcohol-consuming women of childbearing age tend to replace food energy with alcohol energy (P = 0·022). ‘More frequent heavy’ and ‘more frequent moderate/light’ drinkers had higher intakes of total, mono- and polyunsaturated fats with the latter group also consuming higher levels of saturated fats (P < 0·05). Women who were ‘infrequent moderate/light’ drinkers had relatively better diet quality, nutrient intakes and adequate biochemical status in comparison to other drinkers and abstainers. ‘Infrequent heavy’ drinkers, who were predominantly younger in age, had lower serum vitamin B12 levels (P = 0·01) with a higher proportion of women in this category having below-recommended levels of serum folate (P < 0·05).

Alcohol consumption, especially heavy drinking patterns, may compromise nutritional status of women of childbearing age.

The 2009 American Recovery and Reinvestment Act (ARRA) increased monthly Supplemental Nutrition Assistance Program (SNAP) benefits and expanded SNAP eligibility, yet limited evidence exists on the potential impact of ARRA on dietary intake among at-risk individuals. We aimed to examine pre-/post-ARRA differences in food insecurity (FI) and dietary intake by SNAP participation status.

Pre/post analysis.

Boston, MA, USA.

Data were from the longitudinal Boston Puerto Rican Health Study (2007–2015). The US Department of Agriculture ten-item adult module assessed FI. A validated FFQ assessed dietary intake. Diet quality was assessed using the Alternate Healthy Eating Index-2010 (AHEI-2010). Self-reported pre-/post-ARRA household SNAP participation responses were categorized as: sustained (n 249), new (n 95) or discontinued (n 58). We estimated differences in odds of FI and in mean nutrient intakes and AHEI-2010 scores post-ARRA.

Compared with pre-ARRA, OR (95 % CI) of FI post-ARRA were lower for all participants (0·69 (0·51, 0·94)), and within sustained (0·63 (0·43, 0·92)) but not within new (0·94 (0·49, 1·80)) or discontinued (0·63 (0·25, 1·56)) participants. Post-ARRA, total carbohydrate intake was higher, and alcohol intake was lower, for sustained and new participants, and dietary fibre was higher for sustained participants, compared with discontinued participants. Scores for AHEI-2010 and its components did not differ post-ARRA, except for lower alcohol intake for sustained v. discontinued participants.

Post-ARRA, FI decreased for sustained participants and some nutrient intakes were healthier for sustained and new participants. Continuing and expanding SNAP benefits and eligibility likely protects against FI and may improve dietary intake.

To examine changes in the spatial clustering of malnutrition in children under 5 years of age (under-5s) for the period 1999 to 2011 in Bangladesh.

We used data from four nationally representative Demographic and Health Surveys (DHS) conducted in 1999–2000, 2004, 2007 and 2011 in Bangladesh involving a total of 24 211 under-5s located in 1661 primary sampling units (PSU; geographical unit of analysis) throughout Bangladesh. The prevalence of stunting (height/length-for-age Z-score <−2), underweight (weight-for-age Z-score <−2) and wasting (weight-for-height/length Z-score <−2) at each PSU site and for each survey year were estimated based on the WHO child growth standard. The extent of spatial clustering was quantified using semivariograms.

Whole of Bangladesh.

Children under 5 years of age.

Our results demonstrate that in 1999–2000 most PSU throughout Bangladesh experienced stunting, underweight and wasting prevalence which exceeded the WHO thresholds. By 2011, this situation improved, although in two of the six divisions (Barisal and Sylhet) PSU still exhibited higher levels of malnutrition compared with other divisions of the country. The pattern of spatial clustering for stunting, underweight and wasting also changed between 1999 and 2011 both at national and sub-national (division) levels.

We identified divisions where malnutrition indicators (stunting, underweight and wasting) remain highly clustered and other divisions where they are more widely spread in Bangladesh. This has important implications on how interventions for malnutrition need to be delivered (geographically targeted interventions v. random interventions) within each division of the country.

To describe the nutrient intakes of an Italian cohort of infants at 6, 9 and 12 months of age.

Dietary data were collected using a food diary at three follow-ups (6, 9 and 12 months of age of infants). The infants’ dietary data were used to estimate nutrient intakes using the Italian food composition database integrated with data from nutritional labels and the literature. The mean and standard deviation, median and interquartile range, minimum and maximum, and 5th, 25th, 75th and 95th percentiles were calculated for the daily intake of twenty-eight nutrients, with sex differences evaluated using parametric/non-parametric statistical methods.

A prospective population-based birth cohort.

Infants (n 400) living in the urban area of Trieste (Italy).

The sex distribution was fairly balanced at each follow-up. The mean daily intakes of energy and the other twenty-seven nutrients considered were greater in males at all follow-ups. In particular, a significant statistical difference was observed in higher male consumption of cholesterol at 9 months and in energy and carbohydrate intakes at 12 months (P < 0·05). The mean daily intake of proteins was greater than that recommended by the Italian Dietary Reference Values at all follow-ups.

These preliminary results provide a useful basis for understanding the nutrient intake patterns of infants in this area of Italy during the first year of life.

There have been few studies investigating the association between food security and breast-feeding duration and none have been conducted among Canadian Inuit, a population disproportionately burdened with food insecurity. We evaluated the association between household food security and breast-feeding duration in Canadian Inuit children.

Data were obtained from the Nunavut Inuit Child Health Survey, a population-based cross-sectional survey.

The Canadian Territory of Nunavut in 2007 and 2008.

Caregivers of Inuit children aged 3–5 years. Participating children were randomly sampled from community medical centre lists.

Out of 215 children, 147 lived in food-insecure households (68·4 %). Using restricted mean survival time models, we estimated that children in food-secure households were breast-fed for 16·8 (95 % CI 12·5, 21·2) months and children in food-insecure households were breast-fed for 21·4 (95 % CI 17·9, 24·8) months. In models adjusting for social class, traditional knowledge and child health, household food security was not associated with breast-feeding duration (hazard ratio=0·82, 95 % CI 0·58, 1·14).

Our research does not support the hypothesis that children living in food-insecure households were breast-fed for a longer duration than children living in food-secure households. However, we found that more than 50 % of mothers in food-insecure households continued breast-feeding well beyond 1 year. Many mothers in food-secure households also continued to breast-feed beyond 1 year. Given the high prevalence of food insecurity in Inuit communities, we need to ensure infants and their caregivers are being adequately nourished to support growth and breast-feeding, respectively.

To evaluate the nutrient quality of breads and breakfast cereals identified using the wholegrain definition of ≤10:1 carbohydrate:fibre ratio.

Following a cross-sectional study design, nutritional information was systematically gathered from food labels of breads and breakfast cereals that met the ≤10:1 carbohydrate:fibre criterion. The median nutrient content was compared with the UK Food Standards Agency’s nutrient profiling standards and the associations between carbohydrate:fibre ratio and other nutrients were analysed. Subgroup analyses were undertaken for products with and without fruit, nuts and/or seeds.

Products from four major supermarket stores in the UK.

Breads (n 162) and breakfast cereals (n 266).

Breads which met the ≤10:1 criterion typically contained medium fat, low saturated fat, low sugar and medium Na. Breakfast cereals typically contained medium fat, low saturated fat, high sugar and low Na. In both groups, as the carbohydrate:fibre ratio decreased, fat content increased (bread: P=0·029, r=−0·171; breakfast cereal: P=0·033, r=−0·131) and, in breakfast cereals, as the ratio increased, sugar content increased (P<0·0005, r=0·381). Breakfast cereals with fruit, nuts and/or seeds contained, per 100 g, more energy (P=0·002), fat, saturated fat and sugar (all P<0·0005), while seeded breads had more energy, fat and saturated fat (all P<0·0005).

Overall, breads and breakfast cereals meeting the ≤10:1 criterion have good nutritional quality, suggesting that the criterion could be useful in public health and/or food labelling. The utility of applying the ≤10:1 criterion to products containing fruit, nuts and/or seeds is less clear and requires further research.

To examine structured expert judgement (SEJ) elicitation as a method to provide robust, defensible data for three determinants of household food security (food cost, household disposable income and physical access) for quantifying a proof-of-concept integrating decision support system for food security.

SEJ elicitation is a validated method for obtaining unavailable data, but its use in household food security in high-income countries is novel. Investigate Discuss Estimate Aggregate (IDEA) elicitation protocol was implemented, including quantitative and qualitative elements. Using specific questions related to three determinants, food security experts were encouraged to Investigate – estimate individual first-round responses to these questions, Discuss – with each other evidence on the reasoning and logic of their estimates, Estimate – second-round responses, following which these judgements were combined using mathematical Aggregation.

Victoria, Australia.

Five experts with a range of expertise in the area of household food insecurity participated in the SEJ elicitation process.

The experts’ ability to provide reliable estimates was tested and informed the aggregation of the collection of individual estimates into a single quantity of interest for use in decision support. The results of the quantitative elicitation show the impact of combinations of varying household income, food cost and physical access on household food security status and severity and is supported by the experts reasoning during elicitation.

This research provides insight to the application of SEJ where elicited data can inform and support intervention decision-making specific to household food security, especially where evidence is absent or of poor quality.

To evaluate self-assessed knowledge about diabetes mellitus, to assess determinants of health knowledge and to evaluate consequences of health knowledge on appraisal about consequences of the disease.

Population-based computer-assisted web interview survey, supplemented with a paper-and-pencil survey via post.

Representative sample of the general Austrian population aged 15 years and older.

Men (n 1935) and women (n 2065) with and without diabetes mellitus.

Some 20·5 % of men and 17·7 % of women with diabetes, and 46·2 % of men and 36·7 % of women without diabetes, rated their knowledge about diabetes mellitus to be ‘very bad’ or ‘rather bad’. Individuals with diabetes and individuals with a family member with diabetes rated their information level more often as ‘very good’ or ‘rather good’, with adjusted OR (95 % CI) of 1·7 (1·1, 2·8) and 2·1 (1·6, 2·7), respectively, in men and 2·7 (1·5, 4·8) and 2·7 (2·1, 3·5), respectively, in women. Additional significant influencing factors on diabetes knowledge were age and educational level in both sexes, and city size in men. Independent of personal diabetes status, diabetes knowledge was associated with a lower perception of restrictions on daily life of diabetes patients and with a lower probability of underestimating health consequences of diabetes.

Health knowledge is associated with fewer misconceptions and less underestimation of health consequences in individuals both with and without diabetes mellitus. Thus health information about diabetes is important on the individual level towards disease management as well as on the public health level towards disease prevention.

Diets closer aligned with nutritional guidelines could lower the risk of several chronic conditions and improve economic outcomes, such as employment and healthcare costs. However, little is known about the range, order of magnitude and timing of these potential effects.

We used a microsimulation approach to predict US population changes over 30 years in health and economic outcomes that could result from a substantial (but not impossible) improvement in diet quality – an improvement from the third to the fifth quintile of US scores on the Alternate Healthy Eating Index, 2010 version.

Risk ratios from the literature for diabetes, heart disease and stroke were used to modify the Future Adult Model (FAM) to simulate outcomes from a higher-quality diet. Model parameter uncertainty was assessed using bootstrap and sensitivity analysis examined the variation in published risk ratios.

FAM simulates outcomes for the US adult population aged 25 and older.

Improved diet quality initially leads to very small changes in chronic disease prevalence, but these accumulate over time. If diets improved beginning in 2019, after 30 years diabetes prevalence could be reduced by 5·9 million cases (11·5 %), heart disease prevalence by 4·0 million cases (7·2 %) and stroke prevalence by 1·9 million cases (10·3 %). These reductions in disease prevalence would be accompanied that same year by fewer deaths (88 000) and healthcare cost savings of $144·0 billion (2019 USD).

This microsimulation study suggests that improvements in diet are likely to improve health and economic population outcomes over time.

To examine the effects of lunches with different dietary energy densities on food preferences between genders.

Randomized crossover study. Participants were administered the following packed test meals once weekly on a specified day during six sessions: control (150 g of rice with a sautéed beef entrée containing 40 g of raw beef and 240 g of vegetables), high-meat/low-rice, low-vegetable, medium-fat/low-vegetable, high-fat and high-fat/low-vegetable meals. Subjective levels of sensory properties were assessed over time using visual analogue scales.

University of Tokushima Graduate School, Tokushima, Japan.

Sixty-five men and sixty-five women matched by age and BMI.

Men showed significantly stronger desires for salty and fatty foods after meals (P<0·05). Women showed a significantly stronger desire for sweetness from 2 h after the low-vegetable meal, and increasing fat content under high-vegetable conditions caused a significant stimulated sweetness desire in women more than in men (P<0·05). Moreover, after a high-meat/low-rice meal with 100 g of rice, sweetness desire was stronger in women (P=0·024), whereas no significant differences in sweetness desire were shown between genders after another low-energy-density control meal with 150 g of rice.

Men had significantly stronger desires for salty and fatty foods, whereas women preferred sweet food after meals. The sweetness desire in women was stimulated by increasing fat content, even with a high vegetable intake. Low rice intake in a low-energy-density diet also caused a relative stimulation of sweetness desire in women.

To compare the prevalence of overweight or obesity (ow/ob) with WHO BMI cut-off points, International Obesity Task Force (IOTF) cut-off points and Chinese BMI criteria and examine its potential factors among preschool children in Hunan Province.

A cross-sectional survey including anthropometric measurements and questionnaires about children’s information, caregivers’ socio-demographic characteristics and maternal characteristics. χ2 tests and univariate and multivariate binary logistic regression were performed to evaluate the possible factors of ow/ob.

Hunan, China, from September to October 2019.

In total, 7664 children 2 to 6 years of age.

According to Chinese BMI criteria, about 1 in 7–8 children aged 2–6 years had ow/ob in Hunan, China. The overall estimated prevalence of ow/ob among 2- to 6-year-old children was significantly higher when based on the Chinese BMI criteria compared with the WHO BMI cut-off points and IOTF cut-off points. According to Chinese BMI criteria, ow/ob was associated with residing in urban areas, older age, male sex, eating snacking food more frequently, macrosomia delivery, caesarean birth, heavier maternal prepregnancy weight and pre-delivery weight.

The prevalence of ow/ob in preschool children in Hunan Province remains high. More ow/ob children could be screened out according to Chinese BMI cut-offs compared with WHO and IOTF BMI criteria. In the future, targeted intervention studies with matched controls will be needed to assess the long-term effects of intervention measures to provide more information for childhood obesity prevention and treatment.

Unhealthy diet is a modifiable risk factor leading to subclinical arterial damage (SAD), high BP and CVD. It was aimed to investigate the possible associations of dietary patterns (DPs) with SAD in adults having multiple CVD risk factors.

Dietary intake was evaluated through two 24-h dietary recalls and principal component analysis was used to identify DPs. Oscillometry, applanation tonometry with pulse wave analysis and carotid ultrasound were used to assess peripheral and aortic BP, arterial stiffness and pressure wave reflections.

Laiko University Hospital, Athens, Greece.

A total of 470 individuals (53·1 ± 14·2 years) with CVD risk factors were enrolled.

A pattern characterised by increased consumption of whole-grain cereals, white meat and reduced consumption of sugar was positively associated with common carotid compliance (β = 0·01, 95 % CI 0·00, 0·01), whereas a pattern high in refined cereals, red and processed meat was positively associated with brachial but not aortic systolic pressure (β = 1·76, 95 % CI 0·11, 3·42) and mean arterial pressure (MAP) (β = 1·18, 95 % CI 0·02, −2·38). Low consumption of low-fat dairy products, high consumption of full-fat cheese and butter was positively associated with MAP (β = 0·97, 95 % CI 0·01, 1·95). Increased consumption of vegetables, fruits, fresh juices, fish and seafood was inversely associated with augmentation index (AIx) (β = -1·01, 95 % CI -1·93, −0·09).

Consumption of whole grains, white meat, fruits/vegetables, fish/seafood and avoidance of sugar was associated with improved SAD. Preference in refined grains, red/processed meat, high-fat cheese/butter and low intake of low-fat dairy products were associated with BP elevation. Future studies are needed to confirm the present findings.

Survival rates for paediatric cancers have increased dramatically since the 1970s, but childhood cancer survivors (CCS) are at increased risk for several chronic diseases throughout life. Nutrition interventions promoting healthy family meals may support wellness for survivors, but little research has explored CCS family food preparation habits. The goal of the present study was to describe and compare food preparation practices of CCS and non-CCS families.

Observational.

Typical evening meal preparation events were observed and recorded in participant homes. Recordings and notes were analysed using the Healthy Cooking Index (HCI), a measure of nutrition-optimizing food preparation practices relevant to survivor wellness. Demographics, BMI and nutrient composition of prepared meals were also collected.

Forty parents with a CCS or non-CCS child aged 5–17 years were recruited.

There were no major differences between the CCS and non-CCS families with regard to summative HCI score or specific food preparation behaviours. Meals prepared by CCS and non-CCS families had similar nutrient compositions.

The study revealed areas for practical nutrition intervention in CCS and non-CCS families. Future studies should consider adopting and tailoring nutrition intervention methods that have been successful in non-CCS communities.

To estimate the prevalence of thinness, overweight and obesity among Tibetan adolescents aged 12–17 years.

Cross-sectional survey.

Shigatse City of Tibet municipality, with an average altitude of more than 4000 m.

Study participants included 2642 adolescents aged 12–17 years selected from six schools using a convenient cluster sampling method.

The prevalence of thinness/overweight/obesity among Tibetan adolescents was 9·4 %/5·4 %/1·4 % (China definition), 14·7 %/4·4 %/0·7 % (International Obesity Task Force (IOTF) definition), and 2·8 %/5·7 %/0·9 % (WHO definition). The prevalence of thinness and overweight was significantly different between both sexes based on each of three BMI classification criteria (P < 0·001). There was no significant difference in the prevalence of obesity between both sexes according to each of three BMI criteria. There was no clear trend in the prevalence of thinness across ages according to the China or IOTF definition (both P > 0·05), whereas an upward trend was observed for thinness in boys according to the IOTF definition (P for trend <0·05). In contrast, the prevalence of thinness tended to decrease with increasing age in girls according to the IOTF definition and in total sample according to the WHO definition (P for trend <0·05).

Among Tibetan adolescents, the prevalence of overweight and obesity is relatively low, while the prevalence of thinness is high, especially in boys. These data suggest urgent attention is needed to control adolescent thinness in Tibet.

To examine the difference in the rehabilitation rate from underweight by child age at enrolment in the Positive Deviance (PD)/Hearth programme.

This secondary data analysis used programme monitoring records of underweight children aged 6–60 months attending a 2-week PD/Hearth session and followed up for 6 months from September 2018 to March 2019. Data were analysed using multilevel mixed-effect regression and Poisson regression with robust variance.

Rajshahi Division, Bangladesh.

A total of 5227 underweight (weight-for-age Z-score (WAZ) <-2) children attended the PD/Hearth sessions.

From enrolment to 6 months follow-up, the mean WAZ improved from −2·80 to −2·09, and the percentage of underweight children decreased to 54·5 %. Compared to the enrolment age of 6–11 months, the estimated monthly change in WAZ at 6 months of follow-up were 0·05 lower for 12–23 months, 0·06 lower for 24–35 months, and 0·09 lower for 36–60 months of the enrolment age (all P < 0·001). The probability of rehabilitation at 6 months of follow-up were lower by 16·7 % for 12–23 months (RR = 0·83; 95 % CI 0·77, 0·91), 15·5 % for 24–35 months (RR = 0·84; 95 % CI 0·78, 0·92), and 34·9 % for 36–60 months of the enrolment age (RR = 0·65; 95 % CI 0·59, 0·72), compared to the enrolment age of 6–11 months.

Enrolment in the PD/Hearth programme at a younger age had the advantage of greater rehabilitation from underweight than older age. Our findings provide a better understanding of the successes and failures of the PD/Hearth programme to achieve more sustainable and cost-effective impacts.

To determine the impact of an educational programme for primary schools that explored the biodiversity of tomato, by promoting science and sensory education with three distinct varieties of it, in the acceptance of vegetables.

A randomised controlled study in which children were exposed to the educational programme (intervention group) or remained in the class, as usual (control group). The educational programme consisted of three sessions where children explained the observed differences between the three varieties of tomato and individual perceptions of their flavours based on sensory-based food education and by planning and implementing experiments to explain those differences. We tested the effects on both children’s willingness to try and their liking for tomato, and for lettuce and cabbage to study the carry-over effect, compared with the control group (Mann–Whitney U test; P < 0·05).

The study took place in public primary schools in Porto, Portugal.

Children in the third grade (8–13-year-old children) (n 136) were randomly assigned to intervention or control group.

Children in the intervention group reported significant increases in their willingness to try and liking for tomato compared to the control group (P < 0·05), but not for lettuce and cabbage (P > 0·05).

These results highlight the potential for fostering children’s acceptance of a vegetable by exploring biodiversity through science education. Further work may clarify the effects of exploring biodiversity on the consumption of vegetables and establish whether the results are stable over time and replicable across contexts and populations.

To explore whether socio-economic differences exist in cardiometabolic risk markers in children and whether lifestyle-related factors potentially mediate these differences.

Cross-sectional study including measurements of fasting blood lipids, glucose, homeostasis model assessment of insulin resistance (HOMA-IR), blood pressure and heart rate. Potential mediators examined were fat mass index (FMI); intakes of fruit, vegetables, dietary fibre and added sugar; whole-blood n-3 long-chain PUFA (LCPUFA) as a biomarker of fish intake; and physical activity and sedentary time.

Nine primary schools in Denmark.

Children aged 8–11 years (n 715).

Children of parents with the shortest compared with longest education had higher TAG by 0·12 (95 % CI 0·04, 0·21) mmol/l and HOMA-IR by 0·36 (0·10, 0·62), whereas children of parents with a vocational education had higher total cholesterol by 0·14 (0·02, 0·27) mmol/l and LDL cholesterol by 0·14 (0·03, 0·25) mmol/l compared with children of parents with the longest education; all P<0·05. FMI explained 25 % of the difference in TAG, 64 % of the difference in HOMA-IR and 21–29 % of the differences in cholesterols. FMI and whole-blood n-3 LCPUFA combined explained 42 % of the difference in TAG, whereas FMI, whole-blood n-3 LCPUFA and dietary fibre explained 89 % of the difference in HOMA-IR.

Socio-economic differences were present in blood lipids and insulin resistance among 8- to 11-year-olds and were mediated by body fatness, whole-blood n-3 LCPUFA and dietary fibre. These lifestyle factors may be targets in public initiatives to reduce socio-economic differences. Confirmation in longitudinal studies and trials is warranted.

To determine iodine levels in salt and iodine deficiency prevalence in school-aged children in 16 districts in Tanzania with previous severe iodine deficiency.

A cross-sectional study in schoolchildren. Systematic probability sampling was used to select schools and subjects for goitre assessment and urinary iodine determination.

Sixteen districts randomly selected from the 27 categorised as severely iodine-deficient in Tanzania.

The study population was primary-school children aged 6–18 years who were examined for goitre prevalence and urinary iodine concentration (UIC). Salt samples from schoolchildren's homes and from shops were tested for iodine content.

The study revealed that 83.3% of households (n = 21 160) in the surveyed districts used iodised salt. Also, 94% of sampled shops (n = 397) sold iodised salt, with a median iodine level of 37.0 ppm (range 4.2–240 ppm). Median UIC in 2089 schoolchildren was 235.0 μg l− 1 and 9.3% had UIC values below 50 μg l− 1. The overall unweighted mean visible and total goitre prevalence was 6.7% and 24.3%, respectively (n = 16 222). The age group 6–12 years had the lowest goitre prevalence (3.6% visible and 18.0% total goitre, n = 7147). The total goitre prevalence had decreased significantly in all districts from an unweighted mean of 65.4% in the 1980s to 24.3% in 1999 (P < 0.05). We believe this difference was also biologically significant.

These findings indicate that iodine deficiency is largely eliminated in the 16 districts categorised as severely iodine-deficient in Tanzania, and that the iodine content of salt purchased from shops is highly variable.

To investigate the association between the intake of n-6 PUFA and subsequent change in body weight and waist circumference at different levels of the carbohydrate:protein ratio.

Follow-up study with anthropometric measurements at recruitment and on average 5·3 years later. Dietary intake was determined at recruitment by using an FFQ that was designed for the study and validated. We applied linear regression models with 5-year change in weight or waist circumference as outcome and including a two-way interaction term between n-6 PUFA and carbohydrate intakes, lower-order terms, protein intake, long-chain n-3 PUFA intake and other potential confounders. Due to adjustment for intake of protein, levels of carbohydrate indirectly reflect levels of the carbohydrate:protein ratio.

Diet, Cancer and Health follow-up study, Denmark.

Women and men (n 29 152) aged 55 years.

For a high intake of n-6 PUFA (6·9 % of energy) v. a low intake of n-6 PUFA (3·4 % of energy), the difference in 5-year weight change was −189·7 g (95 % CI −636·8, 257·4 g) at a low carbohydrate:protein ratio and −86·7 g (95 % CI −502·9, 329·6 g) at a high carbohydrate:protein ratio; the differences in 5-year waist circumference change were 0·26 cm (95 % CI −0·47, 0·98 cm) and −0·52 cm (95 % CI −1·19, 0·15 cm), respectively. Inclusion of the dietary glycaemic index did not change the results.

No consistent associations between the intake of n-6 PUFA and change in body weight or waist circumference at different levels of the carbohydrate:protein ratio were observed.