Dietary Marine ω-3 Fatty Acids and Incident Sight-Threatening Retinopathy in Middle-Aged and Older Individuals With Type 2 Diabetes: Prospective Investigation From the PREDIMED Trial

Key Points

Question  Do dietary long-chain ω-3 fatty acids protect against diabetic retinopathy?

Findings  A substudy of the PREDIMED randomized clinical trial analyzed as an observational longitudinal cohort, considering only participants with type 2 diabetes at baseline, showed after median follow-up of 6 years that those reporting intake of at least 500 mg/d of long-chain ω-3 fatty acids at baseline had a 46% decreased risk of sight-threatening diabetic retinopathy compared to those not meeting this target.

Meaning  Results suggest fish-derived long-chain ω-3 fatty acids are a healthy fat in this cohort.

Importance  Diabetic retinopathy (DR) is a devastating complication of individuals with type 2 diabetes mellitus. The retina is rich in long-chain ω-3 polyunsaturated fatty acids (LCω3PUFAs), which are substrate for oxylipins with anti-inflammatory and antiangiogenic properties. Experimental models support dietary LCω3PUFA protection against DR, but clinical data are lacking.

Objective  To determine whether LCω3PUFA intake relates to a decreased incidence of sight-threatening DR in individuals with type 2 diabetes older than 55 years.

Design, Setting, and Participants  In late 2015, we conceived a prospective study within the randomized clinical trial Prevención con Dieta Mediterránea (PREDIMED), testing Mediterranean diets supplemented with extra virgin olive oil or nuts vs a control diet for primary cardiovascular prevention. The trial was conducted in primary health care centers in Spain. From 2003 to 2009, 3614 individuals aged 55 to 80 years with a previous diagnosis of type 2 diabetes were recruited. Full data were available for 3482 participants (48% men; mean age 67 years).

Exposures  Meeting the dietary LCω3PUFA recommendation of at least 500 mg/d for primary cardiovascular prevention, as assessed by a validated food-frequency questionnaire.

Main Outcomes and Measures  The main outcome was incident DR requiring laser photocoagulation, vitrectomy, and/or antiangiogenic therapy confirmed by an external adjudication committee.

Results  Of the 3482 participants, 48% were men and the mean age was 67 years. A total of 2611 participants (75%) met target LCω3PUFA recommendation. During a median follow-up of 6 years, we documented 69 new events. After adjusting for age, sex, intervention group, and lifestyle and clinical variables, participants meeting the LCω3PUFA recommendation at baseline (≥500 mg/d) compared with those not fulfilling this recommendation (<500 mg/d) showed a 48% relatively reduced risk of incident sight-threatening DR, with a hazard ratio of 0.52 (95% CI, 0.31-0.88; P = .001). This association was slightly stronger for yearly updated LCω3PUFA intake (relative risk, 0.48; 95% CI, 0.28-0.82; P = .007).

Conclusions and Relevance  In middle-aged and older individuals with type 2 diabetes, intake of at least 500 mg/d of dietary LCω3PUFA, easily achievable with 2 weekly servings of oily fish, is associated with a decreased risk of sight-threatening DR. Our results concur with findings from experimental models and the current model of DR pathogenesis.

Trial Registration  clinicaltrials.gov Identifier: http://www.controlled-trials.com/ISRCTN35739639

The increasing prevalence of type 2 diabetes mellitus, coupled with an increased lifespan, has resulted in a steady rise of disability in older diabetic individuals.¹ A major concern for this population group is diabetic retinopathy (DR), a leading global cause of vision loss.² Given the economic and societal burden of DR, developing effective strategies to prevent or at least delay its onset is a major public health issue.³

The pathogenesis of DR is not yet fully understood, but inflammation, oxidative stress, and hipoxia-driven microvascular alterations play a pivotal role in the worsening of retinal function in individuals with diabetes.⁴ The retina is rich in long-chain ω-3 polyunsaturated fatty acids (LCω3PUFAs), particularly docosahexaenoic acid (DHA; C22:6n-3).⁵ Once released from cell membranes, these fatty acids are transformed to oxylipins with anti-inflammatory and antiangiogenic properties.^6,7 Because cell membrane DHA status is modifiable and dependent on intake, dietary DHA or consumption of its parent food oily fish has been suggested to protect against DR.⁵ Experimental studies have consistently reported a protective role of supplemental DHA or LCω3PUFA against DR^8,9 or neovascularization of the retina,^10,11 a hallmark of proliferative DR. However, to our knowledge, human data are lacking.

We hypothesized that LCω3PUFA intake with the usual diet relates to lower vision-threatening DR. To test this hypothesis, we longitudinally investigated the association of dietary LCω3PUFA intake with incident DR requiring laser photocoagulation, vitrectomy, and/or antiangiogenic therapy in a cohort of middle-aged and older individuals with type 2 diabetes enrolled into the Prevención con Dieta Mediterránea (PREDIMED) study, a nutrition intervention trial for the primary prevention of cardiovascular disease conducted in Spain.¹² Because of the lack of available recommendations for LCω3PUFA intake regarding DR and given the vascular nature of DR pathophysiology, we tentatively set the exposure target to meeting the recommendation for dietary LCω3PUFA for primary cardiovascular protection (500 mg/d),¹³ a goal that can be achieved by following the advice of the American Heart Association to consume 2 weekly servings of fish, preferably oily fish.¹⁴

This substudy, which was conceived in late 2015, was conducted within the frame of the PREDIMED trial (http://www.predimed.es), the design of which has been described in detail¹⁵ (http://www.controlled-trials.com/ISRCTN35739639). From October 2003 to June 2009, a total of 8713 candidates were screened for eligibility, and 7447 were randomly assigned to 1 of 3 interventions. Participants were men aged 55 to 80 years and women aged 60 to 80 years at high cardiovascular risk but with no cardiovascular disease at enrollment. Criteria for eligibility were the presence of either type 2 diabetes or at least 3 cardiovascular risk factors: current smoking, hypertension, dyslipidemia, overweight or obesity, and family history of early-onset coronary heart disease. The study protocol was conducted according to the guidelines laid down in the Declaration of Helsinki and approved by the institutional review boards of Hospital Clínic, Barcelona, Spain; Rovira i Virgili University, Reus, Spain; Hospital del Mar, Barcelona, Spain; University of Valencia, Valencia, Spain; University of Navarra, Navarra, Spain; University of Málaga, Málaga, Spain; Instituto de Investigación Sanitaria de Palma, Palma de Mallorca, Spain; San Pablo Health Center, Sevilla, Spain; University Hospital of Alava, Vitoria, Spain; University of Las Palmas de Gran Canaria, Las Palmas, Spain; and Hospital Universitari de Bellvitge, L'Hospitalet de Llobregat, Spain. Written informed consent was obtained from all study participants. The formal trial protocol can be found in Supplement 1.

The diagnosis of type 2 diabetes was based on at least 1 of the following criteria: current treatment with insulin or oral hypoglycemic drugs; fasting (no caloric intake at least for 8 hours) glucose at least 126 mg/dL in 2 determinations (to convert to millimoles per liter, multiply by 0.0555); casual glucose at least 200 mg/dL with polyuria, polydipsia, or unexplained weight loss; or glucose at least 200 mg/dL in 2 measurements after an oral glucose tolerance test. Participants were considered to have hyperlipidemia or hypertension if they had a previous diagnosis of these conditions and/or they were treated with cholesterol-lowering or antihypertensive agents, respectively. Smoking status was categorized into never, current, or past smoking according to self-reports. Physical activity was determined with a validated Spanish version of the Minnesota Leisure-Time Physical Activity questionnaire and expressed in minutes at a given metabolic equivalent per day.¹⁶ Anthropometric variables (height, weight, and waist circumference) and blood pressure were measured by standard methods.

Dietary intake was assessed at baseline and yearly during follow-up by using a 137-item semiquantitative food-frequency questionnaire validated for the PREDIMED study.¹⁷ In face-to-face interviews, participants were asked about the frequency of consumption of each food item during the past year, specifying usual portion sizes. Nine possibilities of frequency were offered, ranging from never to more than 6 times/d. Information on seafood products was collected in 8 items of the questionnaire (uncanned oily fish; lean fish; smoked/salted fish; mollusks; shrimp, prawn, and crayfish; octopus, baby squid and squid; oily fish canned in oil; and oily fish canned in salted water). Nutrient intakes were computed using Spanish food composition tables. The validation of the food-frequency questionnaire against 4 3-day food records showed energy-adjusted intraclass correlation coefficients of 0.51 for LCω3PUFA (P < .001).

After the screening visit, suitable candidates were randomly assigned to 1 of 3 interventions: Mediterranean diet (MeDiet) supplemented with extra virgin olive oil, MeDiet supplemented with nuts, or control diet (advice to reduce all dietary fat). Quarterly individual and group sessions were scheduled for the 2 MeDiet groups; in them, participants were educated on how to follow the MeDiet and received supplemental foods at no cost. Extra virgin olive oil (1 L/wk) was provided to 1 group and 30 g/d of mixed nuts (15 g walnuts, 7.5 g hazelnuts, and 7.5 g almonds) to the other group. Those allocated to the control group were educated on how to follow a low-fat diet and received small nonfood gifts. In each session, a dietary screener of adherence to the MeDiet was used to track diet changes. The score was determined by 12 questions on food consumption frequency and 2 questions on food consumption habits considered characteristic of the MeDiet (each question scored 0 or 1).¹⁸ For this particular substudy, because a higher adherence to the MeDiet among frequent consumers of seafood could introduce confounding, the question about seafood consumption was omitted from the brief screener; therefore, a 13-point score was used as a covariate (minimum, 0; maximum, 13).

Diabetic retinopathy was not an explicitly prespecified secondary outcome of the PREDIMED trial. However, given that nearly one-half of participants had type 2 diabetes, this complication was always included as a relevant outcome in all interim analyses supervised by the Data and Safety Monitoring Board.

New events were identified prospectively through yearly review of the medical records of each participant. The PREDIMED dietitians of the different teams were responsible for the accurate filling of the follow-up questionnaires, which included 3 items about noncardiovascular complications of diabetes. Suspected DR was investigated by clinical records made by ophthalmologists and hospital discharge reports. All the information was sent to the adjudication committee, which determined the validity of diagnoses based on the information received. Participants were considered to have sight-threatening DR if they had undergone laser photocoagulation, intravitreal antivascular endothelial growth factor injections, and/or vitreoretinal surgery. Only cases that were confirmed by the adjudication committee and that occurred between October 1, 2003, and December 1, 2012 (date of the last update in the extended follow-up of the PREDIMED cohort, 2 years after the end of the trial), were included in the analyses.

Person-time of follow-up was calculated as the interval between the randomization date and the earliest date of the follow-up contact at which a new event was identified, death from any cause, or date of the last contact visit, whichever came first.

After excluding participants free of type 2 diabetes at baseline (n = 3833) with missing data on year after diagnosis of type 2 diabetes (n = 25), food-frequency questionnaires (n = 36), 13-point score of adherence to MeDiet (n = 5), and those who reported total energy intakes outside predefined limits (>4000 or <800 kcal/d in men, and >3500 or <500 kcal/d in women; n = 64) or with implausible intakes of LCω3PUFA (>4 g/d; n = 2), 3482 study participants (n = 1151 control diet, n = 1236 MeDiet plus extra virgin olive oil, and n = 1095 MeDiet plus nuts) remained for inclusion in the analysis (eFigure in Supplement 2).

The exposure of interest was meeting the International Society for the Study of Fatty Acids and Lipids recommendation to consume at least 500 mg/d of LCω3PUFA (DHA + eicosapentaenoic acid [C20:5ω-3]) for primary cardiovascular prevention (yes/no), which was released in June 2004.¹³ Baseline differences in demographic, clinical, and selected dietary variables between groups of exposure were assessed by analysis of variance or χ² tests, as appropriate.

To assess the associations between meeting the recommendation for LCω3PUFA intake (yes/no) at baseline and the risk of incident sight-threatening DR, we used unadjusted, age- and sex-adjusted, and multivariable time-dependent Cox proportional hazard models including age, sex, body mass index, intervention group, year after diagnosis of diabetes (≤5 or >5), use of insulin (yes/no), use of oral hypoglycemic agents (yes/no), smoking status (never, former, or current smoker), systolic blood pressure, history of hypertension (yes/no), use of angiotensin-converting-enzyme inhibitor and/or angiotensin-II receptor blockers (yes/no), physical activity, and adherence to the MeDiet (13-point score) as potential confounders. All analyses were stratified by recruitment center. Additional analyses were done after stratifying for history of hypertension, year after diagnosis of diabetes, and use of insulin and oral hypoglycemic agents at baseline. We also used Cox regression models to assess the risk of the prespecified endpoint according to the joint categories of meeting target intake recommendations for LCω3PUFA intake (yes/no) and intervention group (3 groups, 2 dummy variables).

Several sensitivity analyses were conducted refitting Cox regression analyses for LCω3PUFA after (1) excluding early cases of sight-threatening DR that occurred in the 2 first years of follow-up; (2) including only events occurring after at least 3 years of follow-up; and (3) excluding cases occurring after 5 years of follow-up. Effect modification by sex, intervention group, and LCω3PUFA intake on incident DR was evaluated by calculating the likelihood ratio test between the fully adjusted model and the same model adding the interaction product term.

As a secondary analysis, we repeated the analyses using generalized estimating equations to assess the association for yearly updated LCω3PUFA intake. We assumed a binomial distribution with logit models and the unstructured matrix as the working correlation structure. For each 1-year follow-up period, we used as exposure the average of total LCω3PUFA intake of all repeated measurements from baseline to the beginning of that yearly period. We defined the cohort risk as participants who remained free of DR at the beginning of each 1-year follow-up period. Participants who had been classified as incident cases were excluded from subsequent follow-up analyses.

Finally, we explored the association for meeting the American Heart Association recommendation to consume at least 2 weekly servings of fish (particularly oily fish) at baseline. Analyses were done using SPSS software, version 19.0 (IBM Corp) and STATA software, version 14.0 (StataCorp).

The mean age of participants at inclusion was 67 years, and 48% of them were men. At baseline, from the whole cohort (n = 3482), 2611 participants (75%) met the target recommendation of LCω3PUFA intake. Table 1 summarizes the baseline clinical characteristics and treatment regimens of participants. Those not meeting the recommendation of LCω3PUFA intake were older, smoked less, had a higher prevalence of hypertension (and treatment with antihypertensive drugs), and were treated with insulin more frequently than those meeting the LCω3PUFA recommendation. Intake of energy and nutrients and consumption of key foods are shown in Table 2. In brief, compared with participants meeting the LCω3PUFA recommendation, those who did not also adhered less to the MeDiet, although they consumed less red meat.

During a mean follow-up of 6 years, we documented 69 new events. Table 3 shows the hazard ratios (HRs) and 95% CIs associated with the exposure of interest. After adjusting for intervention group and classic risk factors, participants reporting intake of at least 500 mg/d of eicosapentaenoic acid + DHA at baseline showed a 46% decreased risk of incident sight-threatening DR (HR, 0.52; 95% CI, 0.31-0.88; P = .001) in comparison with participants not meeting the recommendation of LCω3PUFA intake. Higher risk reductions were observed in participants with hypertension, those with diabetes of greater than 5 years duration, and those treated with insulin at baseline (eTable in Supplement 2). The Figure shows the multivariate-adjusted HRs by meeting the recommendation for LCω3PUFA at baseline by intervention group. When compared with the reference category (participants not meeting target intake and allocated to the control diet group, n = 326), those meeting the recommendation had a significant decreased risk regardless of the intervention group to which they were allocated. Notably, compared with the reference category, only participants not meeting the recommendation but allocated to the MeDiet with extra virgin olive oil intervention group (n = 280) had a borderline 63% (95% CI, 87-2) decreased risk of sight-threatening DR.

We conducted several sensitivity analyses to investigate possible sources of bias in the estimation of the relationship between meeting the recommendation for LCω3PUFA intake at baseline and sight-threatening DR (Table 3). The results of these analyses were consistent with the findings of the primary analysis. Additionally, the magnitude of the association was strengthened when using as exposure the yearly updated LCω3PUFA intake (HR, 0.48; 95% CI, 0.28-0.82; P = .007). Finally, a lower risk (HR, 0.41; 95% CI, 0.23- 0.72; P = .002) was also observed when grouping the participants by reporting to meet the advice of the American Heart Association to consume at least 2 weekly servings of oily fish at baseline (Table 4).

In this substudy of the PREDIMED trial analyzed as an observational longitudinal cohort and considering only participants with type 2 diabetes, we found that those reporting intake of at least 500 mg/d of LCω3PUFA (or at least 2 weekly servings of oily fish) at baseline had a significantly decreased risk of incident sight-threatening DR compared with those not meeting this target. The reduction remained significant when using as exposure the yearly updated LCω3PUFA intake. To the best of our knowledge, this is the first study in humans on LCω3PUFA and DR, and the results reinforce a notion heretofore only explored in experimental models. Our findings support the view that regular consumption of oily fish might be beneficial to delay the onset or progression of vascular diseases in arterial beds other than the coronary and cerebrovascular ones.

Epidemiologic studies assessing exposure to LCω3PUFA for the primary prevention of cardiovascular disease (in particular fatal coronary heart disease) prompted the debate of whether increasing LCω3PUFA intake is beneficial for patients with diabetes. The effectiveness of this approach has been questioned in recent years given the null reduction of the rate of cardiovascular events reported in the Outcome Reduction With an Initial Glargine Intervention (ORIGIN) trial of LCω3PUFA intake in patients with diabetes with a much lower background LCω3PUFA intake^19,20 coupled with a suggested increased risk of DR secondary to peroxidation of accumulated PUFA in the retina.²¹ However, no clincal data support a deleterious effect of LCω3PUFA on the retina. Actually, a small cross-sectional study reported that increased PUFA intake (although not distinguishing between ω-3 or ω-6 fatty acids) was associated with reduced odds of prevalent DR and lesser DR severity in patients with well-controlled diabetes, while no associations were found for those poorly controlled.²² Suprisingly, neither epidemiologic studies nor clinical trials focused on dietary LCω3PUFA and DR, although the link has repeatedly been proven in experimental models.^8-11 By reporting that regular LCω3PUFA intake seems to reduce the risk of the most severe form of DR, particularly in patients with hypertension or advanced diabetes, we reinforce the benefits of these particular fatty acid species in a critical population group.

Our results are mechanistically supported by experimental studies showing that anti-inflammatory LCω3PUFA-derived compounds, mostly lipoxins, resolvins, and protectins, delay progression of DR, particularly neovascularization, the hallmark of proliferative DR, and improve diabetic macular edema.¹¹ Nevertheless, we must underline that this study was conducted in a Mediterranean population with high adherence to a plant-based diet, as attested by relatively high mean MeDiet scores (Table 2). Therefore, it seems plausible that a high intake of antioxidants, a main feature of this dietary pattern, might contribute to counteracting the presumed membrane peroxidation of PUFA. Indeed, a prior analysis of the PREDIMED trial concerning diabetic complications by intervention arm described a significant reduction of incident DR in participants allocated to the MeDiet supplemented with extra virgin olive oil.²³ In our study, when considering the 3 intervention groups, a marginal additional reduction in DR was observed by meeting the LCω3PUFA target precisely in the MeDiet with extra virgin olive oil group (Figure 1). This suggests that phytochemicals with antioxidant and anti-inflammatory properties contained in extra virgin olive oil could overrun the protection afforded by LCω3PUFA.

Our study has limitations. First, this is a prospective analysis of a subsample in a larger clinical trial that was not specifically designed to examine incidence of sight-threatening DR. Second, the assessment of diabetes complications was not the primary end point for the PREDIMED trial, and no data on prevalent DR were available at baseline. Nonetheless, we took care to ensure that all cases of incident sight-threatening DR were diagnosed by experienced ophthalmologists and confirmed by an external adjudication committee. Additionally, individuals with DR at baseline would have been more likely to be diagnosed with complicated DR in the first 2 years of follow-up, and results after excluding events reported during this period were similar to those of the primary analysis. Third, we had a relatively low number of events with rates slightly lower than those observed in a Spanish general population,²⁴ resulting in imprecise estimates. Finally, because the study participants were middle-aged and older individuals with type 2 diabetes at high risk for cardiovascular disease living in a Mediterranean country, our findings cannot be easily extrapolated to other populations. There are also strengths to our study, such as a prospective design with a long duration of follow-up, validation of the food frequency questionnaire, adjustment for relevant confounders, and the confirmation of the primary results by both sensitivity analyses and cumulative average estimates of the exposure of interest, that provide a more robust measure than a single baseline dietary assessment.

In summary, we report that meeting the target of at least 500 mg/d of dietary LCω3PUFA is associated with a reduced incidence of severe DR in individuals older than 55 years with type 2 diabetes. Our findings, which are consistent with the current model of the pathogenesis of DR and data from experimental models, add to the notion of fish-derived LCω3PUFA as a healthy fat.

Corresponding Author: Aleix Sala-Vila, DPharm, PhD, Lipid Clinic, Department of Endocrinology and Nutrition, Villarroel 170, Edifici Helios, despatx 8, 08036 Barcelona, Spain (asala@clinic.ub.es).

Accepted for Publication: June 21, 2016.

Published Online: August 18, 2016. doi:10.1001/jamaophthalmol.2016.2906

Author Contributions: Drs Estruch and Ros had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Sala-Vila, Garcia-Layana, Martinez-Gonzalez, Estruch, Lapetra, Fito, Serra-Majem, Pinto, Ros.

Acquisition, analysis, or interpretation of data: Sala-Vila, Diaz-Lopez, Valls-Pedret, Cofan, Garcia-Layana, Lamuela-Raventos, Castaner, Zanon-Moreno, Martinez-Gonzalez, Toledo, Basora-Gallisà, Salas-Salvado, Corella, Gomez-Gracia, Fiol, Estruch, Lapetra, Fito, Aros, Serra-Majem, Pinto.

Drafting of the manuscript: Sala-Vila, Diaz-Lopez, Lamuela-Raventos.

Critical revision of the manuscript for important intellectual content: Diaz-Lopez, Valls-Pedret, Cofan, Garcia-Layana, Lamuela-Raventos, Castaner, Zanon-Moreno, Martinez-Gonzalez, Toledo, Basora-Gallisà, Salas-Salvado, Corella, Gomez-Gracia, Fiol, Estruch, Lapetra, Fito, Aros, Serra-Majem, Pinto, Ros.

Statistical analysis: Sala-Vila, Diaz-Lopez, Martinez-Gonzalez, Gomez-Gracia.

Obtaining funding: Diaz-Lopez, Martinez-Gonzalez, Salas-Salvado, Estruch, Serra-Majem, Ros.

Administrative, technical, or material support: Lamuela-Raventos, Castaner, Martinez-Gonzalez, Toledo, Salas-Salvado, Estruch, Fito, Aros, Pinto.

Study supervision: Garcia-Layana, Martinez-Gonzalez, Basora-Gallisà, Salas-Salvado, Corella, Estruch, Fito, Aros, Serra-Majem, Pinto, Ros.

Group Information: The PREDIMED Investigators are listed below.

Hospital Clínic, Institut d’Investigacions Biomèdiques August Pi i Sunyer, Barcelona, Spain: M. Serra-Mir, RD, A. Pérez-Heras, RD, C. Viñas, RD, R. Casas, BS, PhD, P. Villanueva, RN, E.S. Romero-Mamani, MD, J.M. Baena, MD, M. García, MD, M. Oller, MD, J. Amat, MD, I. Duaso, MD, Y. García, MD, C. Iglesias, MD, C. Simon, MD, Ll. Quinzavos, MD, Ll. Parra, MD, M. Liroz, MD, J. Benavent, MD, J. Clos, MD, I. Pla, MD, M. Amorós, MD, M.T. Bonet, MD, M.T. Martin, MD, M.S. Sánchez, MD, J. Altirruba, MD, E. Manzano, MD, A. Altés, MD, A. Medina-Remon, BS, PhD, M. Doménech, MD, PhD, R. Gilabert, MD, PhD, N. Bargalló, MD, PhD, T.M. Freitas-Simoes, RD, MS, I. Roth, RD, E. Ortega, MD, PhD, and A.J. Amor, MD. University Rovira i Virgili, Reus, Spain: M. Bulló, BS, PhD, R. González, RN, C. Molina, RD, G. Mena, RD, F. Márquez, RD, P. Martínez, RD, N. Ibarrola, RD, M. Sorli, RD, J. García Roselló, MD, F. Martin, MD, N. Tort, RN, A. Isach, MD, N. Babio, RD, PhD, A. Salas-Huetos,, MS, PhD, N. Becerra-Tomás, RD, N. Rosique Esteban, RD, J.J. Cabre, MD, G. Mestres, RD, F. Paris, MD, M. Llauradó, MD, R. Pedret, MD, J. Basells, MD, J. Vizcaino, MD, R. Segarra, MD, P. Hernández-Alonso, MS, S. Giardina, MS, C. Ferreira-Pego, RD, C. Papandreou, RD, and L. Camacho, RD. University of Navarra and Osasunbidea, Servicio Navarro de Salud, Primary Care Centres, Pamplona, Spain: P. Buil-Cosiales, MD, M. Ruiz-Canela, PhD, B. Sanjulian, RN, A. Sánchez-Tainta, RD, J. Diez-Espino, MD, C. Razquin, PhD, A. García-Arellano, MD, E. Goñi, RD, Z. Vázquez, RD, N. Berrade, MD, V. Extremera-Urabayen, MD, S. Eguaras, MD, A. Marti, PhD, C. Arroyo-Azpa, MD, L. García-Perez, MD, J. Villanueva Telleria, MD, F. Cortes Ugalde, MD, T. Sagredo Arce, MD, M.D. García de la Noceda Montoy, MD, M.D. Vigata López, MD, M.T. Arceiz Campo, MD, A. Urtasun Samper, MD, M.V. Gueto Rubio, MD, A. Sola, MD, N. Goñi, MD, and O. Lecea, MD. Institute de Recerca Hospital del Mar, Barcelona, Spain: S. Tello, BS, J. Vila, RN, MS, R. de la Torre, DPharm, PhD, D. Muñoz-Aguayo, BS, R. Elosua, MD, PhD, J. Marrugat, MD, PhD, H. Schröder, RD, PhD, N. Molina, BS, E. Maestre, RD, A. Rovira, RD, and M. Farré, MD, PhD. University of Valencia, Valencia, and University Jaume I, Spain: J.V. Sorlí, MD, PhD, P. Carrasco, PhD, C. Ortega-Azorín, PhD, E.M. Asensio, MS, R. Osma, RN, R. Barragán, RN, MS, F. Francés, MD, PhD, M. Guillén, MD,  PhD, J.I. González, DPharm, PhD, C. Saiz, MD, PhD, O. Portolés, PhD, F.J. Giménez, BS, O.Coltell, PhD, P. Guillem-Saiz, PhD, L. Quiles, MD, PhD, V. Pascual, MD, C. Riera, MD, PhD, M.A. Pages, MD, D. Godoy, MD, A. Carratalá-Calvo, PhD, M.J. Martín-Rillo, MD, E. Llopis-Osorio, MD, J. Ruiz-Baixauli, MD, and A. Bertolín-Muñoz, MD. University Hospital of Alava, Vitoria, Spain: I. Salaverria, RD, T. del Hierro, RN, J. Algorta, BS, S. Francisco, RN, A. Alonso-Gómez, MD, PhD, E. Sanz, MD, J. Rekondo, MD, M.C. Belló, MD, PhD, A. Loma-Osorio, MD, PhD. University of Malaga, Malaga, Spain: J. Wärnberg, BS, PhD, R. Benitez Pont, MD, M. Bianchi Alba, MD, R. Gomez-Huelgas, MD, J. Martínez-González, MD, V. Velasco García, MD, J. de Diego Salas, MD, A. Baca Osorio, MD, J. Gil Zarzosa, MD, J.J. Sánchez Luque, MD, and E. Vargas López, MD. Instituto de la Grasa, Consejo Superior de Investigaciones Científicas, Sevilla, Spain: V. Ruiz-Gutiérrez, PhD); J. Sánchez Perona, BS, PhD, E. Montero Romero, MD, and M. García-García, MD, PhD, E. Jurado-Ruiz, BS. Institute of Health Sciences IUNICS, University of Balearic Islands, and Hospital Son Espases, Palma de Mallorca, Spain: M. García-Valdueza, RD, M. Moñino, RD, A. Proenza, BS, PhD, R. Prieto, DPharm, PhD, G. Frontera, MD, M. Ginard, MD, F. Fiol, MD, PhD, A. Jover, MD, D. Romaguera, DPharm, PhD, and J. García, MD. Department of Family Medicine, Distrito Sanitario Atención Primaria Sevilla, Sevilla, Spain: J.M. Santos-Lozano, MD, PhD, M. Ortega-Calvo, MD, PhD, L. Mellado, RN, M. Leal, RN, E. Martínez, RN, F. José García, MD, P. Román, MD, PhD, P. Iglesias, MD, PhD, Y. Corchado, MD, PhD, L. Miró, DPharm, RD, PhD, C. Domínguez, RN, J.M. Lozano, MD, PhD, and E. Mayoral, MD. School of Pharmacy, University of Barcelona, Barcelona, Spain: M.C. López- Sabater, DPharm, PhD, A.I. Castellote-Bargallo, DPharm, PhD, and A. Tresserra-Rimbau, BS, PhD. University of Las Palmas de Gran Canaria, Las Palmas, Spain: J. Álvarez-Pérez, RD, PhD, E.M. Díaz-Benitez, RN, I. Bautista Castaño, MD, PhD, A. Sanchez-Villegas, DPharm, PhD, M.J. Férnandez-Rodríguez, BS, PhD, T. Casañas Quintana, DPharm, PhD, J. Pérez-Cabrera, RD, M. Nissensohn, RD, PhD, V. Díaz-González, MD, C. Ruano-Rodríguez, DPharm, PhD, A.P. Ortiz-Andrelluchi, DPharm, PhD, B. Macias Gutiérrez, MD, and A.J. Santana-Santana, MD, PhD. Hospital Universitario de Bellvitge, Hospitalet de Llobregat, Barcelona, Spain: E. de la Cruz, RD, A. Galera, RD, Y. Soler, RN, F. Trias, MD, I. Sarasa, MD, E. Padres, MD, and E. Corbella, BS. Primary Care Division, Catalan Institute of Health, Barcelona, Spain: C. Cabezas, MD, E. Vinyoles, MD, M.A. Rovira, MD, L. García, MD, G. Flores, MD, J.M. Verdu, MD, P. Baby, MD, A. Ramos, MD, L. Mengual, MD, P. Roura, MD, M.C. Yuste, MD, A. Guarner, MD, A. Rovira, MD, M.I. Santamaria, MD, M. Mata, MD, C. de Juan, MD, and A. Brau, MD. Other investigators of the PREDIMED network: J.A. Tur, DPharm, PhD, M.P. Portillo, DPharm, PhD) G. Saez, MD, PhD. Clinical End Point Committee —M. Aldamiz, MD, A. Alonso, MD, PhD, J. Berjón, MD, L. Forga, MD, J. Gállego, MD, A. Larrauri, MD, J. Portu, MD, PhD, J. Timiraos, MD, and M. Serrano-Martínez, MD, PhD.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Salas-Salvadó has received research funding and is a nonpaid member of the scientific advisory committee of the International Nut Council. Dr Ros has received research funding from the California Walnut Commission (including an unrestricted educational grant) and is a nonpaid member of the Commission’s Scientific Advisory Committee. Dr Estruch reports 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 Fundación Dieta Mediterránea and Cerveceros de España. No other disclosures were reported.

Funding/Support: This study was funded in part by the Spanish Ministry of Science and Innovation through grants AGL2010-22319-C03-02 and AGL2009-13906-C02-02, and by the Instituto de Salud Carlos III, Spain through grants RTIC G03/140, RTIC RD 06/0045, Centro Nacional de Investigaciones Cardiovasculares CNIC 06/2007, ISCIII FIS PS09/01292, Miguel Servet I (CP12/03299, to Sala-Vila), and Juan Rodés (JR14/00008, to Castaner). CIBER Fisiopatología de la Obesidad y Nutrición is an initiative of Instituto de Salud Carlos III, Spain.

Role of the Funder/Sponsor: The funders had no role in design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.