To identify ingredients from products and by-products derived from GM crops in packaged food products and to analyse the presence of these ingredients in the foods most commonly consumed by the Brazilian population.

Cross-sectional study.

A search of the scientific literature to identify the use of products and by-products derived from GM crops in foods in Brazil and a study of food labels in a supermarket belonging to one of the ten largest supermarket chains in Brazil.

To identify the ingredients present in packaged food products and their nomenclatures, the labels of all packaged food products available for sale in a supermarket were analysed. Subsequently, the presence of potential GM ingredients in the foods most commonly consumed by the Brazilian population was analysed.

A total of twenty-eight GM crops’ by-products with applications in the food industry (from soyabeans, corn, cotton and a yeast) were identified. Such by-products are presented as food ingredients or additives on food labels with 101 distinct nomenclatures. Most of the variety (63·8 %) and the quantity (64·5 %) of the foods most commonly consumed by Brazilians may contain a least one GM ingredient.

The presence of at least one potential GM ingredient was observed in more than half of the variety of foods most commonly consumed by the Brazilian population. Such ingredients were identified with distinct nomenclatures and incomplete descriptions, which may make it difficult to identify potential GM foods and confuse consumers when making food choices.

To test an obesity prevention strategy derived from behavioural economics (optimal defaults plus delay), focused on changing the college dining hall service method.

After a uniform pre-load, participants attended an experimental lunch in groups randomized to one of three conditions: a nutrient-dense, lower-fat/energy lunch as an optimal default (OD); a less-nutrient-dense, higher-fat/energy lunch as a suboptimal default (SD); or a free array (FA) lunch. In the OD condition, students were presented a menu depicting healthier vegetarian and omnivore foods as default, with opt-out alternatives (SD menu) available on request with a 15 min wait. In the SD condition, the same menu format was used with the positioning of food items switched. In the FA condition, all choices were presented in uniform fonts and were available immediately.

Private rooms designed to provide a small version of a college dining hall, on two campuses of a Northeastern US university.

First-year college students (n 129).

There was a significant main effect for condition on percentage of optimal choices selected, with 94 % of food choices in the OD condition optimal, 47 % in the FA condition optimal and none in the SD condition optimal. Similarly, energy intake for those in the SD condition significantly exceeded that in the FA condition, which exceeded that in the OD condition.

Presenting menu items as optimal defaults with a delay had a significant impact on choice and consumption, suggesting that further research into its long-term applicability is warranted.

To investigate relationships between weight resilience (maintaining a normal weight in a food desert environment) and fruit and vegetable (F&V) intake, attitudes and barriers.

Cross-sectional, in-person surveys collected May–December 2011, including self-reported data on F&V-related psychosocial factors, attitudes and barriers. Two 24 h dietary recalls were completed; weight and height were measured. Multivariable regression models estimated prevalence ratios (95 % CI).

Two low-income, predominantly African-American food deserts in Pittsburgh, Pennsylvania, USA.

Women aged 18–49 years (n 279) who were the primary food shopper in a household randomly selected for a parent study.

Fifteen per cent were weight resilient, 30 % were overweight and 55 % were obese. Overall, 25 % reported eating ≥5 F&V servings/d. After adjustment for age, education, parity, employment, living alone, physical activity, per capita income and mean daily energy intake, women eating ≥5 F&V servings/d were 94 % more likely to be weight resilient compared with those eating <5 servings/d (1·94; 1·10, 3·43). Across BMI groups, self-efficacy regarding F&V consumption was high and few F&V barriers were reported. The most frequently reported barrier was concern about the cost of F&V (36 %). Of the attitudinal F&V-related factors, only concern about wasting food when serving F&V was associated with weight resilience in adjusted models (0·29; 0·09, 0·94). In a model predicting consuming ≥5 F&V servings/d, driving one’s own car to the store was the only attitudinal F&V-related factor associated with consumption (1·50; 1·00, 2·24).

In this population, weight resilience may be encouraged by improving access to affordable and convenient F&V options and providing education on ways to make them palatable to the entire household, rather than by shifting women’s F&V perceptions, which are already positive.

To test whether diets achieving recommendations from the UK’s Scientific Advisory Committee on Nutrition (SACN) were associated with higher monetary costs in a nationally representative sample of UK adults.

A cross-sectional study linking 4 d diet diaries in the National Diet and Nutrition Survey (NDNS) to contemporaneous food price data from a market research firm. The monetary cost of diets was assessed in relation to whether or not they met eight food- and nutrient-based recommendations from SACN. Regression models adjusted for potential confounding factors. The primary outcome measure was individual dietary cost per day and per 2000 kcal (8368 kJ).

UK.

Adults (n 2045) sampled between 2008 and 2012 in the NDNS.

On an isoenergetic basis, diets that met the recommendations for fruit and vegetables, oily fish, non-milk extrinsic sugars, fat, saturated fat and salt were estimated to be between 3 and 17 % more expensive. Diets meeting the recommendation for red and processed meats were 4 % less expensive, while meeting the recommendation for fibre was cost-neutral. Meeting multiple targets was also associated with higher costs; on average, diets meeting six or more SACN recommendations were estimated to be 29 % more costly than isoenergetic diets that met no recommendations.

Food costs may be a population-level barrier limiting the adoption of dietary recommendations in the UK. Future research should focus on identifying systems- and individual-level strategies to enable consumers achieve dietary recommendations without increasing food costs. Such strategies may improve the uptake of healthy eating in the population.

To (i) determine differences in the availability of variety of fruits and vegetables (F&V) at home by food security status; and (ii) examine the inter-associations between food security status, availability of variety of F&V at home and frequency of F&V intake, among low-income pregnant women.

Participants were interviewed to collect food security status, home availability of variety of F&V and frequency of F&V intake. Bivariate analyses, multivariate regression and exploratory mediation analyses were conducted using IBM SPSS Statistics version 23.0 and the PROCESS macro.

Low-income pregnant women (n 198) were interviewed if they were ≥18 years of age, in the second trimester of pregnancy, and spoke English or Spanish.

Low/very low food security was found among 43 % of participants. Compared with fully food-secure participants, very low food-secure participants reported a lower variety of fruits (P=0·028) and vegetables (P=0·058) available at home. Mediation analyses indicated that through home availability of variety of fresh F&V, food security status was associated with the daily intake of F&V (indirect effect (95 % CI): fresh fruits, −0·039 (−0·074, −0·013); fresh vegetables, −0·048 (−0·083, −0·023)). As food security worsened, the available variety of fresh F&V decreased, which was associated with lower intake.

The study highlights the interlink between food security, home food environment and diet quality, and the importance of nutrition education intervention to promote a healthy home food environment and improved pregnancy outcomes among low-income women.

Wheat and rye, the most consumed whole grains (WG) in the Nordic countries, contain alkylresorcinols (AR) in their bran. AR concentrations in human adipose tissue might reflect long-term WG rye and wheat intake. We aimed to evaluate AR concentrations in adipose tissue biopsies as a long-term biomarker of WG wheat and rye intake in free-living Swedish men and women.

Cross-sectional study. AR concentrations in adipose tissue biopsies were analysed and compared with long-term WG intake assessed by three FFQ (repeated over a period of 14 years in men, 17 years in women) and with plasma AR concentrations.

The Cohort of Swedish Men between 1997 and 2010 and the Swedish Mammography Cohort between 1987 and 2003, Sweden.

Men (n 149) and women (n 109).

Long-term WG rye intake estimated with repeated FFQ correlated (r=0·31–0·41, P<0·01) with adipose-tissue AR concentrations, while WG wheat intake correlated only weakly (r=0·17–0·33, P<0·05). Total AR concentration in adipose tissue was 61 % lower in women than in men at similar energy-adjusted WG wheat and rye intakes, but plasma concentrations were similar. AR concentrations in adipose tissue correlated well with plasma concentrations (r=0·49–0·81, P<0·001).

AR in adipose tissue reflected long-term WG rye but not WG wheat intake, probably due to poor precision in estimating WG wheat intake by FFQ. AR in adipose tissue appears promising as a biomarker of long-term WG rye intake but should be adjusted for sex.

To identify optimal food choices that meet nutritional recommendations to reduce prevalence of inadequate nutrient intakes.

Linear programming was used to obtain an optimized diet with sixty-eight foods with the least difference from the observed population mean dietary intake while meeting a set of nutritional goals that included reduction in the prevalence of inadequate nutrient intakes to ≤20 %.

Brazil.

Participants (men and women, n 25 324) aged 20 years or more from the first National Dietary Survey (NDS) 2008–2009.

Feasible solution to the model was not found when all constraints were imposed; infeasible nutrients were Ca, vitamins D and E, Mg, Zn, fibre, linolenic acid, monounsaturated fat and Na. Feasible solution was obtained after relaxing the nutritional constraints for these limiting nutrients by including a deviation variable in the model. Estimated prevalence of nutrient inadequacy was reduced by 60–70 % for most nutrients, and mean saturated and trans-fat decreased in the optimized diet meeting the model constraints. Optimized diet was characterized by increases especially in fruits (+92 g), beans (+64 g), vegetables (+43 g), milk (+12 g), fish and seafood (+15 g) and whole cereals (+14 g), and reductions of sugar-sweetened beverages (−90 g), rice (−63 g), snacks (−14 g), red meat (−13 g) and processed meat (−9·7 g).

Linear programming is a unique tool to identify which changes in the current diet can increase nutrient intake and place the population at lower risk of nutrient inadequacy. Reaching nutritional adequacy for all nutrients would require major dietary changes in the Brazilian diet.

To present and discuss results of the most recent studies pertaining to the effects of consumption of different types of fatty acids on the risk of CVD. The aim was also an attempt to answer the question of whether a revision of the current recommendations is necessary.

A review of prospective cohort studies, systematic reviews and meta-analyses published in 2014–2017 on the effects of SFA and trans-fatty acid (TFA) intakes as well as various models of their replacement in the diet on CVD risk.

Results of the new large prospective cohort studies pertaining to the effect of SFA consumption on CVD risk are contradictory. Similarly, the recent meta-analyses of clinical trials related to the effects of SFA substitution on CVD risk provided extremely different results, which is related to the application of different inclusion and exclusion criteria. Differences in results of randomised controlled trials may be caused by different methodologies of dietary parameter changes, varying duration of studies, as well as the time at which they were carried out.

It is extremely difficult to extrapolate results of recent studies to contemporary recommendations. It seems that there is a need for properly randomised studies on large groups, with good control of dietary and non-dietary parameters, which account for not only the sum of SFA and TFA, but also their source. Only such studies will allow for full evaluation of an effect of substituting SFA and TFA on cardiovascular risk.

In the present study we investigated gender-specific associations of low birth weight (LBW) and shorter relative leg length with metabolic syndrome (MetS) after adjusting for sociodemographic characteristics and health-related behaviours. We also investigated whether these associations are independent of age at menarche and BMI at 20 years old.

Cross-sectional analysis.

Baseline data from 12 602 participants (35–74 years) of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), 2008–2010.

MetS was defined according to the revised National Cholesterol Education Program Adult Treatment Panel III guidelines. LBW (<2·5 kg) and age- and sex-standardized relative leg length (high, medium and low) were the explanatory variables studied. The strength of the associations between the explanatory variables and MetS was estimated by Poisson regression with robust variance.

MetS prevalence was 34·2 %; it was more prevalent in men (36·8 %) than in women (32·2 %). In multivariate analysis, LBW was associated (prevalence ratio; 95 % CI) with MetS only in women (1·28; 1·24, 1·45). Shorter leg length was associated with MetS in both men (1·21; 1·09, 1·35 and 1·46; 1·29, 1·65 for low and medium lengths, respectively) and women (1·12; 1·00, 1·25 and 1·40; 1·22, 1·59 for low and medium lengths, respectively). Additional adjustments for age at menarche and BMI at 20 years old did not change the associations.

Poor nutritional status as estimated by LBW and lower leg length in childhood was associated with a higher prevalence of MetS, although LBW was a significant factor only among women.

To evaluate effectiveness of point-of-use water treatment in improving treatment of children affected by severe acute malnutrition (SAM).

Programme sites were randomized to one of four intervention arms: (i) standard SAM treatment; (ii) SAM treatment plus flocculent/disinfectant water treatment; (iii) SAM treatment plus chlorine disinfectant; or (iv) SAM treatment plus ceramic water filter. Outcome measures were calculated based on participant status upon exit or after 120d of enrolment, whichever came first. Child anthropometric data were collected during weekly monitoring at programme sites. Child caregivers were interviewed at enrolment and exit. Use of water treatment products was assessed in a home visit 4–6 weeks after enrolment.

Dadu District, Sindh Province, Pakistan.

Children (n 901) aged 6–59 months with SAM and no medical complications.

Recovery rates were 16·7–22·2 % higher among children receiving water treatment compared with the control group. The adjusted odds of recovery were approximately twice as high for those receiving water treatment compared with controls. Mean length of stay until recovery was 73 (sd 24·6) d and mean rate of weight gain was 4·7 (sd 3·0) g/kg per d. Differences in recovery rate, length of stay and rate of weight gain between intervention groups were not statistically significant.

Incorporating point-of-use water treatment into outpatient treatment programmes for children with SAM increased nutritional recovery rates. No significant differences in recovery rates were observed between the different intervention groups, indicating that different water treatment approaches were equally effective in improving recovery.

To review the available literature on the association between consumption of ultra-processed foods and body fat during childhood and adolescence.

A systematic review was conducted in the PubMed, Web of Science and LILACS databases. Studies that evaluated the association between consumption of ultra-processed food (exposure) and body fat (outcome) during childhood and adolescence were eligible.

Healthy children and adolescents.

Twenty-six studies that evaluated groups of ultra-processed foods (such as snacks, fast foods, junk foods and convenience foods) or specific ultra-processed foods (soft drinks/sweetened beverages, sweets, chocolate and ready-to-eat cereals) were selected. Most of the studies (n 15) had a cohort design. Consumption was generally evaluated by means of FFQ or food records; and body composition, by means of double indirect methods (bioelectrical impedance analysis and skinfolds). Most of the studies that evaluated consumption of groups of ultra-processed foods and soft drinks/sweetened beverages found positive associations with body fat.

Our review showed that most studies have found positive associations between consumption of ultra-processed food and body fat during childhood and adolescence. There is a need to use a standardized classification that considers the level of food processing to promote comparability between studies.

Evidence suggests that improvements to the childcare nutrition environment can have a positive impact on child dietary intake. The primary aim of the present study was to assess, relative to usual care, the effectiveness of a multi-strategy implementation intervention in improving childcare compliance with nutrition guidelines. As a secondary aim, the impact on child dietary intake was assessed.

Parallel-group, randomised controlled trial design. The 6-month intervention was designed to overcome barriers to implementation of the nutrition guidelines that had been identified by applying the theoretical domains framework.

Hunter New England region, New South Wales, Australia.

Forty-five centre-based childcare services.

There were no differences between groups in the proportion of services providing food servings (per child) compliant with nutrition guideline recommendations for all five (5/5) food groups at follow-up (i.e. full compliance). Relative to control services, intervention services were more likely to be compliant with guidelines (OR; 95 % CI) in provision of fruit (10·84; 1·19, 551·20; P=0·0024), meat and meat alternatives (8·83; 1·55, –; P=0·023), dairy (8·41; 1·60, 63·62; P=0·006) and discretionary foods (17·83; 2·15, 853·73; P=0·002). Children in intervention services consumed greater servings (adjusted difference; 95 % CI) of fruit (0·41; 0·09, 0·73; P=0·014) and vegetables (0·70; 0·33, 1·08; P<0·001).

Findings indicate that service-level changes to menus in line with dietary guidelines can result in improvements to children’s dietary intake. The study provides evidence to advance implementation research in the setting as a means of enhancing child public health nutrition.

Nutritional warnings have recently been suggested as a simplified front-of-pack nutrition labelling scheme to facilitate citizens’ ability to identify unhealthful products and discourage their consumption. However, citizens’ perspective on this policy is still under-researched. The objective of the present work was to study how citizens perceive nutritional warnings and to evaluate public support of this policy, with the goal of deriving recommendations for the design of policy measures accompanying the introduction of nutritional warnings.

An online survey with 1416 Uruguayan citizens, aged 18–75 years, 61 % female, was conducted. Participants had to answer a series of questions (open-ended and multiple-choice) related to their perception of warnings as a front-of-package nutrition labelling scheme.

Participants showed a positive attitude towards nutritional warnings, which were regarded as easy to understand and to identify on food packages. The majority of respondents emphasized that they would take nutritional warnings into account when making their food choices, stating that they would allow them to make informed choices and, consequently, to increase the quality of their diet and their health status. Health motivation appeared as a crucial driver for taking nutritional warnings into consideration.

A high level of public support for nutritional warnings was observed. Responses can be used to derive a range of recommendations for a policy mix that should synergistically support the introduction of nutritional warnings and encourage citizens to take them into account when making their food purchases.

We aimed to examine the food-seeking experiences of homeless emerging adults (age 18–24 years) in a US urban context.

The study used a qualitative descriptive design, combining semi-structured interviews with a standardized quantitative measure of food insecurity. Interview data were coded using constant comparative methods to identify patterns across and within interviews. Emerging themes were confirmed and refined through member checking.

Buffalo, a mid-sized city in the Northeastern USA.

A sample of thirty participants was recruited through community-based methods. Eligibility criteria specified that participants were aged 18–24 years and did not have a stable place to live. The sample was demographically diverse and included participants who were couch-surfing, staying on the streets and/or using shelters.

Participants’ food access strategies varied across their living circumstances. Common strategies included purchasing food with cash or benefits (reported by 77 %), using free meal programmes (70 %) and eating at friends’ or relatives’ homes (47 %). Although 70 % of participants received Supplemental Nutrition Assistance Program benefits, several reported access barriers, including initial denials of eligibility due to being listed on a parent’s application even when the participant no longer resided in the household. Participants described a stigma associated with using food pantries and free meal programmes and expressed preference for less institutionalized programmes such as Food Not Bombs.

Given endemic levels of food insecurity among homeless youth and young adults, policy modifications and service interventions are needed to improve food access for this population.

While economic crises can increase socio-economic disparities in health, little is known about the impact of the 2008–09 Great Recession on obesity prevalence among children, especially low-income children. The present study examined whether socio-economic disparities in obesity among children of pre-school age participating in a federal nutrition assistance programme have changed since the recession.

A pre–post observational study using administrative data of pre-school-aged programme participants from 2003 to 2014. Logistic regression was used to examine whether the relationship between obesity prevalence (BMI≥95th percentile of the Centers for Disease Control and Prevention’s growth charts) and three measures of socio-economic status (household income, household educational attainment, neighbourhood-level median household income) changed after the recession by examining the interaction between each socio-economic status measure and a 5-year time-period variable (2003–07 v. 2010–14), stratified by child’s age and adjusted for child’s sociodemographic characteristics.

Los Angeles County, California, USA.

Children aged 2–4 years (n 1 637 788) participating in the Special Supplemental Nutrition Program for Women, Infants, and Children.

The magnitude of the association of household income and household education with obesity increased after 2008–09 among 3- and 4-year-olds and 2- and 3-year-olds, respectively. However, the magnitude of the association of neighbourhood-level median household income with obesity did not change after 2008–09.

Disparities in obesity by household-level socio-economic status widened after the recession, while disparities by neighbourhood-level socio-economic status remained the same. The widening household-level socio-economic disparities suggest that obesity prevention efforts should target the most vulnerable low-income children.

Even though current policy is strongly focused on the crucial first ‘1000 days’, it might be still possible to enhance cognitive function during the pre-adolescent and adolescent years by improving micronutrient status. In Cambodia, nutritional status is poor. Provision of rice fortified with micronutrients through a school meal programme (SMP) could be a cost-effective strategy to help improve health and school performance. The present study aimed to evaluate the effect of three different micronutrient-fortified rice formulations on cognitive function in Cambodian children.

Sixteen Cambodian schools receiving SMP.

The FORISCA-UltraRice®+NutriRice® study was a randomized, double-blind, placebo-controlled trial. Four groups of four schools were randomly allocated to receive normal rice, UltraRice®Original, UltraRice®New or NutriRice®. Within each school, 132 children were randomly selected. Data on cognitive performance (picture completion, block design and Raven’s coloured progressive matrices (RCPM)), anthropometry, parasite infestation and micronutrient status were collected before the intervention and after 6 months.

Cognitive data were available for 1796 children aged 6–16 years.

All cognitive scores improved after 6 months (P<0·001). Block design score improvement was significantly higher in children consuming UltraRice®Original (P=0·03) compared with the other fortified rice groups and placebo. No difference among groups was found on RCPM or picture completion scores. Stunting, parasite infestation and inflammation negatively affected the impact of the intervention.

Combined with other interventions, using SMP to distribute fortified rice to schoolchildren may be a cost-effective way to increase cognitive performance and thereby improve school performance and educational achievements.

The present qualitative study aimed to gain an in-depth understanding of participants’ attitudes, knowledge, perceived effectiveness (a person’s belief that his/her behaviour can contribute to environmental preservation) and behaviours relating to a sustainable eating pattern.

One-to-one interviews (either face-to-face or by telephone) were conducted following a structured interview schedule, audio-recorded, transcribed verbatim and analysed using inductive thematic analysis in NVivo 10.

Victorian (Australia) adult participants recruited via online advertisements, flyers on community advertisement boards and letterbox drops.

Twenty-four participants (mean age 40 years, range 19–69 years; thirteen female, eleven male) were interviewed.

Participants reported that environmental impact was not an important influence on their food choice. Participants displayed limited knowledge about a sustainable eating pattern, with most unaware of the environmental impact of food-related behaviours. Most participants believed sustainable eating would be only slightly beneficial to the environment. Participants reported undertaking limited sustainable food behaviours currently and were more willing to undertake a food behaviour if they perceived additional benefits, such as promoting health or supporting the local community.

The study suggests consumers need further information about a sustainable eating pattern and the environmental impact of food choice. The findings highlight some of the barriers that will need to be addressed when promoting this kind of eating pattern and that a range of interventions will be necessary.

To examine the nutritional quality of menu items promoted in four (US) fast-food restaurant chains (McDonald’s, Burger King, Wendy’s, Taco Bell) in 2010 and 2013.

Menu items pictured on signs and menu boards were recorded at 400 fast-food restaurants across the USA. The Nutrient Profile Index (NPI) was used to calculate overall nutrition scores for items (higher scores indicate greater nutritional quality) and was dichotomized to denote healthier v. less healthy items. Changes over time in NPI scores and energy of promoted foods and beverages were analysed using linear regression.

Four hundred fast-food restaurants (McDonald’s, Burger King, Wendy’s, Taco Bell; 100 locations per chain).

NPI of fast-food items marketed at fast-food restaurants.

Promoted foods and beverages on general menu boards and signs remained below the ‘healthier’ cut-off at both time points. On general menu boards, pictured items became modestly healthier from 2010 to 2013, increasing (mean (se)) by 3·08 (0·16) NPI score points (P<0·001) and decreasing (mean (se)) by 130 (15) kJ (31·1 (3·65) kcal; P<0·001). This pattern was evident in all chains except Taco Bell, where pictured items increased in energy. Foods and beverages pictured on the kids’ section showed the greatest nutritional improvements. Although promoted foods on general menu boards and signs improved in nutritional quality, beverages remained the same or became worse.

Foods, and to a lesser extent, beverages, promoted on menu boards and signs in fast-food restaurants showed limited improvements in nutritional quality in 2013 v. 2010.

To assess (i) the extent to which stakeholders have accepted and implemented a Healthy Food and Drink Policy for schools a decade after its introduction and (ii) any resulting implications for canteen profitability.

Online survey distributed via electronic newsletter to school principals.

Western Australian public schools.

Principals, teachers, canteen managers, and parents and citizens committee presidents (n 307).

Large majorities of respondents reported that the policy has made the foods and drinks provided in schools healthier (85 %) and that the policy constitutes a good opportunity to teach children about healthy eating (90 %). Only small proportions of respondents felt it had been difficult to implement the policy in their schools (13 %) or that the policy fails to accommodate parents’ rights to choose the foods consumed by their children (16 %). Most of the policy outcomes assessed in both the initial post-implementation evaluation (2008) and the 10-year follow-up evaluation (2016) demonstrated significant improvement over time.

The study results indicate that comprehensive school food policies can favourably influence the foods and drinks provided on school premises and can be highly acceptable to key stakeholders, without adversely affecting profitability. The results are encouraging for policy makers in other jurisdictions considering the implementation of similar policies.

To estimate the effect of early, regular breast-milk pumping on time to breast-milk feeding (BMF) and exclusive BMF cessation, for working and non-working women.

Using the Infant Feeding Practices Survey II (IFPS II), we estimated weighted hazard ratios (HR) for the effect of regular pumping (participant defined) compared with non-regular/not pumping, reported at month 2, on both time to BMF cessation (to 12 months) and time to exclusive BMF cessation (to 6 months), using inverse probability weights to control confounding.

USA, 2005–2007.

BMF (n 1624) and exclusively BMF (n 971) IFPS II participants at month 2.

The weighted HR for time to BMF cessation was 1·62 (95 % CI 1·47, 1·78) and for time to exclusive BMF cessation was 1·14 (95 % CI 1·03, 1·25). Among non-working women, the weighted HR for time to BMF cessation was 2·05 (95 % CI 1·84, 2·28) and for time to exclusive BMF cessation was 1·10 (95 % CI 0·98, 1·22). Among working women, the weighted HR for time to BMF cessation was 0·90 (95 % CI 0·75, 1·07) and for time to exclusive BMF cessation was 1·14 (95 % CI 0·96, 1·36).

Overall, regular pumpers were more likely to stop BMF and exclusive BMF than non-regular/non-pumpers. Non-working regular pumpers were more likely than non-regular/non-pumpers to stop BMF. There was no effect among working women. Early, regular pumpers may need specialized support to maintain BMF.