Fiber intake and all-cause mortality in the Prevención con Dieta Mediterránea (PREDIMED) study 1 3

See corresponding editorial on page 1409. Fiber intake and all-cause mortality in the Prevención con Dieta Mediterránea (PREDIMED) study 1 3 Pilar Buil-Cosiales, Itziar Zazpe, Estefanía Toledo, Dolores Corella, Jordi Salas-Salvadó, Javier Diez-Espino, Emilio Ros, Joaquin Fernandez-Creuet Navajas, José Manuel Santos-Lozano, Fernando Arós, Miquel Fiol, Olga Casta~ner, Lluis Serra-Majem, Xavier Pintó, Rosa M Lamuela-Raventós, Amelia Marti, F Javier Basterra-Gortari, José V Sorlí, Jose M a Verdú-Rotellar, Josep Basora, Valentina Ruiz-Gutierrez, Ramón Estruch, and Miguel A Martínez-González ABSTRACT Background: Few observational studies have examined the effect of dietary fiber intake and fruit and vegetable consumption on total mortality and have reported inconsistent results. All of the studies have been conducted in the general population and typically used only a single assessment of diet. Objective: We investigated the association of fiber intake and whole-grain, fruit, and vegetable consumption with all-cause mortality in a Mediterranean cohort of elderly adults at high cardiovascular disease (CVD) risk by using repeated measurements of dietary information and taking into account the effect of a dietary intervention. Design: We followed up 7216 men (55 75 y old) and women (60 75 y old) at high CVD risk in the Prevención con Dieta Mediterránea (PREDIMED) trial for a mean of 5.9 y. Data were analyzed as an observational cohort. Participants were initially free of CVD. A 137-item validated food-frequency questionnaire administered by dietitians was repeated annually to assess dietary exposures (fiber, fruit, vegetable, and whole-grain intakes). Deaths were identified through the continuing medical care of participants and the National Death Index. An independent, blinded Event Adjudication Committee adjudicated causes of death. Cox regression models were used to estimate HRs of death during follow-up according to baseline dietary exposures and their yearly updated changes. Results: In up to 8.7 y of follow-up, 425 participants died. Baseline fiber intake and fruit consumption were significantly associated with lower risk of death [HRs for the fifth compared with the first quintile: 0.63 (95% CI: 0.46, 0.86; P = 0.015) and 0.59 (95% CI: 0.42, 0.82; P = 0.004), respectively]. When the updated dietary information was considered, participants with fruit consumption.210 g/d had 41% lower risk of all-cause mortality (HR: 0.59; 95% CI: 0.44, 0.78). Associations were strongest for CVD mortality than other causes of death. Conclusion: Fiber and fruit intakes are associated with a reduction in total mortality. PREDIMED was registered at controlled-trials.com as ISRCTN35739639. Am J Clin Nutr 2014;100:1498 507. Keywords dietary fiber intake, fruit consumption, mortality, Mediterranean diet, olive oil, monounsaturated fat, tree nuts, feeding trial, cohort INTRODUCTION Prospective cohort studies have consistently shown that high consumption of dietary fiber (DF) 4, fruit, and vegetables lowers risk of stroke (1, 2) and heart disease (3, 4). In epidemiologic studies, dietary fiber has been associated with lower cardiovascular disease (CVD) mortality (5) and lower incidences of type 2 diabetes (6), some cancers (7) and inflammatory diseases (8 10). Fruit, vegetables, legumes, and grains are the main sources of DF, but some investigators have argued that the fiber content is 1 From the Prevención con Dieta Mediterránea Research Network (Red 06/ 0045) (PB-C, IZ, ET, JS-S, JD-E, FA, LS-M, XP, RML-R, AM, FJB-G, JB, VR-G, RE, and MÁM-G) and the Centro de Investigación Biomédica en Red Fisiopatología de la Obesidad y Nutrición (PB-C, IZ, ET, DC, JS-S, JD-E, ER, JF-CN, JMS-L, FA, MF, OC, LS-M, XP, RML-R, AM, FJB-G, JVS, JMV-R, JB, VR-G, RE, and MÁM-G), Instituto de Salud Carlos III, Madrid, Spain; the Servicio Navarro de Salud-Osasunbidea (PB-C, JD-E, and FJB-G), Pamplona, Spain; the Departments of Preventive Medicine and Public Health (IZ, ET, and MÁM-G) and Nutrition (AM), University of Navarra, Pamplona, Spain; the Department of Preventive Medicine, University of Valencia, Valencia, Spain (DC and JVS); the Human Nutrition Department, Hospital Universitari Sant Joan, Institut d Investigació Sanitaria Pere Virgili, Universitat Rovira i Virgili, Reus, Spain (JS-S and JB); the Lipid Clinic, Department of Endocrinology and Nutrition (ER), the Department of Internal Medicine (RE), Institut d Investigacions Biomèdiques August Pi i Sunyer, Hospital Clinic, and the Department of Nutrition and Food Science, School of Pharmacy, Xarxa de Referència en Tecnologia dels Aliments, Instituto de Investigación ennutrición y Seguridad Alimentaria (RML-R), University of Barcelona, Barcelona, Spain (ER); the Department of Preventive Medicine, University of Malaga, Malaga, Spain (JF-CN); the Department of Family Medicine, Primary Care Division of Seville, San Pablo Health Center, Seville, Spain (JMS); the Department of Cardiology, University Hospital of Alava, Vitoria, Spain (FA); the Institute of Health Sciences, University of Balearic Islands, and Hospital Son Espases, Palma de Mallorca, Spain (MF); the Cardiovascular and Nutrition Research Group, Institut de Recerca Hospital del Mar, Barcelona, Spain (OC); the Department of Clinical Sciences, University of Las Palmas de Gran Canaria, Las Palmas, Spain (LS-M); the Lipids and Vascular Risk Unit, Internal Medicine, Hospital Universitario de Bellvitge, Hospitalet de Llobregat, Barcelona, Spain (XP); the Instituto de la Grasa, Consejo Superior de Investigaciones Científicas, Sevilla, Spain (VR-G); and the Primary Care Division, Catalan Institute of Health, Barcelona, Spain (JMV-R). 2 Supported by the official funding agency for biomedical research of the Spanish government, Instituto de Salud Carlos III, through grants provided to research networks specifically developed for the trial [Red Temática de Investigación Cooperativa en Salud (RTIC) G03/140 to RE; RTIC RD 06/0045 to MÁM-G] and the Centro de Investigación Biomédica en Red de Fisiopatología de la Obesidad y Nutrición and grants from Centro Nacional de Investigaciones Cardiovasculares (06/2007), the Fondo de Investigación Sanitaria Fondo Europeo de Desarrollo Regional [Proyecto de Investigación 1498 Am J Clin Nutr 2014;100:1498 507. Printed in USA. Ó 2014 American Society for Nutrition

FIBER AND MORTALITY 1499 not the sole reason for the preventive effects of these foods. Different components present in their natural matrix may be relevant and even more important than their fiber contents. This possibility raises the interest of simultaneously assessing the role of fiber-rich foods and not only total DF intake. In addition, only a limited number of observational studies have examined the effect of DF intake on total mortality and have reported inconsistent results (7, 9, 11). Few prospective studies have shown an inverse association between fruit and vegetable consumption and all-cause mortality. In some of these studies, fruit and vegetable intakes were analyzed together (12 15), and fruit and vegetable consumption was inversely associated with total mortality. Contrarily, in studies in which the association between fruit and vegetable intakes were analyzed separately, results were inconsistent. In some of these studies, a high vegetable consumption or high vegetable fiber intake were inversely associated with total mortality, whereas no significant association between fruit consumption and mortality was observed (7, 16). In other studies, fruit consumption but not vegetable consumption was inversely associated with total mortality (17 19). These available prospective studies have been conducted in cohorts from the general population with low CVD risk, and some of them explicitly excluded participants with baseline diabetes or dyslipidemia (3). To the best of our knowledge, no prospective study has assessed this association in a cohort entirely composed of subjects at high CVD risk. Furthermore, the usual approach in large cohorts has been limited to a single baseline assessment of diet. The need for updating the information on dietary habits by using repeated questionnaires has been advocated because the use of a single measurement of dietary habits at baseline may lead to an increased measurement error, unrealistic assumptions about the induction period, and, consequently, a potential attenuation of the true association (20, 21). In this context, it is likely that the use of cumulative averages of intake, which reflect long-term diet, is more relevant etiologically than the use of either the most-remote (baseline) or most-recent diet. We used yearly repeated measurements of dietary information to investigate the association of fiber intake and whole grain, fruit, and vegetable consumption with all-cause (PI) 04-2239, PI05/2584, CP06/00100, PI07/0240, PI07/1138, PI07/0954, PI 07/0473, PI10/01407, PI10/02658, PI10/01407, PI11/01647, PI11/02505, and PI13/00462], the Ministerio de Ciencia e Innovación [Recursos y Tecnologías Agroalimentairas (AGL)-2009-13906-C02, AGL2010-22319-C03, and AGL2013-49083-C3-1-R], the Fundación Mapfre 2010, the Consejería de Salud de la Junta de Andalucía (PI0105/2007), the Public Health Division of the Department of Health of the Autonomous Government of Catalonia, the Generalitat Valenciana [Generalitat Valenciana AyudaComplementaria (GVACOMP) 06109, GVACOMP2010-181, GVACOMP2011-151, Conselleria de Sanidad y AP, AtenciónPrimaria (CS) 2010-AP-111, and CS2011- AP-042], and the Regional Government of Navarra (P27/2011). ET is supported by a Rio Hortega postresidency fellow of the Instituto de Salud Carlos III, Ministry of Economy and Competitiveness, Spanish Government. 3 Address correspondence to MÁ Martínez-González, Department of Preventive Medicine & Public Health, School of Medicine, University of Navarra, Irunlarrea 1, 31008-Pamplona, Navarra, Spain. E-mail: mamartinez@unav.es. 4 Abbreviations used: CVD, cardiovascular disease; DF, dietary fiber; FFQ, food-frequency questionnaire; PREDIMED, Prevención con Dieta Mediterránea. Received June 23, 2014. Accepted for publication August 21, 2014. First published online September 10, 2014; doi: 10.3945/ajcn.114.093757. mortality in a Mediterranean cohort of elderly adults at high CVD risk. SUBJECTS AND METHODS The current cohort study was conducted within the framework of the Prevención con Dieta Mediterránea (PREDIMED) trial, the design of which has been described in detail elsewhere (22, 23). Briefly, the PREDIMED study was a large, parallel-group, multicenter, randomized, controlled field trial that aimed to assess the effects of the Mediterranean diet on the primary prevention of CVD (www.predimed.es) (controlled-trials.com; ISRCTN35739639). The PREDIMED study was conducted in Spain. Recruitment took place between October 2003 and January 2009 in Primary Health Clinics by their family practitioners. We randomly assigned 7447 participants to 1 of 3 interventions (2 Mediterranean diets enriched with extra-virgin olive oil or mixed nuts and a control low-fat diet). Participants were men aged 55 80 y and women aged 60 80 y who were free of CVD at baseline but who had either type 2 diabetes or met $3 ofthe following criteria for CVD risk factors: current smoking, hypertension (blood pressure $140/90 mm Hg or treatment with an antihypertensive medication), a high plasma LDLcholesterol concentration ($160 mg/dl or lipid-lowering therapy), a low plasma HDL-cholesterol concentration (#40 in men and #50 mg/dl in women), overweight or obesity [BMI (in kg/m 2 ) $25], and family history of premature CVD (aged #55yinmenand#60yinwomen).Exclusioncriteria for the PREDIMED study were the presence of any severe chronic illness, previous history of CVD, alcohol or drug abuse, BMI $40, and history of allergy or intolerance to olive oil or nuts. The intervention took place in 11 different centers. At baseline and in yearly follow-up visits, trained personnel performed anthropometric and blood pressure measurements and obtained samples of fasting blood. Dietary assessment Dietary intake was measured with the use of a validated foodfrequency questionnaire (FFQ) repeatedly administered every year. The validation study indicated a correlation between intakes of fiber, fruit, and vegetables on the dietary questionnaire and intake assessed on four 3-d dietary records of 0.66, 0.72, and 0.81, respectively (24). The reproducibility was also validated (25). The FFQ included 137 food items, and frequencies of consumption of food items were reported on an incremental scale with 9 levels (never or almost never; 1 3 times/mo; 1, 2 4, and 5 6 times/wk; and1, 2 3, 4 6, and.6 times/d). Nutrient intakes were computed by using Spanish food-composition tables (26). Primary endpoint The primary endpoint was death from any cause. We used the following 4 sources of information to identify deaths: contacts with families of participants, contacts with general practitioners who were responsible for the routine clinical care of participants, yearly consultation of the National Death Index, and a comprehensive yearly review of medical records of all participants by medical doctors who were blinded with respect to the group allocation and all nutritional information. All medical records related to endpoints were examined by the Event Adjudication Committee, whose

1500 BUIL-COSIALES ET AL. members were unaware of the dietary information. Only endpoints that were confirmed by the Event Adjudication Committee and occurred between 1 October 2003 and 30 June 2012 (maximum: 8.7y;mean:5.9y)wereincludedinthisanalysis. Statistical analysis Participants were categorized into fifths of consumption for each food group. The sample size was estimated by assuming a 2-tailed a error = 0.05, RR = 0.60, absolute risk (cumulative incidence) = 6%, and statistical power = 0.80. Under these assumptions, the required sample size in each compared group was 1327 subjects, which was covered with the number of participants in each of the extreme fifths. Tests for trend included all data and had higher statistical power. We used the residuals method to adjust DF, fruit, vegetable, and whole-grain consumption for total energy intake (27). We excluded participants with total energy intake out of predefined limits (800 and 4000 kcal/d for men and 500 and 3500 kcal/d for women) Baseline characteristics are presented according to fifths of baseline intake of DF as the mean (6SD) for quantitative traits and n (%) for categorical variables. Cox regression models were used to assess the relation between categories of baseline fiber, fruit, vegetable, and wholegrain intakes and death during follow-up. We also analyzed separately the consumption of high and low glycemic index fruit (28) and, in addition, considered different causes of death. HRs and their 95% CIs were calculated by using the group with the lowest consumption (,20th percentile) as the reference category. The entry time was defined as the date at recruitment. The exit time was defined as the date at death or the date when completing the last interview was completed or 30 June 2012, whichever came first. To minimize any effects of a variation in diet, we also calculated the average of food consumption by using yearly updated information for each participant from repeated FFQs collected at baseline and during follow-up after 1, 2, 3, 4, 5, or 6 y. Multivariable models were adjusted for known or suspected predictors of death. First we adjusted only for age and sex and stratified by recruitment center. In model 2, additional adjustments for smoking status (nonsmokers, former smokers, and current smokers), diabetes (yes or no), BMI, and systolic and diastolic arterial blood pressures (mm Hg) were done. Model 3 was the same as model 2 with additional adjustments for the use of statins (yes or no), alcohol consumption (g/d), educational level (up to primary school, university graduate, or others), physical activity (metabolic equivalent tasks per day), and total energy intake (kcal/d). In a final model, we also adjusted for vegetable (g/d) and whole-grain (g/d) consumption in the fruitconsumption analysis, fruit (g/d) and whole-grain (g/d) consumption in analyses that assessed vegetable consumption, and vegetable (g/d) and fruit (g/d) consumption in analyses of wholegrain intake. To test for the linear trend across successive categories, we created a continuous variable with the median value within each category and regressed risk of death on this variable. Finally, we estimated the average of all repeated measurements of intake during follow-up (i.e., by excluding the baseline questionnaire and using only questionnaires that assessed dietary habits after the intervention) and created 4 joint categories as follows: participants whose intakes was low (lowest fifth) in the baseline questionnaire and did not increase their intakes (low/ low), those whose intakes were adequate (4 upper categories) at baseline but whose intakes during follow-up went down below the lowest baseline quintile (20 g for fiber intake and 210 g for fruit consumption) (adequate/low), those with low consumption at baseline but who increased their consumption beyond the rank of these same limits (fiber.20 g/d; fruit.210 g/d) during follow-up (low/adequate), and participants with an appropriate consumption in both baseline and follow-up measurements (adequate/adequate). Cox regression models were used to assess the relation between these categories and all-cause mortality after adjustment for the same variables mentioned previously. Statistical tests were 2-sided, and P, 0.05 was considered to indicate statistical significance. Statistical analyses were performed with STATA software (version SE 11, 2009; StataCorp LP). RESULTS From October 2003 through June 2009, a total of 8713 candidates were screened for eligibility, and 7447 subjects were randomly assigned to 1 of 3 study groups. Of these subjects, 153 participants who were outside the limits for total energy intake at baseline were excluded. We also excluded 78 participants with missing baseline dietary information. Thus, analyses were carried out in 7216 participants. Over 5.9 y of mean follow-up, 425 participants died. An analysis of causes of death showed that 169 participants died of cancer (40%); 153 subjects died from no-cancer, no-cvd death (36%); and 103 subjects died of CVD (24%). The comparison of participants characteristics according to their baseline DF intakes is shown in Table 1. No differences were observed across categories of DF intake in the distribution of classical confounding factors. As expected, participants with higher DF intake consumed more fruit, vegetables, and whole grains and less refined grains. They also consumed more omega-3 fatty acids and less MUFAs. They were also more likely to use statins. In our population, the most-frequently consumed vegetables were in this order: tomatoes; the group of lettuce, endive, and curly endive; and the group zucchini, eggplant, and cucumber. The most-frequently consumed fruit was as follows: the group of apples and pears, the citrus fruit group (oranges, tangerines, and grapefruit), and watermelon. Finally, main sources of cereal fiber were whole-grain bread, white bread, and pasta. When fiber intake was considered as a whole, main sources of fiber were whole-grain bread, apples and pears, and the citrus fruit group. Age and sex-adjusted and multivariable-adjusted analyses showed a significant inverse association between baseline fiber consumption and all-cause mortality (Table 2). Compared with the lowest group of DF intake, participants in the highest fifth of DF intake had 37% lower RR of death. In separate analyses for the consumption of fruit, vegetables and whole grains, the inverse association was similar for the consumption of fruit, but results neither for vegetables nor whole grains were significant. After follow-up, we showed higher fiber intake because of increased consumption of fruit, vegetables, and whole grains that was reflected in higher values for medians of the first quartile for 4 nutritional variables. In the cumulative analysis that used repeated measurements of diet, we showed significant inverse associations of DF intake and fruit consumption with all-cause

FIBER AND MORTALITY 1501 TABLE 1 Demographic and lifestyle characteristics and daily nutrient intakes according to baseline quintiles of dietary fiber intake among participants in the PREDIMED study 1 Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 Dietary fiber intake, 2 g/d 17 21 24 28 35 n 1444 1443 1443 1443 1443 Age, y 67 6 6.3 3 67 6 6.4 67 6 6.1 67 6 6.2 67 6 6.0 Male sex, % 43 43 43 43 43 BMI, 4 kg/m 2 30 6 3.9 30 6 3.8 30 6 3.8 30 6 4.0 30 6 3.8 Waist circumference, cm 101 6 11 101 6 10 100 6 10 100 6 11 100 6 9.9 Serum total cholesterol, 5 mg/dl 213 6 39 210 6 37 208 6 38 210 6 39 209 6 38 Serum triglycerides, 5 mg/dl 140 6 81 137 6 72 135 6 75 137 6 73 134 6 73 Serum HDL cholesterol, 5 mg/dl 54 6 14 53 6 15 54 6 15 54 6 14 54 6 14 Serum LDL cholesterol, 5 mg/dl 132 6 34 130 6 33 127 6 34 129 6 36 129 6 35 Systolic blood pressure, mm Hg 150 6 21 151 6 21 150 6 20 149 6 20 147 6 21 Diastolic blood pressure, mm Hg 84 6 11 84 6 11 83 6 11 83 6 10 82 6 11 Use of statin medication, % 36 39 40 44 43 Hypercholesterolemia, % 69 70 72 75 75 Olive oil intervention group, % 35 34 33 34 36 Nuts intervention group, % 30 33 34 33 33 Current smokers, % 19 16 13 11 11 Former smokers, % 24 24 25 25 25 Educational level, % University 5 3 3 5 4 High school 21 19 17 18 18 Physical activity, METs-min/d 194 6 210 218 6 215 236 6 241 250 6 247 258 6 271 Energy intake, kcal/d 2392 6 539 2184 6 534 2152 6 526 2154 6 523 2299 6 557 Carbohydrate, % of energy 39 6 7 40 6 6 41 6 7 43 6 7 45 6 7 Protein, % of energy 16 6 3 16 6 3 17 6 3 17 6 3 17 6 3 Fat, % of energy 42 6 7 40 6 6 40 6 7 38 6 7 36 6 6 SFAs, % of energy 11 6 2 10 6 2 10 6 2 10 6 2 9 6 2 MUFAs, % of energy 21 6 4 20 6 4 20 6 5 19 6 4 17 6 4 PUFAs, % of energy 6 6 2 6 6 2 6 6 2 6 6 2 6 6 2 Long-chain omega-3 fatty acids, g/d 0.77 6 0.47 0.78 6 0.45 0.78 6 0.55 0.81 6 0.49 0.85 6 0.51 Short-chain omega-3 fatty acids, g/d 1.40 6 0.60 1.31 6 0.65 1.33 6 0.64 1.38 6 0.76 1.46 6 0.77 Phytosterols, mg/d 327 6 101 320 6 99 332 6 989 350 6 101 399 6 113 Alcohol, g/d 13 6 20 9 6 14 7 6 12 7 6 11 6 6 10 Refined grains, g/d 141 6 88 132 6 90 119 6 91 103 6 87 71 6 78 Whole grains, g/d 4 6 14 7 6 21 17 6 32 31 6 45 89 6 93 Vegetables, g/d 244 6 99 285 6 99 331 6 112 366 6 126 445 6 191 Fruit, g/d 234 6 129 304 6 135 357 6 156 430 6 179 518 6 252 1 METs-min, metabolic equivalent task-minutes; PREDIMED, Prevención con Dieta Mediterránea. 2 All values are medians. 3 Mean 6 SD (all such values). 4 BMI is weight divided by the square of height. 5 Data were available for only 4260 participants (quintile 1: 927 subjects; quintile 2: 827 subjects; quintile 3: 813 subjects; quintile 4: 792 subjects; and quintile 5: 901 subjects). mortality in age- and sex-adjusted models (Table 3). The association became nonsignificant in the fully adjusted multivariable model for total DF intake, but it was still significant for fruit. When we compared participants with updated fruit consumption.210 g/d and those with consumption #210 g/d (below the first quintile at baseline), we showed an HR of all-cause death of 0.59 (95% CI: 0.44, 0.78) in the multivariable-adjusted model. We also compared participants who were below the first quintile at baseline and exhibited no increase in their consumption of fruit (or further decreased it) during the intervention compared with subjects who had an appropriate consumption at baseline and maintained it during the intervention (Figure 1). Participants who had low baseline fruit consumption and also low consumption of fruit during the trial had an HR of death of 2.64 (95% CI: 1.67, 4.17) compared with participants with adequate consumption both at baseline and during follow-up. HRs for other combinations of baseline fruit consumption and changes during the trial are also shown in Figure 1. Participants with low fruit consumption at baseline but adequate consumption during followup had significantly lower risk of mortality than did participants with steady low fruit consumption (HR: 0.62; 95% CI: 0.38, 1.00; P = 0.049). Regarding dietary fiber intake, compared with the reference category of adequate intake both at baseline and follow-up, participants who had low baseline intake but increased it during follow-up showed significantly higher mortality risk (HR: 1.42; 95% CI: 1.03, 1.95), as did participants with steady low dietary fiber intake (HR: 2.18; 95% CI: 1.45, 3.29). In additional analyses by the specific cause of death (Table 4), the significant protection provided by fiber intake and fruit consumption was apparent only for CVD mortality. Participants in

1502 BUIL-COSIALES ET AL. TABLE 2 HRs (95% CIs) of all-cause mortality according to baseline quintiles of fiber, fruit, vegetable, and whole-grain intakes Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 P-trend Dietary fiber intake (g/d) 1 17 21 24 28 35 n 1444 1443 1443 1443 1443 No. of deaths 116 80 88 74 67 Person-years 8618 8659 8563 8358 8267 Age, sex, and center adjusted 1 (reference) 2 0.63 (0.48, 0.84) 0.71 (0.54, 0.94) 0.63 (0.47, 0.85) 0.60 (0.44, 0.82) 0.004 Adjusted for risk factors 3 1 (reference) 0.63 (0.47, 0.83) 0.71 (0.54, 0.95) 0.64 (0.47, 0.87) 0.62 (0.45, 0.85) 0.010 Multivariable adjusted 4 1 (reference) 0.63 (0.47, 0.84) 0.72 (0.54, 0.96) 0.65 (0.48, 0.89) 0.63 (0.46, 0.86) 0.015 Fruit consumption (g/d) 1 153 256 339 439 613 n 1444 1443 1443 1443 1443 No. of cases (death) 121 77 90 73 64 Person-years 8794 8742 8411 8302 8215 Age, sex, and center adjusted 1 (reference) 0.61 (0.46, 0.81) 0.73 (0.55, 0.96) 0.58 (0.43, 0.79) 0.55 (0.40, 0.75),0.001 Adjusted for risk factors 3 1 (reference) 0.61 (0.46, 0.81) 0.73 (0.55, 0.96) 0.60 (0.44, 0.80) 0.56 (0.41, 0.76) 0.001 Multivariable adjusted 4 1 (reference) 0.61 (0.46, 0.81) 0.72 (0.54, 0.96) 0.60 (0.44, 0.82) 0.58 (0.42, 0.79) 0.002 Further adjusted 5 1 (reference) 0.61 (0.46, 0.82) 0.73 (0.55, 0.97) 0.61 (0.44, 0.84) 0.59 (0.42, 0.82) 0.004 Vegetable consumption (g/d) 1 178 255 316 386 503 n 1444 1443 1443 1443 1443 No. of cases (death) 104 82 96 82 61 Person-years 8669 8705 8443 8356 8291 Age, sex, and center adjusted 1 (reference) 0.82 (0.62, 1.10) 1.00 (0.76, 1.32) 0.87 (0.65, 1.16) 0.69 (0.50, 0.96) 0.052 Adjusted for risk factors 3 1 (reference) 0.84 (0.63, 1.12) 1.00 (0.76, 1.33) 0.86 (0.64, 1.16) 0.70 (0.50, 0.97) 0.051 Multivariable adjusted 4 1 (reference) 0.85 (0.63, 1.13) 1.01 (0.77, 1.34) 0.89 (0.66, 1.19) 0.72 (0.52, 1.00) 0.083 Further adjusted 6 1 (reference) 0.87 (0.65, 1.16) 1.04 (0.79, 1.38) 0.92 (0.68, 1.25) 0.77 (0.55, 1.08) 0.203 Whole-grain consumption (g/d) 1,7 0 5 19 84 n 2887 1443 1443 1443 No. of cases (death) 168 96 90 71 Person-years 17,146 8446 8293 8580 Age, sex, and center adjusted 1 (reference) 0.73 (0.49, 1.10) 0.66 (0.42, 1.02) 0.90 (0.62, 1.30) 0.139 Adjusted for risk factors 3 1 (reference) 0.75 (0.50, 1.12) 0.68 (0.44, 1.05) 0.89 (0.62, 1.29) 0.141 Multivariable adjusted 4 1 (reference) 0.76 (0.51, 1.14) 0.65 (3.42, 1.01) 0.92 (0.64, 1.32) 0.102 Further adjusted 8 1 (reference) 0.78 (0.52, 1.17) 0.67 (0.43, 1.04) 0.92 (0.64, 1.33) 0.124 1 All values are medians. 2 HR; 95% CI in parentheses (all such values). HRs were obtained from Cox models. 3 Adjusted for age, sex, smoking status, diabetes, BMI, baseline systolic and diastolic arterial blood pressures, and intervention group and stratified by recruitment center. 4 Further adjusted for use of statins, alcohol intake, educational level, physical activity, and total energy intake. 5 Further adjusted for vegetable and whole-grain consumption. 6 Further adjusted for fruit and whole-grain consumption. 7 Quintiles 1 and 2 were merged because of the large number of participants who did not consume whole grains. 8 Further adjusted for fruit and vegetable consumption. the highest category of fiber intake showed 54% lower risk of CVD mortality than that of participants in the lowest fifth of fiber intake (P-trend = 0.059). Also, participants in the highest fruitconsumption category showed significantly lower cardiovascular death risk (HR: 0.44; 95% CI: 0.22, 0.89). DISCUSSION In our study, that included elderly adults at high risk of CVD, higher baseline intakes of fiber or fruit were inversely associated with all-cause mortality. The observed dose-response pattern was compatible with a threshold effect. Thus, a person who consumed $210 g fruit/d or had fiber intakes of $20 g/d exhibited significantly lower risk of death. In our models, we adjusted for the more-important risk factors, and estimates did not substantially change. We observed reduced risk of death for participants who had low baseline fruit consumption and increased their consumption during the trial compared with for participants with steady low fruit consumption. When we analyzed separately low and high glycemic index fruit, we observed no substantial differences between both types of fruit. The observed reduction in all-cause mortality was mainly driven by reduced CVD mortality. We did not show any significant association between vegetable or whole-grain consumption and total mortality. This result might be explained by the already high consumption of vegetables in our population at baseline (the lowest quintile of vegetable consumption was 224 and 220 g/d for women and men, respectively) and low between-subject variability in whole grains in this population who consumed almost exclusively refined cereals (mainly white bread) (28). Our results for fiber intake were consistent in direction and magnitude with those from other cohort studies (7, 10, 11, 30, 31) (ORs between 0.57 and 0.96) and in the same direction but with a greater magnitude for fruit consumption (ORs between 0.79 and 0.92) (14, 16, 18, 19, 32, 33). The results also were consistent with results for vegetable consumption in these studies; only

FIBER AND MORTALITY 1503 TABLE 3 HRs (95% CIs) of death according to cumulative quintiles of fiber, fruit, vegetable, and whole-grain intakes Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5 P-trend Fiber consumption (g/d) 1 19 22 25 28 33 Age, sex, and center adjusted 1 (reference) 2 0.71 (0.54, 0.94) 0.65 (0.48, 0.87) 0.62 (0.45, 0.84) 0.67 (0.49, 0.91) 0.014 Adjusted for risk factors 3 1 (reference) 0.70 (0.53, 0.93) 0.67 (0.50, 0.90) 0.64 (0.47, 0.87) 0.71 (0.52, 0.97) 0.045 Multivariable adjusted 4 1 (reference) 0.72 (0.54, 0.96) 0.70 (0.51, 0.95) 0.66 (0.48, 0.92) 0.74 (0.54, 1.02) 0.095 Fruit consumption (g/d) 1 203 299 365 443 585 Age, sex, and center adjusted 1 (reference) 0.64 (0.48, 0.85) 0.51 (0.37, 0.68) 0.63 (0.47, 0.85) 0.56 (0.41, 0.77) 0.001 Adjusted for risk factors 3 1 (reference) 0.65 (0.49, 0.86) 0.52 (0.38, 0.70) 0.66 (0.49, 0.88) 0.58 (0.43, 0.80) 0.004 Multivariable adjusted 4 1 (reference) 0.66 (0.49, 0.88) 0.52 (0.38, 0.71) 0.68 (0.50, 0.93) 0.59 (0.43, 0.82) 0.007 Further adjusted 5 1 (reference) 0.66 (0.50, 0.89) 0.53 (0.39, 0.73) 0.70 (0.51, 0.96) 0.61 (0.43, 0.85) 0.016 Vegetable consumption (g/d) 1 206 276 328 381 477 Age, sex, and center adjusted 1 (reference) 0.72 (0.54, 0.96) 0.71 (0.53, 0.95) 0.70 (0.52, 0.94) 0.74 (0.54, 1.02) 0.057 Adjusted for risk factors 3 1 (reference) 0.74 (0.56, 0.98) 0.72 (0.53, 0.96) 0.72 (0.53, 0.97) 0.76 (0.55, 1.05) 0.083 Multivariable adjusted 4 1 (reference) 0.74 (0.56, 0.99) 0.74 (0.55, 0.99) 0.73 (0.54, 0.98) 0.75 (0.54, 1.05) 0.092 Further adjusted 6 1 (reference) 0.76 (0.57, 1.01) 0.77 (0.57, 1.03) 0.77 (0.56, 1.05) 0.81 (0.57, 1.15) 0.240 Whole-grain consumption (g/d) 1 0 1 7 33 89 Age, sex, and center adjusted 1 (reference) 0.99 (0.74, 1.33) 0.90 (0.67, 1.21) 0.93 (0.69, 1.25) 0.91 (0.67, 1.24) 0.389 Adjusted for risk factors 3 1 (reference) 0.96 (0.72, 1.29) 0.86 (0.63, 1.16) 0.93 (0.68, 1.26) 0.90 (0.66, 1.23) 0.437 Multivariable adjusted 4 1 (reference) 0.89 (0.64, 1.24) 0.79 (0.54, 1.15) 0.86 (0.60, 1.24) 0.88 (0.62, 1.24) 0.651 Further adjusted 7 1 (reference) 0.90 (0.65, 1.26) 0.81 (0.55, 1.19) 0.90 (0.63, 1.30) 0.93 (0.65, 1.31) 0.882 1 All values are medians. Medians for each group were calculated by using the food residual (energy-adjusted) model as proposed by Willett et al. (27, 29). We computed residuals of food intake by removing the variation caused by each individual s total energy intake; thus the food intakes of the individuals were regressed on their total energy intakes. We added a constant (the predicted nutrient intake at the mean total energy intake) to the residuals from this regression (differences between each individual s actual intake and the intake predicted by their total energy intake, with mean ¼ 0) to obtain energy-adjusted intakes. 2 HR; 95% CI in parentheses (all such values). HRs were obtained from Cox models. 3 Adjusted for age, sex, smoking status, diabetes, BMI, baseline systolic and diastolic arterial blood pressures, and intervention group and stratified by recruitment center. 4 Further adjusted for use of statins, alcohol intake, educational level, physical activity, and total energy intake. 5 Further adjusted for vegetable and whole-grain consumption. 6 Further adjusted for fruit and whole-grain consumption. 7 Further adjusted for fruit and vegetable consumption. Sayoun et al. (16) and Lenders et al. (34) showed an inverse significant association between vegetable consumption and allcause mortality. Large cohort studies conducted in the United States and Europe (7, 9) showed a strong linear trend for the inverse association between fiber and mortality. In our study, we showed a threshold effect between the first and the second quintiles. The consumption observed in the PREDIMED trial was higher than that observed in these surveys. Concretely, participants in the PREDIMED trial in the second fifth of fiber intake consumed already.20 g/d, whereas in other studies this amount of fiber intake was reached only by participants between the second and third fifths (7) or between the third and fourth categories (9). It is known that a steady state plasma concentration of vitamin C is achieved at a dose of 200 mg acid ascorbic/d, but between 100 and 200 mg acid ascorbic/d, there is little increase in the plasma concentration (35). The fruit consumption in our study was higher than those observed in other previous studies (16, 17, 34) but similar to the reported amounts from studies conducted in Greece (18), Italy (15), and Spain (36). Agudo et al. (19), in the Spanish cohort of the European Prospective Investigation into Cancer and Nutrition study with fruit consumption similar to that FIGURE 1 HRs (95% CIs) for all-cause mortality according to fruit consumption in the Prevención con Dieta Mediterránea trial, 2003 2012. HRs were obtained from Cox models. Ref., reference.

1504 BUIL-COSIALES ET AL. TABLE 4 HRs (95% CIs) for specific causes of death according to baseline quintiles of fiber, fruit, vegetable, and whole-grain intake CVD deaths (n = 103) Cancer deaths (n = 169) Other-cause deaths (n = 153) Fiber intake 1 Quintile 1 (17 g/d) 2 1 (reference) 3 1 (reference) 1 (reference) Quintile 2 (21 g/d) 0.55 (0.31, 0.95) 0.67 (0.42, 1.06) 0.66 (0.41, 1.09) Quintile 3 (24 g/d) 0.57 (0.32, 1.02) 0.78 (0.50, 1.23) 0.79 (0.49, 1.29) Quintile 4 (28 g/d) 0.65 (0.35, 1.19) 0.59 (0.35, 0.98) 0.77 (0.46, 1.29) Quintile 5 (35 g/d) 0.46 (0.23, 0.93) 0.69 (0.42, 1.14) 0.72 (0.44, 1.19) P-trend 0.059 0.177 0.379 Fruit consumption 4 Quintile 1 (153 g/d) 1 (reference) 1 (reference) 1 (reference) Quintile 2 (257 g/d) 0.51 (0.28, 0.93) 0.66 (0.43, 1.03) 0.68 (0.41, 1.12) Quintile 3 (339 g/d) 0.64 (0.36, 1.12) 0.61 (0.38, 0.99) 0.98 (0.61, 1.58) Quintile 4 (439 g/d) 0.33 (0.17, 0.66) 0.83 (0.52, 1.34) 0.66 (0.38, 1.14) Quintile 5 (613 g/d) 0.44 (0.22, 0.89) 0.63 (0.37, 1.07) 0.71 (0.42, 1.21) P-trend 0.015 0.219 0.250 Vegetable consumption 5 Quintile 1 (178 g/d) 1 (reference) 1 (reference) 1 (reference) Quintile 2 (255 g/d) 0.88 (0.50, 1.57) 0.79 (0.50, 1.25) 0.97 (0.58, 1.63) Quintile 3 (316 g/d) 1.02 (0.58, 1.80) 1.11 (0.72, 1.71) 0.96 (0.57, 1.61) Quintile 4 (386 g/d) 0.81 (0.43, 1.51) 0.66 (0.40, 1.10) 1.42 (0.87, 2.32) Quintile 5 (503 g/d) 0.83 (0.40, 1.72) 0.60 (0.34, 1.03) 0.98 (0.57, 1.69) P-trend 0.568 0.051 0.630 Whole-grain consumption 6 Quintile 1 (0 g/d) 1 (reference) 1 (reference) 1 (reference) Quintile 2 (1 g/d) 0.79 (0.39, 1.59) 0.94 (0.55, 1.62) 0.89 (0.49, 1.63) Quintile 3 (5 g/d) 0.76 (0.33, 1.76) 0.91 (0.48, 1.73) 1.03 (0.52, 2.02) Quintile 4 (19 g/d) 0.81 (0.35, 1.90) 0.83 (0.40, 1.69) 1.03 (0.51, 2.10) Quintile 5 (84 g/d) 0.73 (0.34, 1.58) 0.75 (0.40, 1.41) 0.76 (0.38, 1.53) P-trend 0.608 0.318 0.291 1 Adjusted for age, sex, smoking status, diabetes, BMI, baseline systolic and diastolic arterial blood pressure, and intervention group use of statins, alcohol intake, educational level, physical activity, and total energy intake and stratified by recruitment center. 2 Median (all such values). Medians for each group were calculated by using the food residual (energy-adjusted) model as proposed by Willett et al. (27, 29). We computed residuals of food intake by removing the variation caused by each individual s total energy intake; thus the food intakes of the individuals were regressed on their total energy intakes. We added a constant (the predicted nutrient intake at the mean total energy intake) to the residuals from this regression (differences between each individual s actual intake and the intake predicted by their total energy intake, with mean ¼ 0) to obtain energy-adjusted intakes. 3 HR for cause specific mortality; 95% CI in parentheses (all such values). 4 Adjusted for age, sex, smoking status, diabetes, BMI, baseline systolic and diastolic arterial blood pressures, and intervention group use of statins, alcohol intake, educational level, physical activity, total energy intake, and vegetable and whole-grain consumption and stratified by recruitment center. 5 Adjusted for age, sex, smoking status, diabetes, BMI, baseline systolic and diastolic arterial blood pressures, and intervention group use of statins, alcohol intake, educational level, physical activity, total energy intake, and fruit and whole-grain consumption and stratified by recruitment center. 6 Adjusted for age, sex, smoking status, diabetes, BMI, baseline systolic and diastolic arterial blood pressures, and intervention group use of statins, alcohol intake, educational level, physical activity, total energy intake, and fruit and vegetable consumption and stratified by recruitment center. in our study, showed an inverse linear trend, but after the second fifth (median: 225 g/d), a threshold effect was suggested. Some investigators (20, 21, 37) have postulated that the cumulative average of nutrient intake represents better the longterm integration of dietary exposures and helps to minimize any effects of within-person variations. Unlike other authors, we did not stop updating the dietary information after an intermediate outcome occurred because our participants had risk factors before study entry (e.g., dyslipidemia and diabetes). Also, we assumed that nutrient intakes before and after cardiovascular outcomes were equally important regarding death risk. Nevertheless, we are aware that this model likely underestimated the magnitude of the inverse association between nutrient intake and death (20). We showed an inverse association between fruit consumption and total mortality in the cumulative analyses. However, the inverse association in multiple-adjusted analyses was not statistically significant for fiber intake, although the magnitude of this association was similar to that shown in previous studies. Our participants received an intervention that encouraged the consumption of fiber, fruit, and vegetables, and we did enhance the consumption of these 3 elements, and thus, the contrast between the lowest and highest categories in the cumulative analysis was narrower than when we used only baseline dietary information. Thus, the narrower between-group contrast after follow-up may not

FIBER AND MORTALITY 1505 have been sufficient to detect a significant association. Moreover, the median in the lowest fifth with updated information was higher than the threshold shown in the analysis with baseline information. In contrast, in our cumulative analysis, participants who consumed highest amounts of fiber at baseline (probably reflecting that they were high-fruit consumers during most of their previous lives) were considered together with participants who increased only recently their fiber intake, and this merging of both types of participants could have brought on some degree of misclassification bias. This possibility could explain the nonsignificant results observed in multiple-adjusted cumulative average analyses. Neither the baseline nor cumulative consumption of vegetables or whole grains was significantly associated with mortality. Very few studies (16, 34) had shown a significant association between vegetable consumption and total mortality. Wang et al. (38) recently assessed, in a meta-analysis of prospective studies, the association between fruit and vegetable consumption and all-cause mortality. They reported significant reductions in total mortality and CVD mortality for fruit or vegetables, but they did not observe any significant association with cancer mortality. Some investigators (39) think that only some vegetables or some specific fruit are associated with risk of cancer. Our data were consistent with a wealth of existing studies in supporting the health benefits of fiber and fruit consumption; indeed, clinical trials have shown that both elements have beneficial effects on some intermediate markers of chronic diseases. In a subsample of the PREDIMED study, DF intake after a 3-mo behavioral intervention to enhance the Mediterranean diet (40) was associated with significant reductions in body weight, waist circumference, systolic and diastolic blood pressures, and fasting glucose as well as a greater increase in HDL cholesterol. In addition, when changes in the consumption of soluble DF were specifically assessed, participants in the highest category (.80th percentile) also showed a significantly greater reduction in serum LDL-cholesterol concentrations than did those,20th percentile. Fruit and vegetables contain a wide array of nutrients and phytochemicals that may act together to produce a benefit; vitamins, fiber, and minerals may confer cardioprotective, anticarcinogenic, anti-inflammatory and antioxidant properties. Rissanen et al. (14) showed that intakes of energy-adjusted folate, lycopene, and vitamins C and E explained 28% of the protective effect of fruit, berries, and vegetables against all-cause mortality. However, other nutrients besides them may be driving the association of fruit and vegetables with mortality. The current study had also some limitations. First, the method used for dietary assessment was an FFQ with the potential for misclassification bias. However, the FFQ was extensively validated, and the cumulative average estimate of consumptions may have been more robust than with a one-time assessment. Second, residual confounding might be a possible explanation of our findings. However, we have adjusted for many possible confounders. And third, the generalizability of our study results is limited because all participants lived in a Mediterranean country and were at high CVD risk. Our study had the following important strengths: a large sample size, the relative homogeneity of participants, the prospective design, the use of repeated dietary measurements during follow-up, and adjustment for a wide array of potential confounders. In conclusion, fiber and fruit intakes are associated with a reduced rate of total mortality. To the best of our knowledge, this study is the first to provide epidemiologic evidence with yearly repeated measurements of intake that increasing fruit and fiber consumption is associated with decreasing risk of all-cause mortality. The PREDIMED Study Investigators are as follows the University of Navarra, Primary Care Centers, Pamplona, Spain: B Sanjulian, JA Martínez M Serrano-Martinez, A Sánchez-Tainta, A Garcia-Arellano, JV Extremera- Urabayen, M Ruiz-Canela, Garcia-Perez L, Arroyo-Azpa C, Sola-Larraza A, Barcena F, Oreja-Arrayago C, Lasanta-Saez MJ, Amézqueta Go~ni C, Cia-Lecumberri P, Elcarte-Lopez T, Artal-Moneva F, Esparza-Lopez JM, Figuerido-Garmendia E, Tabar-Sarrias JA, Fernandez-Urzainqui L, Ariz- Arnedo MJ, Cabeza-Beunza JA, Pascual-Pascual P, Martinez-Mazo MD, Arina-Vergara E, Macua-Martínez T, and Parra-Oses A; the Hospital Clinic, Institut d Investigacions Biomediques August Pi i Sunyer, Barcelona, Spain: M Serra, A Perez-Heras, C Vinas, R Casas, L de Santamaria, S Romero, JM Baena, M Garcia, M Oller, J Amat, I Duaso, Y Garcia, C Iglesias, C Simon, Ll Quinzavos, Ll Parra, M Liroz, J Benavent, J Clos, I Pla, M Amoros, MT Bonet, MT Martin, MS Sanchez, J Altirruba, E Manzano, A Altes, M Cofan, C Valls-Pedret, A Sala-Vila, and M Domenech; the University Rovira i Virgili, Reus, Spain: M Bulló, R González, C Molina, FMárquez, N Babio, M Sorli, J García Roselló, M Guasch-Ferré, A Diaz- López, P Martínez, R Balanzà, BF Martin, R Tort, A Isach, B Costa, JJ Cabré, J Fernández-Ballart, N Ibarrola-Jurado, C Alegret, P Martínez, S Millán, JL Pi~nol, T Basora, and JM Hernández; the Institute de Recerca Hospital del Mar, Barcelona, Spain: M Fitó, S Tello, J Vila, R de la Torre, D Munoz-Aguayo, R Elosua, J Marrugat, and M Ferrer; the University of Valencia, Valencia, Spain: P Carrasco, R Osma, M Guillen, P Guillem-Saiz, O Portoles, V Pascual, C Riera, J Valderrama, A Serrano, E Lazaro, A Sanmartin, A Girbes, V Santamaria, C Sanchez, Z Pla, EM Asensio, and JI González; the University Hospital of Alava, Vitoria, Spain: I Salaverria, T del Hierro, J Algorta, S Francisco, A Alonso, J San Vicente, A Casi, E Sanz, I Felipe, J Rekondo, and A Loma-Osorio; the University of Malaga, Malaga, Spain: E Gómez-Gracia, J Wärnberg, R Benitez Pont, M Bianchi Alba, Navajas, R Gomez-Huelgas, J Martinez-Gonzalez, V Velasco Garcia, J de Diego Salas, A Baca Osorio, J Gil Zarzosa, JJ Sanchez Luque, and E Vargas Lopez; the Instituto de la Grasa, Consejo Superior de Investigaciones Cientificas, Sevilla, Spain: J Sanchez Perona, E Montero Romero, M Garcia-Garcia, and E Jurado-Ruiz; Instituto de Investigación Sanitaria de Palma (IdISPa), University of Balearic Islands, and Hospital Son Espases, Palma de Mallorca, Spain: M Garcia-Valdueza, M Monino, A Proenza, R Prieto, G Frontera, M Ginard, A Jover, D Romaguera, and J Garcia; the Department of Family Medicine, Primary Care Division of Sevilla, Sevilla, Spain: J Lapetra M Leal, E Martinez, M Ortega-Calvo, P Roman, P Iglesias, Y Corchado, E Mayoral, L Mellado, L Miró, JM Lozano, and C Lama; the School of Pharmacy, University of Barcelona, Barcelona, Spain: AI Castellote- Bargallo, A Medina-Remon, and A Tresserra-Rimbau; the University of Las Palmas de Gran Canaria, Las Palmas, Spain: J Alvarez-Perez, E díaz Benitez, I Bautista Casta~no, I Maldonado Diaz, A Sanchez-Villegas, MJ Férnandez- Rodríguez, F Sarmiento de la Fe, C Simon Garcia, I Falcon Sanabria, B Macias Gutierrez, AJ Santana Santana, MJ Rodríguez-Fernández, and JGarcía-Pastor; the Hospital Universitario de Bellvitge, Hospitalet de Llobregat, Barcelona, Spain: E de la Cruz, A Galera, Y Soler, F Trias, I Sarasa, E Padres, and E Corbella; the Primary Care Division, Catalan Institute of Health, Barcelona, Spain: MA Mu~noz, C Cabezas, E Vinyoles, MA Rovira, L Garcia, G Flores, P Baby, A Ramos, L Mengual, P Roura, MC Yuste, A Guarner, A Rovira, MI Santamaria, M Mata, C de Juan, and A Brau; and other investigators of the PREDIMED network: JA Tur (University of Balearic Islands), MP Portillo (University of Basque Country), and G Saez (University of Valencia). The authors responsibilities were as follows PB-C, DC, JS-S, JD-E, ER, RE, and MÁM-G: conceived the project; PB-C, IZ, ET, DC, JS-S, JD-E, JF-CN, JMS, OC, FA, MF, LS-M, XP, RE, and MÁM-G: conducted the research; PB-C, IZ, ET, and MÁM-G: analyzed the data; PB-C, IZ, ET, and MÁM-G: wrote the manuscript and had primary responsibility for the final content of the manuscript; and all authors: read and approved the final manuscript. ER reported serving on the board of and receiving travel support, as well as grant support through his institution, from the California

1506 BUIL-COSIALES ET AL. Walnut Commission; serving on the board of the Flora Foundation (Unilever); receiving lecture fees from Danone; and receiving grant support through his institution from Nutrexpa, Feiraco, Unilever, and Karo Bio. JS-S reported serving on the board of and receiving grant support through his institution from the International Nut and Dried Fruit Council, receiving consulting fees from Danone, and receiving grant support through his institution from Eroski and Nestle. LS-M reported serving on the boards of the Mediterranean Diet Foundation and the Beer and Health Foundation. XP reported serving on the board of and receiving grant support through his institution from the Residual Risk Reduction Initiative Foundation, serving on the board of Omegafort, serving on the board of and receiving payment for the development of educational presentations, receiving lecture fees from Danone, and receiving grant support through his institution from Unilever and Karo Bio. RE reported serving on the board of and receiving lecture fees from the Research Foundation on Wine and Nutrition, serving on the boards of the Beer and Health Foundation and the European Foundation for Alcohol Research, and receiving lecture fees from Cerveceros de Espanã. PB-C, IZ, ET, DC, JD-E, JF-CN, JMS-L, FA, MF, OC, RML-R, AM, FJB-G, JVS, JMV-R, JB, VR-G, and MAM-G had no conflicts of interest. Funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of data; or preparation, review, or approval of the manuscript. REFERENCES 1. Dauchet L, Amouyel P, Dallongeville J. Fruit and vegetable consumption and risk of stroke: a meta-analysis of cohort studies. Neurology 2005;65:1193 7. 2. He FJ, Nowson CA, MacGregor GA. Fruit and vegetable consumption and stroke: meta-analysis of cohort studies. Lancet 2006;367:320 6. 3. Pereira MA, O Reilly E, Augustsson K, Fraser GE, Goldbourt U, Heitmann BL, Hallmans G, Knekt P, Liu S, Pietinen P, et al. Dietary fiber and risk of coronary heart disease: a pooled analysis of cohort studies. Arch Intern Med 2004;164:370 6. 4. Dauchet L, Amouyel P, Hercberg S, Dallongeville J. Fruit and vegetable consumption and risk of coronary heart disease: a meta-analysis of cohort studies. J Nutr 2006;136:2588 93. 5. Streppel MT, Ocké MC, Boshuizen HC, Kok FJ, Kromhout D. Dietary fiber intake in relation to coronary heart disease and all-cause mortality over 40 y: the Zutphen Study. Am J Clin Nutr 2008;88:1119 25. 6. de Munter JS, Hu FB, Spiegelman D, Franz M, van Dam RM. Whole grain, bran, and germ intake and risk of type 2 diabetes: a prospective cohort study and systematic review. PLoS Med 2007;4:e261. 7. Chuang SC, Norat T, Murphy N, Olsen A, Tjønneland A, Overvad K, Boutron-Ruault MC, Perquier F, Dartois L, Kaaks L, et al. Fiber intake and total and cause-specific mortality in the European Prospective Investigation into Cancer and Nutrition cohort. Am J Clin Nutr 2012;96:164 74. 8. Jacobs DR Jr, Andersen LF, Blomhoff R. Whole-grain consumption is associated with a reduced risk of noncardiovascular, noncancer death attributed to inflammatory diseases in the Iowa Women s Health Study. Am J Clin Nutr 2007;85:1606 14. 9. Park Y, Subar A, Hollenbeck A, Schatzkin A. Dietary fiber intake and mortality in the NIH-AARP Diet and Health Study. Arch Intern Med 2011;171:1061 8. 10. Martínez-González MÁ, de la Fuente-Arrillaga C, López-del-Burgo C, Vázquez-Ruiz C, Benito S, Ruiz-Canela M. Low consumption of fruits and vegetables and risk of chronic disease: a review of the epidemiological evidence and temporal trends among Spanish graduates. Public Health Nutr 2011;14:2309 15. 11. Bazzano LA, He J, Ogden LG, Loria CM, Whelton PK. Dietary fiber intake and reduced risk of coronary heart disease in us men and women: the National Health and Nutrition Examination Survey I Epidemiologic Follow-up Study. Arch Intern Med 2003;163:1897 904. 12. Bazzano LA, He J, Ogden L, Loria CM, Vupputuri S, Myers L, Whelton PK. Fruit and vegetable intake and risk of cardiovascular disease in US adults: the first National Health and Nutrition Examination Survey Epidemiologic Follow-up Study. Am J Clin Nutr 2002;76:93 9. 13. Genkinger JM, Platz EA, Hoffman SC, Comstock JW, He KJ. Fruit, vegetable, and antioxidant intake and all-cause, cancer, and cardiovascular disease mortality in a community-dwelling population in Washington County, Maryland. Am J Epidemiol 2004;160:1223 33. 14. Rissanen TH, Voutilainen S, Virtanen JK, Venho B, Vanharanta M, Mursu J, Salonen JT. Low intake of fruits, berries and vegetables is associated with excess mortality in Men: the Kuopio Ischaemic Heart Disease Risk Factor (KIHD) Study. J Nutr 2003;133:199 204. 15. Bamia C, Trichopoulos D, Ferrari P, Overvad K, Bjerregaard L, Tjønneland A, Halkjær J, Clavel-Chapelon F, Kesse E, Boutron-Ruault MC, et al. Dietary patterns and survival of older Europeans: the EPIC- Elderly Study (European Prospective Investigation into Cancer and Nutrition). Public Health Nutr 2007;10:590 8. 16. Sahyoun NR, Jacques PF, Russell RM. Carotenoids, vitamins C and E, and mortality in an elderly population. Am J Epidemiol 1996;144: 501 11. 17. Strandhagen E, Hansson PO, Bosaeus I, Isaksson B, Eriksson E. High fruit intake may reduce mortality among middle-aged and elderly men. The Study of Men Born in 1913. Eur J Clin Nutr 2000;54:337 41. 18. Trichopoulou A, Costacou T, Bamia C, Trichopoulos D. Adherence to a Mediterranean diet and survival in a Greek population. N Engl J Med 2003;348:2599 608. 19. Agudo A, Slimani N, Ocké MC, Naska A, Miller AB, Kroke A, Bamia C, Karalis D, Vineis P, Palli D, et al. Consumption of vegetables, fruits and other plant foods in the European Prospective Investigation into Cancer and Nutrition (EPIC) cohorts from 10 European countries. Public Health Nutr 2002;5:1179 96. 20. Bernstein AM, Rosner B, Willet W. Cereal fiber and coronary heart disease: a comparison of modeling approaches for repeated dietary measurements, intermediate outcomes and long follow-up. Eur J Epidemiol 2011;26:877 86. 21. Hu FB, Stampfer MJ, Rimm E, Ascherio A, Rosner BA, Spiegelman D, Willett W. dietary fat and coronary heart disease: a comparison of approaches for adjusting for total energy intake and modeling repeated dietary measurements. Am J Epidemiol 1999;149:531 40. 22. Martínez-González MÁ, Corella D, Salas-Salvadó J, Ros E, Covas MI, Fiol M, Wärnberg J, Arós F, Ruíz-Gutiérrez V, Lamuela-Raventós RM, et al. Cohort profile: design and methods of the PREDIMED study. Int J Epidemiol 2012;41:377 85. 23. Estruch R, Ros E, Salas-Salvadó J, Covas MI, Corella D, Arós F, Gómez-Gracia E, Ruiz-Gutiérrez V, Fiol M, Lapetra J, et al. Primary prevention of cardiovascular disease with a Mediterranean diet. N Engl J Med 2013;368:1279 90. 24. Fernández-Ballart JD, Pi~nol JL, Zazpe I, Corella D, Carrasco P, Toledo E, Perez-Bauer M, Martínez-González MA, Salas-Salvadó J, Martín- Moreno JM. Relative validity of a semi-quantitative food-frequency questionnaire in an elderly Mediterranean population of Spain. Br J Nutr 2010;103:1808 16. 25. de la Fuente-Arrillaga C, Vázquez Ruiz Z, Bes-Rastrollo M, Sampson L, Martinez-González MA. Reproducibility of an FFQ validated in Spain. Public Health Nutr 2010;13:1364 72. 26. Moreiras O, Carbajal A, Cabrera L. Tablas de composición de alimentos. [Food composition tables.] 9th ed. Madrid, Spain: Editorial Pirámide, 2005 (in Spanish). 27. Willet W, Stampfer MJ. Implications of total energy intake for epidemiologic analisys. In: Willet W, ed. Nutritional epidemiology, 2nd ed. New York: Oxford University Press, 1998:273 301. 28. Jenkins DJ, Srichaikul K, Kendall CW, Sievenpiper JL, Abdulnour S, Mirrahimi A, Meneses C, Nishi S, He X, Lee S, et al. The relation of low glycaemic index fruit consumption to glycaemic control and risk factors for coronary heart disease in type 2 diabetes. Diabetologia 2011;54:271 9. 29. Willett WC, Howe GR, Kushi LH. Adjustment for total energy intake in epidemiologic studies. Am J Clin Nutr l997;65(suppl):1220s 8S. 30. Lubin F, Lusky A, Chetrit A, Dankner R. Lifestyle and ethnicity play a role in all-cause mortality. J Nutr 2003;133:1180 5. 31. Todd S, Woddward M, Tunstall-Pedoe H, Bolton-Smith C. Dietary antioxidant vitamins and fiber in the etiology of cardiovascular disease and all-cause mortality: results from the Scottish heart health study. Am J Epidemiol 1999;150:1073 80. 32. Key TJ, Thorogood M, Appleby PN, Burr ML. Dietary habits and mortality in 11000 vegetarians and health conscious people: results of a 17 year follow up. BMJ 1996;313:775 9. 33. Nagura J, Iso H, Watanabe Y, Maruyama K, Date CH, Toyoshima H. Fruit, vegetable and bean intake and mortality from cardiovascular disease among Japanese men and women: the JACC Study. Br J Nutr 2009;102:285 92. 34. Leenders M, Sluijs I, Ros M, Boshuizen HC, Slersema PD, Ferrari P, Tjønneland A, Olsen A, Boutron-Ruault MC, Clavel-Chapelon F, et al.