Dietary intake assessment is often complicated by intrinsic bias. This study investigated whether food purchase data could constitute a valid indication of dietary intake, by evaluating the extent to which diet quality as measured by supermarket food purchases is correlated with diet quality as measured by reported dietary intake. We used data from the Supreme Nudge cluster-randomised controlled supermarket trial (n 227). Data were collected at baseline from supermarket purchases (loyalty cards) and a dietary questionnaire (short forty-item FFQ) to compute two scores reflecting diet quality from purchasing data (purchased diet quality) and FFQ (consumed diet quality). Both scores constituted thirteen food groups and could theoretically range from 0 (low diet quality) to 130 (high diet quality). The relationship between purchased diet quality and consumed diet quality was assessed using correlation coefficients and the Bland–Altman limits-of-agreement method. Multiple linear regression was fitted between purchased diet quality and consumed diet quality, adjusted for age, sex, waist circumference, educational level and household size. Consumed and purchased diet qualities were modestly positively correlated (Pearson’s ρ = 0·31, 95 % CI 0·18, 0·42). A positive association from linear regression was found after confounding adjustments (βbaseline = 0·22, 95 % CI 0·10, 0·34). The purchased diet quality was systematically lower than the consumed diet quality. This study found that diet quality as measured by supermarket purchases provided a reasonable indication of diet quality as reported by short-FFQ, albeit modest.

The association between intake of different dairy products and the risk of stroke remains unclear. We therefore investigated substitutions between dairy product subgroups and risk of stroke. We included 36 886 Dutch men and women. Information about dairy product intake was collected through a FFQ. Dairy products were grouped as low-fat milk, whole-fat milk, buttermilk, low-fat yogurt, whole-fat yogurt, cheese and butter. Incident stroke cases were identified in national registers. We used Cox proportional hazards regression to calculate associations for substitutions between dairy products with the rate of stroke. During a median follow-up of 15·2 years we identified 884 stroke cases (503 ischaemic and 244 haemorrhagic). Median intake of total dairy products was four servings/d. Low-fat yogurt substituted for whole-fat yogurt was associated with a higher rate of ischaemic stroke (hazard ratio (HR) = 2·58 (95 % CI 1·11, 5·97)/serving per d). Whole-fat yogurt as a substitution for any other subgroup was associated with a lower rate of ischaemic stroke (HR between 0·33 and 0·36/serving per d). We did not observe any associations for haemorrhagic stroke. In conclusion, whole-fat yogurt as a substitution for low-fat yogurt, cheese, butter, buttermilk or milk, regardless of fat content, was associated with a lower rate of ischaemic stroke.

Dietary guidelines for pure fruit juice consumption differ between countries, regarding the question whether pure fruit juice is an acceptable alternative for fruit. Currently, little is known about pure fruit juice consumption and the risk of CVD. In this prospective cohort study, we studied the association of pure fruit juice and fruit consumption with the incidence of fatal and non-fatal CVD, CHD and stroke and investigated the differences in association with pure fruit juice consumption between low and high fruit consumers. A validated FFQ was used to estimate dietary intake of 34 560 participants (26·0 % men and 74·0 % women) aged 20–69 years from the European Prospective Investigation into Cancer and Nutrition–Netherlands study. Adjusted hazard ratios (HR) were estimated using Cox regression after average follow-up of 14·6 years. Compared with no consumption, pure fruit juice consumption up to 7 glasses/week – but not consumption of ≥8 glasses – was significantly associated with reduced risk of CVD and CHD, with HR from 0·83 (95 % CI 0·73, 0·95) to 0·88 (95 % CI 0·80, 0·97). Consumption of 1–4 and 4–8 glasses/week was significantly associated with lower risk of stroke with HR of 0·80 (95 % CI 0·64, 0·99) and 0·76 (95 % CI 0·61, 0·94), respectively. Associations did not differ considerably between low and high fruit consumers. The highest three quintiles of fruit consumption (≥121 g/d) were significantly associated with lower incidence of CVD, with HR of 0·87 (95 % CI 0·78, 0·97) and 0·88 (95 % CI 0·80, 0·98). In conclusion, although we observed favourable associations of moderate pure fruit juice consumption with CVD, for now consumption of whole fruit should be preferred because the evidence of the health benefits of fruit is more conclusive.

Guidelines for a healthy diet aim to decrease the risk of chronic diseases. It is unclear as to what extent a healthy diet is also an environmentally friendly diet. In the Dutch sub-cohort of the European Prospective Investigation into Cancer and Nutrition, the diet was assessed with a 178-item FFQ of 40 011 participants aged 20–70 years between 1993 and 1997. The WHO’s Healthy Diet Indicator (HDI), the Dietary Approaches to Stop Hypertension (DASH) score and the Dutch Healthy Diet index 2015 (DHD15-index) were investigated in relation to greenhouse gas (GHG) emissions, land use and all-cause mortality risk. GHG emissions were associated with HDI scores (−3·7 % per sd increase (95 % CI −3·4, −4·0) for men and −1·9 % (95 % CI −0·4, −3·4) for women), with DASH scores in women only (1·1 % per sd increase, 95 % CI 0·9, 1·3) and with DHD15-index scores (−2·5 % per sd increase (95 % CI −2·2, −2·8) for men and −2·0 % (95 % CI −1·9, −2·2) for women). For all indices, higher scores were associated with less land use (ranging from −1·3 to −3·1 %). Mortality risk decreased with increasing scores for all indices. Per sd increase of the indices, hazard ratios for mortality ranged from 0·88 (95 % CI 0·82, 0·95) to 0·96 (95 % CI 0·92, 0·99). Our results showed that adhering to the WHO and Dutch dietary guidelines will lower the risk of all-cause mortality and moderately lower the environmental impact. The DASH diet was associated with lower mortality and land use, but because of high dairy product consumption in the Netherlands it was also associated with higher GHG emissions.

We investigated the validity and reproducibility of the FFQ used in the Dutch European Investigation of Cancer and Nutrition cohort, in order to rank subjects according to intakes of fatty acid classes and individual fatty acids. In total, 121 men and women (23–72 years) filled out three FFQ at 6-month intervals between 1991 and 1992. As a reference method, they filled out twelve monthly 24-h dietary recalls (24HDR) during the same year. Intra-class correlation coefficients for the FFQ showed moderate to good reproducibility across all fatty acids (classes and individual) in men (0·56–0·81) and women (0·57–0·83). In men, Spearman’s correlation coefficients (rs) for the FFQ compared with the 24HDR indicated moderate to good relative validity (rs=0·45–0·71) for all fatty acids, except for arachidonic acid and marine PUFA (rs<0·40). In women, relative validity was moderate to good for MUFA and trans-fatty acids (TFA) and the majority of SFA (rs=0·40–0·66), was fair for the short-chain SFA and lauric acid (rs=0·30–0·33) and was fair to moderate for PUFA (rs=0·22–0·47). Bland–Altman plots showed good agreement between the FFQ and 24HDR, and proportional bias for fatty acids with very low intakes. In conclusion, the FFQ showed good reproducibility for subject ranking based on intakes of fatty acids (classes and individual). The relative validity measures indicated that the FFQ is an adequate tool to rank subjects according to intakes of high-abundant fatty acids, but less for low-abundant fatty acids.

The objective of the present study was to investigate the relationship between total and subtypes of bacterial fermented food intake (dairy products, cheese, vegetables and meat) and mortality due to all causes, total cancer and CVD. From the European Prospective Investigation into Cancer and Nutrition-Netherlands cohort, 34 409 Dutch men and women, aged 20–70 years who were free from CVD or cancer at baseline, were included. Baseline intakes of total and subtypes of fermented foods were measured with a validated FFQ. Data on the incidence and causes of death were obtained from the national mortality register. Cox proportional hazards models were used to analyse mortality in relation to the quartiles of fermented food intake. After a mean follow-up of 15 (sd 2·5) years, 2436 deaths occurred (1216 from cancer and 727 from CVD). After adjustment for age, sex, total energy intake, physical activity, education level, hypertension, smoking habit, BMI, and intakes of fruit, vegetables and alcohol, total fermented food intake was not found to be associated with mortality due to all causes (hazard ratio upper v. lowest quartile (HRQ4 v. Q1) 1·00, 95 % CI 0·88, 1·13), cancer (HRQ4 v. Q1 1·02, 95 % CI 0·86, 1·21) or CVD (HRQ4 v. Q1 1·04, 95 % CI 0·83, 1·30). Bacterial fermented foods mainly consisted of fermented dairy foods (78 %) and cheese (16 %). None of the subtypes of fermented foods was consistently related to mortality, except for cheese which was moderately inversely associated with CVD mortality, and particularly stroke mortality (HRQ4 v. Q1 0·59, 95 % CI 0·38, 0·92, Ptrend= 0·046). In conclusion, the present study provides no strong evidence that intake of fermented foods, particularly fermented dairy foods, is associated with mortality.