Hostname: page-component-7c8c6479df-94d59 Total loading time: 0 Render date: 2024-03-28T08:27:58.932Z Has data issue: false hasContentIssue false

Hypothesis-oriented food patterns and incidence of hypertension: 6-year follow-up of the SUN (Seguimiento Universidad de Navarra) prospective cohort

Published online by Cambridge University Press:  06 August 2009

Estefanía Toledo
Affiliation:
Department of Preventive Medicine and Public Health, Medical School – Clinica Universidad de Navarra, University of Navarra, c/Irunlarrea, 1 Ed. Investigacion, E-31008 Pamplona(Navarra), Spain Department of Preventive Medicine and Quality Management, Hospital Virgen del Camino, c/Irunlarrea 4, E-31008 Pamplona(Navarra), Spain
Francisco de A Carmona-Torre
Affiliation:
Department of Preventive Medicine and Public Health, Medical School – Clinica Universidad de Navarra, University of Navarra, c/Irunlarrea, 1 Ed. Investigacion, E-31008 Pamplona(Navarra), Spain
Alvaro Alonso
Affiliation:
Department of Preventive Medicine and Public Health, Medical School – Clinica Universidad de Navarra, University of Navarra, c/Irunlarrea, 1 Ed. Investigacion, E-31008 Pamplona(Navarra), Spain Division of Epidemiology and Community Health, School of Public Health, University of Minnesota, Minneapolis, MN 55454, USA
Blanca Puchau
Affiliation:
Department of Nutrition and Food Sciences, Physiology and Toxicology, University of Navarra, c/Irunlarrea, 1. Ed. Investigacíon, E-31008 Pamplona (Navarra), Spain
María A Zulet
Affiliation:
Department of Nutrition and Food Sciences, Physiology and Toxicology, University of Navarra, c/Irunlarrea, 1. Ed. Investigacíon, E-31008 Pamplona (Navarra), Spain
J Alfredo Martinez
Affiliation:
Department of Nutrition and Food Sciences, Physiology and Toxicology, University of Navarra, c/Irunlarrea, 1. Ed. Investigacíon, E-31008 Pamplona (Navarra), Spain
Miguel A Martinez-Gonzalez*
Affiliation:
Department of Preventive Medicine and Public Health, Medical School – Clinica Universidad de Navarra, University of Navarra, c/Irunlarrea, 1 Ed. Investigacion, E-31008 Pamplona(Navarra), Spain
*
*Corresponding author: Email mamartinez@unav.es
Rights & Permissions [Opens in a new window]

Abstract

Objective

To study the association between adherence to several a priori-defined healthy food patterns and the risk of hypertension.

Design

Prospective, multipurpose, dynamic cohort study (recruitment permanently open). We followed up 10 800 men and women (all of them university graduates), who were initially free of hypertension, for a variable period (range 2–6 years, median 4·6 years). During follow-up, 640 participants reported a new medical diagnosis of hypertension. Baseline diet was assessed using a validated 136-item FFQ. Validated information about non-dietary potential confounders was also gathered. We calculated adherence to fifteen different hypothesis-oriented food patterns and assessed the association between each of them and incident hypertension using multivariable Cox models.

Setting

The SUN (Seguimiento Universidad de Navarra – University of Navarra Follow-up) Project, Spain.

Subjects

Participants recruited to the SUN cohort before October 2005 were eligible for inclusion; after excluding those with self-reported hypertension or CVD at baseline, or with extreme total energy intake, data of 10 800 were analysed.

Results

Higher adherence to the DASH (Dietary Approaches to Stop Hypertension) diet (range of the score: 0 to 5) was significantly associated with a lower risk for developing hypertension (P for trend = 0·02). The other food patterns showed no significant association with incident hypertension.

Conclusions

Our results support a long-term protection of the DASH diet against the incidence of hypertension, but we found no evidence of a similar inverse association with hypertension for any other a priori-defined healthy food pattern.

Type
Research Paper
Copyright
Copyright © The Authors 2009

Approximately one billion individuals worldwide are affected by elevated values of blood pressure (BP)(Reference Chobanian, Bakris and Black1). BP is a classical, strong and independent risk factor for CVD: a continuous and consistently progressive positive association with the risk of CVD is observed throughout the range of BP, with no evidence of a threshold. Hypertension is a well-known and modifiable determinant of myocardial infarction, heart failure, stroke and kidney disease.

Preventive strategies for lowering BP, reducing BP-related events and preventing clinical hypertension should be reasonably priced, low-risk and easily implemented. This is one of the reasons why much of the effort to reduce the population burden of hypertension focuses on implementing non-pharmacological approaches. It is well established that lifestyle modifications such as weight loss, increased physical activity, moderation of alcohol consumption, reduction in sodium intake, or a combination of these modalities, decrease BP, enhance antihypertensive drug efficacy and decrease cardiovascular risk(Reference Chobanian, Bakris and Black1). A salient element incorporated into these interventions is dietary advice following the Dietary Approaches to Stop Hypertension (DASH) diet(Reference Svetkey, Simons-Morton, Vollmer, Appel, Conlin, Ryan, Ard and Kennedy2, Reference Appel, Moore and Obarzanek3). The so-called DASH diet (rich in fruits, vegetables, low-fat dairy and whole grains, but low in saturated fat and red meats) has been proved to reduce average levels of BP and to reduce the incidence of hypertension in short-term trials(Reference Svetkey, Simons-Morton, Vollmer, Appel, Conlin, Ryan, Ard and Kennedy2Reference Appel, Brands, Daniels, Karanja, Elmer and Sacks5). However, the epidemiological evidence regarding the long-term effects of a DASH-type diet on the prevention of hypertension is not completely consistent. In fact, no apparent inverse linear trend was found in a large cohort study(Reference Schulze, Hoffmann, Kroke and Boeing6). Another study found that the DASH diet was not more effective in preventing hypertension than was high fruit and vegetable consumption alone(Reference Dauchet, Kesse-Guyot and Czernichow7). Also, others found that general established lifestyle and dietary recommendations were similarly effective in reducing BP as adding the DASH diet to these recommendations(Reference Elmer, Obarzanek and Vollmer8). Moreover, adherence to other healthy food patterns has sometimes been related with reductions in average BP levels or reduced risk of hypertension, but the evidence is even less consistent(Reference McNaughton, Ball, Mishra and Crawford9Reference Nettleton, Schulze, Jiang, Jenny, Burke and Jacobs13). In addition, most large previous epidemiological reports about these associations are based on cross-sectional designs(Reference McNaughton, Ball, Mishra and Crawford9Reference Nettleton, Schulze, Jiang, Jenny, Burke and Jacobs13) and the possibility of reverse causation bias cannot be discarded. In this context, there is no universal consensus about which pattern must be recommended for the long-term prevention of hypertension. There is also a need to ascertain if some of these healthy food patterns may be equally effective in reducing the long-term risk of developing hypertension.

Diet indices or food patterns can be built a priori (as opposed to patterns derived from exploratory factor or cluster analyses) because they are hypothesis-oriented food patterns and reflect known or suspected diet and disease associations(Reference Willett and McCullough14, Reference Jacobs and Steffen15). The approach to build these patterns consists in summarizing the diet by means of a single score that results from a function of different components, such as foods, food groups or a combination of foods and nutrients. These components are selected based on prior knowledge or scientific evidence. This approach has been thus referred to as an ‘a priori approximation’(Reference Schulze, Hoffmann, Kroke and Boeing6, Reference Willett and McCullough14Reference Schulze and Hoffmann16). Some of these indices are based on adherence to existing dietary models, such as the Mediterranean diet(Reference Bach, Serra-Majem, Carrasco, Roman, Ngo, Bertomeu and Obrador17); on adherence to existing Dietary Guidelines(Reference Britten, Marcoe, Yamini and Davis18); or on diversity in dietary intake(Reference Katanoda, Kim and Matsumura19).

The assessment of the association between the original and most commonly used definition for the Mediterranean Diet Score (MDS), developed by Trichopoulou et al.(Reference Trichopoulou, Costacou, Bamia and Trichopoulos20), and the risk of hypertension in the SUN (Seguimiento Universidad de Navarra – University of Navarra Follow-up) cohort has been the topic of a previous report by our group(Reference Núñez-Córdoba, Valencia-Serrano, Toledo, Alonso and Martínez-González21). We found no significant association between adherence to this original MDS and the incidence of hypertension(Reference Núñez-Córdoba, Valencia-Serrano, Toledo, Alonso and Martínez-González21). However, there are several other definitions and operational scores proposed to estimate adherence to the Mediterranean diet(Reference Bach, Serra-Majem, Carrasco, Roman, Ngo, Bertomeu and Obrador17). In addition to the MDS there are several other indices available to assess the compliance with a variety of recommended healthy dietary patterns.

The aim of the present study was to provide evidence to clarify which of the most frequently proposed healthy dietary indices is more effective for the reduction of the long-term incidence of hypertension in the SUN cohort.

Methods

Study population

The SUN project comprises an ongoing, multipurpose, prospective and dynamic cohort of university graduates conducted in Spain. The study protocol was approved by the Institutional Review Board of the University of Navarra.

The study methods have been published in detail elsewhere(Reference Segui-Gomez, de la Fuente, Vazquez, de Irala and Martinez-Gonzalez22). In short, beginning in December 1999, participants – all of them university graduates – have been periodically contacted through mailed questionnaires, which ask for comprehensive baseline characteristics of the study participants. Besides the questionnaire, they receive an invitation letter to participate. Voluntary completion of the first questionnaire is considered as informed consent. The enrolment is permanently open and each year an average of 2000–2500 new participants are newly admitted in the cohort. Follow-up is conducted through mailed questionnaires every 2 years. Non-respondents receive up to five additional mailings requesting their follow-up questionnaire.

Up to July 2008, 20 095 participants were enrolled in the SUN cohort. To warrant a minimum follow-up of 2 years, 15 829 participants recruited before October 2005 were candidates to be eligible for the present analysis because they had spent enough time in the study to be able to complete at least the 2-year follow-up questionnaire. Among them, the retention rate was 88 %. Therefore, we had follow-up information of 13 898 participants. Retention rates at 4- and 6-year follow-up were above 80 %. We excluded 1505 participants due to self-reported baseline prevalent hypertension and 1366 participants with extreme total energy intake (<2092 kJ/d or >14 644 kJ/d in women; <3347 kJ/d or >16 736 kJ/d in men)(Reference Willett23). Finally, 362 participants were excluded due to prevalent CVD at baseline. Thus, the effective sample size for the analyses was 10 800 participants. Among them, 5113 had completed the 6-year follow-up, 2494 the 4-year follow-up but not the 6-year follow-up, and 3193 only the 2-year follow-up.

Exposure assessment

Habitual diet was assessed at baseline with a semi-quantitative 136-item FFQ previously validated in Spain(Reference Martin-Moreno, Boyle, Gorgojo, Maisonneuve, Fernández-Rodríguez, Salvini and Willett24). Each item in the questionnaire included a typical portion size. Daily food consumption was estimated by multiplying the portion size by the consumption frequency for each food item. Nutrient composition of the food items was derived from Spanish food composition tables(Reference Moreiras, Carvajal and Cabrera25, Reference Mataix Verdú and Mañas Almendros26).

We tested a slightly modified definition of the original MDS proposed by Trichopoulou et al.(Reference Trichopoulou, Costacou, Bamia and Trichopoulos20), the Modified Mediterranean Diet Score (MMDS). This score was calculated by developing an a priori score (range: 0 to 9 points) using olive oil instead of the MUFA:SFA ratio originally used in the MDS; we also restricted the negative weighting given to the dairy products group to only whole-fat dairy. A value of 0 or 1 was assigned to each of the nine indicated components with the use of the sex-specific medians as cut-off points. For allegedly beneficial components (vegetables, legumes, fruits, cereals, fish, olive oil), participants whose consumption was below the median were assigned a value of 0, and a value of 1 otherwise. For components presumed to be detrimental (meats and meat products, whole-fat dairy products), participants whose consumption was below the median were assigned a value of 1, and a value of 0 otherwise. We also lowered the upper cut-off points of the original definition of the MDS for alcohol intake and considered only alcohol coming from red wine. A value of 1 was given to men consuming from 5 to <30 g alcohol/d and to women consuming from 2·5 to 15 g alcohol/d exclusively from red wine. Participants were categorized into a low (0–2), intermediate (3–6) or high adherence (7–9) to this MMDS.

Dietary information in our cohort was updated after 2 years of follow-up with brief questions in which participants reported whether they had increased, maintained or decreased the consumption of key food groups. With this available updated information we calculated an Updated Modified Mediterranean Diet Score (UMMDS) as follows. For changes in the consumption of fruits and vegetables, fish, alcohol or olive oil, we summed another point for each item when the participant increased his/her consumption whereas we subtracted a point for each of these items that the participant reported to have reduced his/her consumption. For any decrease in the consumption of dairy products, meats and meat products, butter or sweets we added a further point for each item; increases in the consumption of these items were computed by subtracting a point for each from the baseline score. Accordingly, this updated score (UMMDS) potentially ranged from −8 to +17.

We also looked at the association between other previously published food patterns dealing with the Mediterranean diet and the incidence of hypertension, metabolic syndrome or obesity. Thus, we calculated the Mediterranean Adequacy Index (MAI)(Reference Bach, Serra-Majem, Carrasco, Roman, Ngo, Bertomeu and Obrador17, Reference Alberti-Fidanza and Fidanza27, Reference Alberti, Fruttini and Fidanza28), the Mediterranean Diet Quality Index (MDQI)(Reference Scali, Richard and Gerbeer29), the Mediterranean Food Pattern (MFP) proposed by Sanchez-Villegas et al.(Reference Sánchez-Villegas, Martínez, De Irala and Martínez-González30) and the Mediterranean Score proposed by Panagiotakos et al.(Reference Panagiotakos, Pitsavos and Stefanadis31) (MSP). Further information on how to calculate these indices can be found in the Appendix and the cited references.

In order to cover a wider spectrum, we also considered several dietary patterns that were not based on the Mediterranean diet hypothesis and assessed their association with incident hypertension. Specifically, we computed the Diet Quality Index–International (DQI-I)(Reference Kim, Haines, Siega-Riz and Popkin32); the Recommended Food Score (RFS)(Reference Kant, Schatzkin, Graubard and Schairer33); the Quantitative Index for Dietary Diversity, both in terms of total energy intake (QIDD-k) and in grams of intake (QIDD-g)(Reference Katanoda, Kim and Matsumura19); the Healthy Eating Index (HEI)(Reference Kennedy, Ohls, Carlson and Fleming34); the Alternate Healthy Eating Index (AHEI)(Reference McCullough and Willett35); and the Dietary Guidelines for Americans Adherence Index (DGAI)(Reference Fogli-Cawley, Dwyer, Saltzman, McCullough, Troy and Jacques36). Again, detailed information on how to estimate these scores can be found in the Appendix and the cited references.

The DASH food pattern is based on recommendations originating from the DASH trial(Reference Appel, Moore and Obarzanek3, Reference Appel, Brands, Daniels, Karanja, Elmer and Sacks5). Similarly to the definition of the MDS, we defined a score of adherence to the DASH diet by creating an a priori 6-point score. For the DASH score, a value of 0 or 1 was assigned to each of six indicated components with the use of the results of the DASH trial and the available DASH dietary recommendations (www.dashdiet.org). Thus, daily consumption of ≥5 servings of fruit, ≥4 servings of vegetables, 2–3 servings of low-fat or non-fat dairy products, ≤1/2 serving of sweets and ≥1 serving of whole grains, and consumption of 1–3 servings of lean meat, poultry or fish, were considered as optimal and were scored with 1 point each.

In the baseline questionnaire, the following short questions concerning attitudes towards a healthy diet were included: (i) ‘Do you try to eat much fruit?’; (ii) ‘Do you try to eat many vegetables?’; (iii) ‘Do you try to eat much fish?’; (iv) ‘Do you usually snack between meals?’; (v) ‘Do you try to avoid consuming butter?’; (vi) ‘Do you try to reduce your fat intake?’; (vii) ‘Do you try to reduce your meat consumption?’; (viii) ‘Do you try to reduce your consumption of sweets?’. Another question gathered information about the frequency of eating outside the home. With the answers to these questions, we built up another score: Score of Attitudes Towards a Healthy Diet (ATHD). Attitudes towards increased fruit, vegetable or fish consumption, or reduced butter, fat, meat, snacking or frequency of eating outside the home (<1/week), each contributed 1 point to this score. Consequently, this score (ATHD) ranged from 0 to 9 points.

Ascertainment of incident hypertension

The outcome was defined by the self-report of a medical diagnosis of hypertension in any follow-up questionnaire. Self-reported diagnosed hypertension has been previously validated in a subsample of this cohort(Reference Alonso, Beunza, Delgado-Rodriguez and Martinez-Gonzalez37). Briefly, two physicians, blinded to the information reported by participants in the questionnaires, did direct measurements of BP in the participants’ home and thus confirmed self-reported hypertension or self-reported hypertension-free status in a subsample of the cohort. With the conventional measurement of BP, 82·3 % (95 % CI 72·8, 92·8 %) of those self-reporting a diagnosis of hypertension in the questionnaires were confirmed. Among those who did not report a diagnosis of hypertension in the questionnaires, 85·4 % (95 % CI 72·4, 89·1 %) were confirmed as non-hypertensives(Reference Alonso, Beunza, Delgado-Rodriguez and Martinez-Gonzalez37).

Assessment of other covariates

Age, sex, smoking habit, family history of hypertension, height and weight were collected in the baseline questionnaire. BMI was then calculated as the ratio between weight and the square of height (kg/m2).

Information regarding physical activity was gathered at baseline with a specific questionnaire previously validated in Spain(Reference Martinez-Gonzalez, Lopez-Fontana, Varo, Sánchez-Villegas and Martinez38) which assessed the time spent in seventeen different activities. Each of these activities was assigned a multiple of the resting metabolic rate (MET score). For this purpose, we used information on average intensity of each activity from previously published guidelines(Reference Ainsworth, Haskell and Whitt39).

Statistical analysis

Participants were divided into categories according to previous categorizations of these scores. In the cases in which evidence was not available, participants were divided taking into account sample sizes of each category.

Food and nutrient adjustment for total energy intake was performed with the residual method(Reference Willett23).

We fitted Cox regression models to assess the relative incidence of hypertension across increasing categories of the a priori-defined scores of adherence to healthy food patterns. When addressing the association between the UMMDS and the outcome, we used as exposure the updated diet after 2-year follow-up and we used as outcome only the incidence of hypertension after 4-year or 6-year follow-up (i.e. we excluded subjects who had only 2-year follow-up). In all analyses, we fitted a first Cox regression model adjusted only for age and sex. In a second model we additionally adjusted for BMI (kg/m2), family history of hypertension, total energy intake, smoking (in three categories: never, past and current smokers) and physical activity. For the linear trend tests, we treated the exposures (scores) as continuous variables.

All P values are two-tailed and statistical significance was set at P < 0·05. Analyses were performed with the SPSS statistical software package version 15·0 (SPSS Inc., Chicago, IL, USA).

Results

Median follow-up in this cohort was 4·6 years. During 50 304 person-years of follow-up, 640 cases of incident hypertension were observed.

Baseline characteristics of the study participants are presented in Table 1. Subjects with a higher adherence to the DASH diet were more likely to be female, older, more physically active and hypercholesterolaemic, and less likely to be current smokers. Family history of hypertension was more frequent among them. They also had a lower consumption of alcohol, a lower total fat intake and higher intakes of total energy, potassium, carbohydrate, vegetable protein and fibre. On the other hand, participants with a higher adherence to the MMDS were more likely male, older, hypercholesterolaemic and physically active. Family history of hypertension was more common among them and they were less likely to be current smokers. These subjects presented higher intakes of total energy, sodium, carbohydrate, vegetable protein, fibre and MUFA:SFA ratio and a lower total fat intake.

Table 1 Baseline characteristicsFootnote * of the SUN study population according to adherence to food patterns (participants recruited during 1999–2005)

SUN, Seguimiento Universidad de Navarra (University of Navarra Follow-up); DASH, Dietary Approaches to Stop Hypertension; MMDS, Modified Mediterranean Diet Score; MET, metabolic equivalent.

* Mean and standard deviation unless otherwise stated.

Based on the recommendations originating from the DASH trial, we defined a score of adherence to the DASH diet by creating an a priori 6-point score. A value of 0 or 1 was assigned to each of six indicated components with the use of the results of the DASH trial and the available DASH dietary recommendations (www.dashdiet.org). Thus, a daily consumption of ≥5 servings of fruit, ≥4 servings of vegetables, 2–3 servings of low-fat or non-fat dairy products, ≤1/2 serving of sweets and ≥1 serving of whole grains, and consumption of 1–3 servings of lean meat, poultry or fish, were considered as optimal and were scored with 1 point each.

The MMDS was calculated by assigning a value of 0 or 1 to each of the nine indicated components with the use of the sex-specific medians as cut-off points. For allegedly beneficial components (vegetables, legumes, fruits, cereals, fish, olive oil), participants whose consumption was below the median were assigned a value of 0, and a value of 1 otherwise. For components presumed to be detrimental (meat and meat products, whole-fat dairy products), participants whose consumption was below the median were assigned a value of 1, and a value of 0 otherwise. For alcohol, a value of 1 was given to men consuming from 5 to <30 g/d and to women consuming from 2·5 to 15 g/d exclusively from red wine.

Hazard ratios for the incidence of hypertension according to adherence to the different patterns are shown in Table 2. A higher adherence to the DASH diet was significantly associated with a lower risk for developing hypertension in the multivariable-adjusted model. Specifically, there was a significant inverse linear trend for this association. When we additionally adjusted for alcohol consumption, the results did not change materially (Table 2). Regarding the AHEI, the comparison between extreme quintiles showed an increased risk of hypertension among those subjects with a higher adherence to this pattern. Nevertheless, there was no significant linear trend for this association. Unexpectedly, hazard ratios relating adherence to the UMMDS with the risk of hypertension showed a significant direct association (multivariable-adjusted hazard ratio = 1·34, 95 %CI 1·04, 1·73, P for trend = 0·002). However, none of the other healthy food patterns, including five other indices, assessing adherence to the Mediterranean diet (MMDS, MAI, MDQI, MFP and MSP) showed any significant association with the incidence of hypertension.

Table 2 Hazard ratios (HR) and 95 % confidence intervals of hypertension according to adherence to a priori-defined food patterns, the SUN Study, 1999–2008

SUN, Seguimiento Universidad de Navarra (University of Navarra Follow-up) Study; DASH, Dietary Approaches to Stop Hypertension; DQI-I, Diet Quality Index–International; RFS, Recommended Food Score; QIDD, Quantitative Index for Dietary Diversity (in terms of total energy intake (QIDD-k) and in grams of intake (QIDD-g)); HEI,: Healthy Eating Index; AHEI, Alternate Healthy Eating Index; DGAI, Dietary Guidelines for Americans Index; MMDS, Modified Mediterranean Diet Score; UMMDS, Updated Modified Mediterranean Diet Score; MAI, Mediterranean Adequacy Index; MDQI, Mediterranean Diet Quality Index; MFP, Mediterranean Food Pattern (Sanchez-Villegas et al.); MSP, Mediterranean score (Panagiotakos et al.); ATHD, Attitudes Towards a Healthy Diet.

*Reference category.

†Adjusted for age, sex, BMI, family history of hypertension, total energy intake, physical activity, smoking and hypercholesterolaemia.

‡Additionally adjusted for alcohol intake.

Discussion

These data from the SUN cohort with more than 50 000 person-years of follow-up showed that higher adherence to a DASH-type diet was associated with a reduction in the risk of hypertension in the long term. Although an updated score for the Mediterranean diet including only the subset of the cohort who completed 4-year or 6-year follow-up was unexpectedly associated with a modestly increased risk of hypertension, all other indices built to appraise adherence to the Mediterranean food pattern (MMDS, MAI, MDQI, MFP and MSP) which included all participants did not show any apparent association with the incidence of hypertension.

All assessed food patterns share some characteristics such as encouraging the consumption of high amounts of fruits and vegetables. However, they try to gather some diverse aspects of diet and thus they can be separated into two main groups: (i) those that aim to capture the healthy aspects of the Mediterranean diet (MMDS, UMMDS, MAI, MDQI, MFP and MSP); and (ii) those trying to merge existing evidence and recommendations about promoting healthy and avoiding deleterious foods and nutrients (DASH diet, DQI-I, RFS, QIDD, HEI, AHEI and DGAI).

It has long been postulated that the Mediterranean diet may be protective against CVD(Reference Keys40, Reference Kromhout, Keys and Aravanis41). In fact, several large cohorts have found that higher adherence to the Mediterranean diet was associated with a significant reduction in total and cardiovascular mortality(Reference Trichopoulou, Costacou, Bamia and Trichopoulos20, Reference Martínez-González and Sánchez-Villegas42Reference Sofi, Cesari, Abbate, Gensini and Casini44). However, the inconsistency of these previous results with our findings regarding hypertension can be explained because other pathways can constitute alternative explanations of the cardioprotective effect of classical Mediterranean diets, such as those related to inflammatory status, cardiac rhythm thrombotic mechanisms, lipid levels, insulin sensitivity or endothelial function. Our results are not consistent with a previous report by Psaltoupoulou et al. where an index that tried to capture the nature of the traditional Mediterranean diet – the original MDS – was found to be inversely associated with average systolic and diastolic BP(Reference Psaltopoulou, Naska, Orfanos, Trichopoulos, Mountokalakis and Trichopoulou45). The cross-sectional design of the study by Psaltoupoulou et al.(Reference Psaltopoulou, Naska, Orfanos, Trichopoulos, Mountokalakis and Trichopoulou45) together with the fact that they assessed BP average levels instead of the risk of hypertension does not allow a direct and proper comparison with our findings. On the other hand, in a previous report by another group of researchers, higher adherence to the Mediterranean diet (assessed using the MAI) was shown to be cross-sectionally associated with higher average systolic BP levels among older women(Reference Di Giuseppe, Bonanni, Olivieri, Di Castelnuovo, Donati, de Gaetano, Cerlettu and Iacoviello46). Similarly to our results regarding the UMMDS, this unexpected cross-sectional finding does not support that any protection against hypertension can be expected from a higher adherence to the Mediterranean diet(Reference Núñez-Córdoba, Valencia-Serrano, Toledo, Alonso and Martínez-González21). Our interpretation of the results regarding the Mediterranean diet and hypertension is that we found no evidence to support the hypothesis that a Mediterranean-type diet may reduce the long-term risk of hypertension, because the association was essentially null for all other indices of Mediterranean diet adherence that we tested. It is also possible that unmeasured or uncontrolled residual confounding may explain the unexpected positive association between UMMDS and hypertension. In fact, it is likely that small increases in BP, some slight weight gain or the diagnosis of some incident minor disease may have prompted decisions of participants to change their dietary habits or, because of these reasons, they may have received medical advice to improve their adherence to a Mediterranean-type diet.

The RFS has been previously associated with lower risk of CVD in women(Reference McCullough, Feskanich, Stampfer, Giovannucci, Rimm, Hu, Spiegelman, Hunter, Colditz and Willett47). While the HEI has been associated with lower risk of CVD only in women(Reference McCullough, Feskanich, Stampfer, Rosner, Hu, Hunter, Variyam, Colditz and Willett48, Reference McCullough, Feskanich, Rimm, Giovannucci, Ascherio, Variyam, Spiegelman, Stampfer and Willett49), its variant – the AHEI – has been associated with lower risks of CVD in both women and men(Reference McCullough and Willett35).

Adherence to a DASH-type diet has been the only dietary pattern shown to be inversely associated with the long-term incidence of hypertension in a large prospective cohort, the Iowa Women’s Health Study, including 20 993 women(Reference Folsom, Parker and Harnack50). Not surprisingly, we also found a protective association also for this pattern against the risk of hypertension. However, in the Iowa cohort, the inverse association was apparent only in the model adjusted for age and total energy intake; after adjustment for other potential confounders, there was little evidence that the long-term incidence of hypertension was independently related to the baseline DASH diet(Reference Folsom, Parker and Harnack50). Our findings are also in agreement with the results reported by two other smaller cohorts. The first study, a German cohort of the EPIC (European Prospective Investigation into Cancer and Nutrition) project, including 8552 women followed for 2–4 years, found that participants in the third quartile of a DASH diet had significantly less hypertension incidence than those in the first quartile(Reference Schulze, Hoffmann, Kroke and Boeing6). The other cohort study was conducted in France (SU.VI.MAX; SUpplementation en VItamines et Minéraux AntioXydants study) and included 2341 men and women followed-up for 5·4 years. They reported that the DASH pattern was inversely associated with changes in average BP, but no assessment was reported about the incidence of hypertension(Reference Dauchet, Kesse-Guyot and Czernichow7).

We are aware that our study has some limitations. First, we relied on self-reported information in the ascertainment of exposure and outcome. However, previous validation studies have shown adequate quality of this information. The FFQ that we used has been previously validated in Spain(Reference Martin-Moreno, Boyle, Gorgojo, Maisonneuve, Fernández-Rodríguez, Salvini and Willett24) and the self-report of hypertension had been previously validated in a subsample of the SUN cohort(Reference Alonso, Beunza, Delgado-Rodriguez and Martinez-Gonzalez37). The results of the validation study suggest that self-reported hypertension can be considered a valid tool for assessing a medical diagnosis of hypertension in this highly educated cohort. Second, our sample is not representative of the general population since it is a young cohort formed entirely of university graduates. However, there is no biological reason to think that our results might not be generalizable to other population groups and this is the major support for the external validity of our findings(Reference Rothman, Greenland and Lash51). Third, as in all observational studies, residual confounding might be an alternative potential explanation of the results found. Nevertheless, we were able to adjust for the main known risk factors for hypertension and for this reason we do not consider residual confounding as a likely important cause of the observed results. Fourth, non-differential measurement error in nutritional variables, inherent to the methodology in nutritional studies, might have occurred and we acknowledge that it may represent a difficulty for identifying associations of very low magnitude between healthy dietary patterns and the risk of hypertension. Fifth, since we have tested several dietary patterns it could be argued that multiple testing might play a role in our findings. Certainly, this issue could explain the presence of significant results it that were the case; however, it is not likely to be a major problem in our study where we found mainly non-significant results. Besides this, we have applied previously defined patterns with a clear rationale for their development. Thus, taking into account the consistency with previous studies(Reference Appel, Moore and Obarzanek3, Reference Schulze, Hoffmann, Kroke and Boeing6, Reference Folsom, Parker and Harnack50) and substantial mechanistic reasons, the significant inverse linear trend found for the DASH diet is more likely to be supported by biological plausibility than to be explained just because of multiple testing.

Our findings do not support recommending the Mediterranean diet for the prevention of hypertension, but provide evidence in favour of the long-term effectiveness of the DASH diet.

Acknowledgements

Sources of funding: The SUN Project is funded by the Spanish Government (Instituto de Salud Carlos III, Fondo de Investigaciones Sanitarias projects PI070240, PI081943 and RD 06/0045). Conflict of interest: None of the authors had any conflicts of interest in connection with this study. Authors’ contributions: E.T. and M.A.M.-G. were the main authors responsible for the study design, the statistical analysis and writing the manuscript. F.A.C.-T., A.A., B.P., M.A.Z. and J.A.M. contributed to the interpretation and discussion of the results. M.A.M.-G. obtained funding, is the main researcher in the SUN cohort, and revised the manuscript providing expert advice. E.T., F.A.C.-T., A.A., B.P., M.A.Z., J.A.M. and M.A.M.-G. declare that they participated sufficiently in the work to take full and public responsibility for its content. Acknowledgements: We are indebted to the participants of the SUN study for their continued cooperation and participation. We also thank other members of the SUN study group including: J.M. Nuñez-Cordoba, C. de la Fuente, Z. Vazquez, S. Benito, J. de Irala, M. Segui-Gomez, A. Marti, F. Guillen-Grima and M. Serrano-Martinez, University of Navarra; M. Delgado-Rodriguez, University of Jaen; J. Llorca, University of Cantabria; and A. Sanchez-Villegas, University of Las Palmas. We thank the members of the Department of Nutrition of the Harvard School of Public Health (A. Ascherio, F.B. Hu and W.C. Willett) who helped us to design the SUN study.

Appendix

Calculation of the dietary indices

References

1.Chobanian, AV, Bakris, GL, Black, HR et al. (2003) The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA 289, 25602572.CrossRefGoogle ScholarPubMed
2.Svetkey, LP, Simons-Morton, D, Vollmer, WM, Appel, LJ, Conlin, PR, Ryan, DH, Ard, J & Kennedy, BMfor the DASH Research Group (1999) Effects of food patterns on blood pressure: subgroup analysis of the Dietary Approaches to Stop Hypertension (DASH) randomized clinical trial. Arch Intern Med 159, 285293.CrossRefGoogle ScholarPubMed
3.Appel, LJ, Moore, TJ, Obarzanek, E et al. (1997) A clinical trial of the effects of food patterns on blood pressure. DASH Collaborative Research Group. N Engl J Med 336, 11171124.CrossRefGoogle ScholarPubMed
4.Shah, M, Adams-Huet, B & Garg, A (2007) Effect of high-carbohydrate or high-cis-monounsaturated fat diets on blood pressure: a meta-analysis of intervention trials. Am J Clin Nutr 85, 12511256.CrossRefGoogle ScholarPubMed
5.Appel, LJ, Brands, MW, Daniels, SR, Karanja, N, Elmer, PJ, Sacks, FM & American Heart Association (2006) Dietary approaches to prevent and treat hypertension: a scientific statement from the American Heart Association. Hypertension 47, 296308.CrossRefGoogle ScholarPubMed
6.Schulze, MB, Hoffmann, K, Kroke, A & Boeing, H (2003) Risk of hypertension among women in the EPIC-Potsdam Study: comparison of relative risk estimates for exploratory and hypothesis-oriented dietary patterns. Am J Epidemiol 158, 365373.CrossRefGoogle ScholarPubMed
7.Dauchet, L, Kesse-Guyot, E, Czernichow, S et al. (2007) Dietary patterns and blood pressure change over 5-y follow-up in the SU.VI.MAX cohort. Am J Clin Nutr 85, 16501656.CrossRefGoogle ScholarPubMed
8.Elmer, PJ, Obarzanek, E, Vollmer, WM et al. .; PREMIER Collaborative Research Group (2006) Effects of comprehensive lifestyle modification on diet, weight, physical fitness, and blood pressure control: 18-month results of a randomized trial. Ann Intern Med 144, 485495.CrossRefGoogle ScholarPubMed
9.McNaughton, SA, Ball, K, Mishra, GD & Crawford, DA (2008) Dietary patterns of adolescents and risk of obesity and hypertension. J Nutr 138, 364370.CrossRefGoogle ScholarPubMed
10.van Dam, RM, Grievink, L, Ocké, MC & Feskens, EJ (2003) Patterns of food consumption and risk factors for cardiovascular disease in the general Dutch population. Am J Clin Nutr 77, 11561163.CrossRefGoogle ScholarPubMed
11.Fogli-Cawley, JJ, Dwyer, JT, Saltzman, E, McCullough, ML, Troy, LM, Meigs, JB & Jacques, PF (2007) The 2005 Dietary Guidelines for Americans and risk of the metabolic syndrome. Am J Clin Nutr 86, 11931201.CrossRefGoogle ScholarPubMed
12.Kant, AK & Graubard, BI (2005) A comparison of three dietary pattern indexes for predicting biomarkers of diet and disease. J Am Coll Nutr 24, 294303.CrossRefGoogle ScholarPubMed
13.Nettleton, JA, Schulze, MB, Jiang, R, Jenny, NS, Burke, GL & Jacobs, DR Jr (2008) A priori-defined dietary patterns and markers of cardiovascular disease risk in the Multi-Ethnic Study of Atherosclerosis (MESA). Am J Clin Nutr 88, 185194.CrossRefGoogle ScholarPubMed
14.Willett, WC & McCullough, ML (2008) Food pattern analysis for the evaluation of dietary guidelines. Asia Pac J Clin Nutr 17, Suppl. 1, 7578.Google ScholarPubMed
15.Jacobs, DR Jr & Steffen, LM (2003) Nutrients, foods, and dietary patterns as exposures in research: a framework for food synergy. Am J Clin Nutr 78, 3 Suppl., 508S513S.CrossRefGoogle ScholarPubMed
16.Schulze, MB & Hoffmann, K (2006) Methodological approaches to study dietary patterns in relation to risk of coronary heart disease and stroke. Br J Nutr 95, 860869.CrossRefGoogle ScholarPubMed
17.Bach, A, Serra-Majem, L, Carrasco, JL, Roman, B, Ngo, J, Bertomeu, I & Obrador, B (2006) The use of indexes evaluating the adherence to the Mediterranean diet in epidemiological studies: a review. Public Health Nutr 9, 132146.CrossRefGoogle Scholar
18.Britten, P, Marcoe, K, Yamini, S & Davis, C (2006) Development of food intake patterns for the MyPyramid food guidance system. J Nutr Educ Behav 38, 6 Suppl., S78S92.CrossRefGoogle ScholarPubMed
19.Katanoda, K, Kim, HS & Matsumura, Y (2006) New Quantitative Index for Dietary Diversity (QUANTIDD) and its annual changes in the Japanese. Nutrition 22, 283287.CrossRefGoogle ScholarPubMed
20.Trichopoulou, A, Costacou, T, Bamia, C & Trichopoulos, D (2003) Adherence to a Mediterranean diet and survival in a Greek population. N Engl J Med 348, 25992608.CrossRefGoogle Scholar
21.Núñez-Córdoba, JM, Valencia-Serrano, F, Toledo, E, Alonso, A & Martínez-González, MA (2009) Mediterranean diet and incidence of hypertension: the Seguimiento Universidad de Navarra (SUN) Study. Am J Epidemiol 169, 339346.CrossRefGoogle Scholar
22.Segui-Gomez, M, de la Fuente, C, Vazquez, Z, de Irala, J & Martinez-Gonzalez, MA (2006) Cohort profile: the ‘Seguimiento Universidad de Navarra’ (SUN) study. Int J Epidemiol 35, 14171422.CrossRefGoogle Scholar
23.Willett, WC (1998) Nutritional Epidemiology, 2nd ed. New York: Oxford University Press.CrossRefGoogle Scholar
24.Martin-Moreno, JM, Boyle, P, Gorgojo, L, Maisonneuve, P, Fernández-Rodríguez, JC, Salvini, S & Willett, WC (1993) Development and validation of a food frequency questionnaire in Spain. Int J Epidemiol 22, 512519.CrossRefGoogle ScholarPubMed
25.Moreiras, O, Carvajal, A & Cabrera, L (2005) Tablas de composición de alimentos (Food Composition Tables), 9th ed. Madrid: Pirámide.Google Scholar
26.Mataix Verdú, J & Mañas Almendros, M (2003) Tablas de composición de alimentos (Food Composition Tables), 4th ed. Granada: Universidad de Granada.Google Scholar
27.Alberti-Fidanza, A & Fidanza, F (2004) Mediterranean Adequacy Index of Italian diets. Public Health Nutr 7, 937941.CrossRefGoogle ScholarPubMed
28.Alberti, A, Fruttini, D & Fidanza, F (2009) The Mediterranean Adequacy Index: further confirming results of validity. Nutr Metab Cardiovas Dis 19, 6166.CrossRefGoogle ScholarPubMed
29.Scali, J, Richard, A & Gerbeer, M (2001) Diet profiles in a population sample from Mediterranean southern France. Public Health Nutr 4, 173182.CrossRefGoogle Scholar
30.Sánchez-Villegas, A, Martínez, JA, De Irala, I & Martínez-González, MA (2002) Determinants of the adherence to an ‘a priori’ defined Mediterranean food pattern. Eur J Nutr 41, 249257.CrossRefGoogle Scholar
31.Panagiotakos, DB, Pitsavos, C & Stefanadis, C (2006) Food patterns: a Mediterranean diet score and its relation to clinical and biological markers of cardiovascular disease risk. Nutr Metab Cardiovasc Dis 16, 559568.CrossRefGoogle ScholarPubMed
32.Kim, S, Haines, P, Siega-Riz, AM & Popkin, BM (2003) The Diet Quality Index–International (DQI-I) provides an effective tool for cross-national comparison of diet quality as illustrated by China and the United States. J Nutr 133, 34763484.CrossRefGoogle ScholarPubMed
33.Kant, AK, Schatzkin, A, Graubard, BI & Schairer, C (2000) A prospective study of diet quality and mortality in women. JAMA 283, 21092115.CrossRefGoogle ScholarPubMed
34.Kennedy, ET, Ohls, J, Carlson, S & Fleming, K (1995) The Healthy Eating Index: design and applications. J Am Diet Assoc 95, 11031108.CrossRefGoogle ScholarPubMed
35.McCullough, ML & Willett, WC (2006) Evaluating adherence to recommended diets in adults: the Alternate Healthy Eating Index. Public Health Nutr 9, 152157.CrossRefGoogle ScholarPubMed
36.Fogli-Cawley, JJ, Dwyer, JT, Saltzman, E, McCullough, ML, Troy, LM & Jacques, PF (2006) The 2005 Dietary Guidelines for Americans Adherence Index: development and application. J Nutr 136, 29082915.CrossRefGoogle ScholarPubMed
37.Alonso, A, Beunza, JJ, Delgado-Rodriguez, M & Martinez-Gonzalez, MA (2005) Validation of self reported diagnosis of hypertension in a cohort of university graduates in Spain. BMC Public Health 5, 94.CrossRefGoogle Scholar
38.Martinez-Gonzalez, MA, Lopez-Fontana, C, Varo, JJ, Sánchez-Villegas, A & Martinez, JA (2005) Validation of the Spanish version of the physical activity questionnaire used in the Nurses’ Health Study and the Health Professionals’ Follow-up Study. Public Health Nutr 8, 920927.CrossRefGoogle ScholarPubMed
39.Ainsworth, BE, Haskell, WL, Whitt, MC et al. (2000) Compendium of physical activities: an update of activity codes and MET intensities. Med Sci Sports Exerc 32, 9 Suppl., S498S504.CrossRefGoogle ScholarPubMed
40.Keys, A (1980) Seven Countries: A Multivariate Analysis of Death and Coronary Heart Disease. Cambridge, MA: Harvard University Press.CrossRefGoogle Scholar
41.Kromhout, D, Keys, A, Aravanis, C et al. (1989) Food consumption patterns in the 1960s in seven countries. Am J Clin Nutr 49, 889894.CrossRefGoogle ScholarPubMed
42.Martínez-González, MA & Sánchez-Villegas, A (2004) The emerging role of Mediterranean diets in cardiovascular epidemiology: monounsaturated fats, olive oil, red wine or the whole pattern? Eur J Epidemiol 19, 913.CrossRefGoogle ScholarPubMed
43.Mitrou, PN, Kipnis, V, Thiébaut, AC et al. (2007) Mediterranean dietary pattern and prediction of all-cause mortality in a US population: results from the NIH-AARP Diet and Health Study. Arch Intern Med 167, 24612468.CrossRefGoogle Scholar
44.Sofi, F, Cesari, F, Abbate, R, Gensini, GF & Casini, A (2008) Adherence to Mediterranean diet and health status: meta-analysis. BMJ 337, a1344.CrossRefGoogle ScholarPubMed
45.Psaltopoulou, T, Naska, A, Orfanos, P, Trichopoulos, D, Mountokalakis, T & Trichopoulou, A (2004) Olive oil, the Mediterranean diet, and arterial blood pressure: the Greek European Prospective Investigation into Cancer and Nutrition (EPIC) study. Am J Clin Nutr 80, 10121018.CrossRefGoogle ScholarPubMed
46.Di Giuseppe, R, Bonanni, A, Olivieri, M, Di Castelnuovo, A, Donati, MB, de Gaetano, G, Cerlettu, C & Iacoviello, L (2008) Adherence to Mediterranean diet and anthropometric and metabolic parameters in an observational study in the ‘Alto Molise’ region: the MOLI-SAL project. Nutr Metab Cardiovasc Dis 18, 415421.CrossRefGoogle Scholar
47.McCullough, ML, Feskanich, D, Stampfer, MJ, Giovannucci, EL, Rimm, EB, Hu, FB, Spiegelman, D, Hunter, DJ, Colditz, GA & Willett, WC (2002) Diet quality and major chronic disease risk in men and women: moving toward improved dietary guidance. Am J Clin Nutr 76, 12611271.CrossRefGoogle ScholarPubMed
48.McCullough, ML, Feskanich, D, Stampfer, MJ, Rosner, BA, Hu, FB, Hunter, DJ, Variyam, JN, Colditz, GA & Willett, WC (2000) Adherence to the Dietary Guidelines for Americans and risk of major chronic disease in women. Am J Clin Nutr 72, 12141222.CrossRefGoogle Scholar
49.McCullough, ML, Feskanich, D, Rimm, EB, Giovannucci, EL, Ascherio, A, Variyam, JN, Spiegelman, D, Stampfer, MJ & Willett, WC (2000) Adherence to the Dietary Guidelines for Americans and risk of major chronic disease in men. Am J Clin Nutr 72, 12231231.CrossRefGoogle Scholar
50.Folsom, AR, Parker, ED & Harnack, LJ (2007) Degree of concordance with DASH diet guidelines and incidence of hypertension and fatal cardiovascular disease. Am J Hypertens 20, 225232.CrossRefGoogle ScholarPubMed
51.Rothman, KJ, Greenland, S & Lash, TL (2008) Modern Epidemiology, 3rd ed. Philadelphia, PA: Lippincott Williams & Wilkins.Google Scholar
Figure 0

Table 1 Baseline characteristics* of the SUN study population according to adherence to food patterns (participants recruited during 1999–2005)

Figure 1

Table 2 Hazard ratios (HR) and 95 % confidence intervals of hypertension according to adherence to a priori-defined food patterns, the SUN Study, 1999–2008