Mediterranean Dietary Patterns and Impaired Physical Function in Older Adults

Mediterranean Dietary Patterns and Impaired Physical Function in Older Adults Abstract Background Information about nutritional risk factors of functional limitation is scarce. The aim of this study was to examine the association between the Mediterranean diet and risk of physical function impairment in older adults. Methods We used data from 1,630 participants in the Seniors-ENRICA cohort aged ≥60 years. In 2008–2010, adherence to the Mediterranean diet pattern was measured with the Mediterranean Diet Score (MDS) and the Mediterranean Diet Adherence Screener (MEDAS). Study participants were followed up through 2012 to assess incident impairment in agility and mobility as well as impairment in overall physical functioning, defined as a ≥5-point decrease from baseline to follow-up in the physical component summary of the 12-Item Short-Form Health Survey. Results Over a median follow-up of 3.5 years, we identified 343 individuals with agility limitation, 212 with mobility limitation, and 457 with decreased overall physical functioning. No association was found between the MDS score and the likelihood of impaired agility or mobility, although a 2-point increment in the MDS score was marginally associated with lower likelihood for decreased overall physical function. Compared to individuals in the lowest tertile of the MEDAS score, those in the highest tertile showed a lower odds of agility limitation (odds ratio: 0.67, 95% confidence interval: 0.48; 0.94, p trend = .02), mobility limitation (odds ratio: 0.69, 95% confidence interval: 0.40; 0.88, p trend = .01), and decreased overall physical functioning (odds ratio: 0.60, 95% confidence interval: 0.45; 0.79, p trend < .001). Conclusions In this prospective cohort study, a Mediterranean-style dietary pattern, especially when measured with the MEDAS, was associated with a lower likelihood of physical function impairment in older adults. MEDAS, MDS, Mobility, Agility Life expectancy in Spain is among the highest in Europe and has increased from 69.0 to 78.4 years in men and from 74.6 to 84.2 years in women in the period between 1970 and 2010 (1). Nevertheless, persons who live longer are not necessarily in good health (2,3). In fact, many older adults suffer from multimorbidity that, in combination with the ageing-related decline in many biological systems, may result in impairments of physical functioning (4,5). In turn, these functional limitations can lead to disability, defined as a difficulty or dependency in carrying out activities essential to independent living (6). Moreover, functional limitations have also been associated with increased risk of institutionalization and death (7,8). Several modifiable behavioral factors have been associated with limitations in physical functioning in older adults, including obesity and low physical activity (9,10). In addition, previous research suggests that certain nutrients (antioxidants and B vitamins) and food products (fruit, vegetables, and dairy) play a role in maintaining physical functioning in old age (10–14). One limitation of investigating the effect of individual dietary factors is that nutrients or food products tend to correlate or even interact with each other (15). Also, the effects of individual dietary factors may be too small to be detected. Thus, the use of dietary patterns could overcome these limitations (15). So far, only a few studies have investigated the association of dietary patterns with impaired physical function (16–21); their results suggested a protective effect of the Mediterranean diet (17–21). However, the majority of these studies were limited by their cross-sectional design (17–20), and the previous longitudinal studies performed focused on only one domain of physical performance (21,22). Therefore, the objective of this study was to examine the prospective association between a Mediterranean-style diet and the impairment of physical function, specifically of agility, mobility, and overall physical functioning, in a large population of older adults from Spain. Methods Study Design and Participants Data were taken from the Seniors-ENRICA cohort, whose methods have been reported elsewhere (23,24). In brief, the cohort was derived from the ENRICA study, a survey conducted in 2008–2010 among 12,948 individuals representative of the noninstitutionalized adult population of Spain. The study participants aged 60 years or older (n = 2,614) from the ENRICA study comprised the Seniors-ENRICA cohort. At baseline, information on sociodemographic variables, lifestyle, health status, and morbidity was collected through a phone interview. In two subsequent home visits, appropriately trained research staff collected dietary information, conducted a physical exam, and obtained blood and urine samples. Participants were followed up through 2012, when a second wave of data collection was performed to update information from baseline. Ninety-five participants (3.6%) died during follow-up. From the 2,519 Seniors-ENRICA cohort participants alive, we excluded 9 participants with baseline medical diagnosis of dementia or Alzheimer’s disease and 441 participants without information about their cognitive function. We also excluded 8 individuals without diet information or an implausibly high or low energy intake (outside the range of 800–5,000 kcal/d for men and 500–4,000 kcal/d for women) and 160 participants who lacked data on mobility (n = 139), agility (n = 2), or overall physical function score (n = 19). Individuals excluded were less educated and reported a higher prevalence of diagnosed diabetes, osteomuscular diseases, cardiovascular diseases, and cancer than individuals without missing values. Additionally, we excluded participants with baseline limitations in physical functioning: 186 with impaired mobility and 85 with impaired agility. Thus, the analyses were conducted on 1,630 individuals. Study participants gave written informed consent. The study was approved by the Clinical Research Ethics Committee of the La Paz University Hospital in Madrid. Study Variables Diet At baseline, information on food consumption was obtained through a validated computer-assisted face-to-face diet history, which was developed from that used in the EPIC cohort study in Spain (25). Additional information about validation of this diet history has been provided in the online Supplementary Material. Two scores were used to measure accordance with a Mediterranean-style diet, the Mediterranean Diet Score (MDS) and the Mediterranean Diet Adherence Screener (MEDAS) (26–28). The MDS was developed by Trichopoulou and colleagues and has been widely used in Mediterranean as well as non-Mediterranean countries (26–28). The MDS is based on nine different items (vegetables, fruit and nuts, legumes, grains, fish and seafood, the ratio of unsaturated fatty acids to saturated fatty acids, meat and poultry, dairy, and alcohol) (27). A value of 1 was assigned to moderate alcohol intake (5–25 g/d for women and 10–50 g/d for men) and to an intake above the sex-specific median for all other items, except for meat and poultry and for dairy. For these food groups, considered detrimental for health, 1 point was assigned to an intake below the sex-specific median. Scores for all nine items were summed, resulting in a range from 0 to 9, whereby a higher score reflects better accordance with the Mediterranean diet. Due to the use of sample medians, this score is highly dependent on the sample characteristics. The MEDAS score was developed to assess compliance with the dietary intervention of the PREDIMED trial and uses defined cutoff values for its 14 components (29). The MEDAS score includes two questions on food intake habits and 12 questions on food consumption. One point was given for each of the following components: using olive oil as the principal source of fat for cooking; preferring white meat over red meat; ≥4 tablespoons of olive oil/d; ≥2 servings of vegetables/d; ≥3 pieces of fruit/d; <1 serving of red meat, a hamburger or sausage/d; <1 serving of butter, margarine, or cream/d; <1 time of sugar-sweetened or carbonated beverages/d; ≥1 servings of red wine/d; ≥3 servings of legumes/wk; ≥3 servings of fish or seafood/wk; <2 commercial baked goods/wk; ≥3 servings of nuts/wk; and ≥2 servings/wk of a dish with a traditional sauce of tomatoes, garlic, onion, or leeks sautéed in olive oil (sofrito). The MEDAS score ranges from 0 to 14, with a higher score indicating better accordance with the Mediterranean diet. Correlation between both diet scores was 0.41, which indicates a modest concordance. Physical function We assessed three different domains of physical function: agility, mobility, and overall physical functioning. Persons were defined as having impaired agility when they answered “a lot” to the following question from the Rosow and Breslau scale (30): “On an average day with your current health, would you be limited in bending and kneeling?”, whose categories of response were “yes, a lot,” “yes, a little,” and “not at all,” In the same way, impairment in mobility was defined as answering “a lot” to any of the following questions from the Rosow and Breslau scale: “On an average day with your current health, would you be limited in the following activities: 1) picking up or carrying a shopping bag?; 2) climbing one flight of stairs?; 3) walking several city blocks (a few hundred meters)?” Lastly, a limitation in overall physical function was deemed to exist when the score on the physical function score of the 12-Item Short-Form Health Survey (SF-12) decreased ≥5 points from baseline to follow-up (31). Other variables At baseline, we obtained information on sociodemographic variables, lifestyle, anthropometrics, and disease history (32,33). Weight and height were measured in each subject under standardized conditions. Body mass index was calculated as weight (kilogram) divided by square height (square meter). Physical activity during leisure time (metabolic equivalent hours per week) was ascertained with the EPIC cohort questionnaire, validated in Spain (32). Sedentary behavior was approximated by the time (hours per week) spent watching TV. Total energy intake (kilocalories per day) was estimated with standard composition tables of foods in Spain. Blood pressure was measured with a validated sphygmomanometer using standardized procedures, and hypertension was defined as systolic blood pressure ≥ 140 mm Hg, diastolic blood pressure ≥ 90 mm Hg, or being on antihypertensive drug treatment. Twelve-hour fasting serum glucose was centrally measured with standard techniques, and diabetes mellitus was defined as glucose ≥126 mg/dL or being treated with oral drugs or insulin. Cognitive function was assessed with the Mini-Mental State Examination (MMSE), and cognitive impairment was defined as a MMSE score of <23 (33). Participants also reported the following physician-diagnosed diseases: osteomuscular disease (osteoarthritis, arthritis, and hip fracture), cardiovascular disease (ischemic heart disease, stroke, and heart failure), cancer, chronic lung disease (asthma and chronic bronchitis), and depression requiring treatment. Statistical Analysis We used logistic regression to estimate the odds ratios and the 95% confidence interval of the association between the Mediterranean dietary patterns and incident limitation in physical function. Participants were categorized into sex-specific tertiles of each dietary pattern, and the first tertile (lowest accordance with the Mediterranean diet) was used as the reference in the analyses. To investigate the linear dose–response relationship, we modeled the tertiles of the diet scores as a continuous variable. In addition, we calculated the risk of physical impairment associated with a 2-point increase in the scores (corresponding to approximately 1 SD). The measures to define physical function were agility, mobility, and overall physical functioning as well as impairment in any of these outcomes. Several logistic models were built: the first one adjusting for age and sex; a second model with additional adjustment for educational level, smoking status, sedentary behavior, energy intake, and body mass index; a third model, further adjusting for chronic diseases, to understand their impact on the studied association; and a forth model by further adjusting the analyses for time spent on leisure physical activity and for cognitive impairment, which are closely related to the outcome, to understand their contribution to the studied association. For the models that included the change in overall physical functioning as the outcome, additional adjustment for the baseline value of the SF-12 physical function score was performed and compared with the same model without this adjustment (data not shown) because it is unclear whether analyses of change should be adjusted for baseline values (34). We conducted several sensitivity analyses to assess the robustness of the results. Thus, analyses were replicated in subgroups of individuals at high risk of impaired physical function: age ≥ 70 years, low physical activity, current or former smoking, body mass index ≥ 30 kg/m2, and important morbidity, compared to individuals at low risk. A possible modifying effect of sex was also tested. Differences between subgroups were tested by using likelihood-ratio tests, which compared models with and without cross-product interaction terms. We also repeated the analyses among participants who were not hospitalized during the follow-up to observe if the effect of diet on physical function would be stronger when participants with severe incident diseases were excluded. Finally, we examined the independent association of individual components of the Mediterranean diet scores with impaired physical function. Analyses were conducted using the SAS software, version 9.2 (SAS Institute). Results Supplementary Table 1 and Table 1 show the characteristics of study participants according to the sex-specific tertiles of the MEDAS and MDS scores. Compared to those in the lowest tertile of MDS, those in the highest tertile showed a lower frequency of current smoking, obesity, and diabetes. Those with a higher MEDAS score were also less often current smokers, showed less sedentary behavior, and had a lower frequency of obesity, diabetes, cognitive impairment, and depression. Table 1. Population Characteristics of the Study Participants Across the Tertiles of the Mediterranean Diet (N = 1,630)   MEDAS   Population Characteristics  Tertile 1  Tertile 2   Tertile 3   p Trend   Score, men/women  0–6/0–6  7–8/7  9–14/8–14    Participants, n  545  541  544    Age, y  67.9 ± 6.4  68.1 ± 6.0  68.2 ± 5.5  .34  Educational level, %   ≤Primary  50.1  50.3  50.2  .79   Secondary  25.9  25.9  27.0     University  24.0  23.8  22.8    Smoking status, %   Current smoker  14.9  13.6  9.0  .01   Former smoker  30.0  34.6  31.1     Never smoker  55.1  51.8  59.9    Leisure-time physical activity, MET-h/wk  21.1 ± 15.2  24.5 ± 15.8  22.9 ± 15.3  .06  Time spent watching TV, h/wk  17.9 ± 11.8  17.9 ± 10.5  16.4 ± 9.8  .03  Energy intake, kcal/d  2,117 ± 610  2,049 ± 553  2,009 ± 536  .002  BMI, kg/m2  28.4 ± 4.2  28.2 ± 3.7  27.7 ± 4.0  .004  Diagnosed diseases, %   Hypertension  63.5  68.7  61.7  .54   Diabetes  17.4  12.5  12.3  .02   Cognitive impairmenta  3.1  1.9  1.3  .03   Osteomuscular diseaseb  45.7  42.5  48.4  .30   Cardiovascular diseasec  4.8  5.4  4.8  .99   Cancer  2.4  3.3  3.1  .47   Chronic lung disease  9.9  8.1  9.4  .76   Depression  9.2  6.7  4.6  .003    MEDAS   Population Characteristics  Tertile 1  Tertile 2   Tertile 3   p Trend   Score, men/women  0–6/0–6  7–8/7  9–14/8–14    Participants, n  545  541  544    Age, y  67.9 ± 6.4  68.1 ± 6.0  68.2 ± 5.5  .34  Educational level, %   ≤Primary  50.1  50.3  50.2  .79   Secondary  25.9  25.9  27.0     University  24.0  23.8  22.8    Smoking status, %   Current smoker  14.9  13.6  9.0  .01   Former smoker  30.0  34.6  31.1     Never smoker  55.1  51.8  59.9    Leisure-time physical activity, MET-h/wk  21.1 ± 15.2  24.5 ± 15.8  22.9 ± 15.3  .06  Time spent watching TV, h/wk  17.9 ± 11.8  17.9 ± 10.5  16.4 ± 9.8  .03  Energy intake, kcal/d  2,117 ± 610  2,049 ± 553  2,009 ± 536  .002  BMI, kg/m2  28.4 ± 4.2  28.2 ± 3.7  27.7 ± 4.0  .004  Diagnosed diseases, %   Hypertension  63.5  68.7  61.7  .54   Diabetes  17.4  12.5  12.3  .02   Cognitive impairmenta  3.1  1.9  1.3  .03   Osteomuscular diseaseb  45.7  42.5  48.4  .30   Cardiovascular diseasec  4.8  5.4  4.8  .99   Cancer  2.4  3.3  3.1  .47   Chronic lung disease  9.9  8.1  9.4  .76   Depression  9.2  6.7  4.6  .003  Note: BMI = body mass index; MEDAS = Mediterranean Diet Adherence Screener; MET = metabolic equivalent; MMSE, Mini-Mental State Examination. For continuous variables, mean and standard deviation are reported. aCognitive impairment is defined as a MMSE score < 23. bOsteoarthritis, arthritis, and hip fracture. cIschemic heart disease, stroke, and heart failure. View Large Over a median follow-up of 3.5 years, we identified 663 individuals (40.7%) with incident physical impairment in any of the three domains considered: 408 (25.0%) had limitation in one domain, 161 (9.9%) in two domains, and 94 (5.8%) in the three of them. Specifically, 343 participants had incident impairment in agility, 212 had impaired mobility, and 457 showed a decline in overall physical function. We found no association between the MDS and the risk of impaired agility or mobility, although a 2-point increment in the MDS was marginally associated with lower likelihood for decreased overall physical function (Supplementary Table 2). By contrast, compared to persons in the lowest tertile of the MEDAS score, those in the highest tertile showed a lower odds of impaired agility, impaired mobility, and decreased overall physical functioning, in the fully adjusted analyses (Table 2). Results for impairment in overall physical functioning did not materially change when we removed the adjustment for the baseline SF-12 physical function score. A 2-point increment of the MEDAS score was associated with a 17% lower likelihood for impaired agility, a 21% lower likelihood for impaired mobility, a 17% lower likelihood for decreased overall function, and a 13% lower likelihood of impairment in any of the above domains. Repeating the analyses additionally adjusting for both physical activity and cognitive impairment did not materially change the results (Table 2). Neither did the results change when physical activity and cognitive impairment were entered separately (data not shown). Table 2. Odds Ratios (95% Confidence Interval) for the Association Between the Mediterranean Diet Adherence Screener and Physical Function Impairment During a 3.5-Year Follow-up of Older Adults (N = 1,630) Physical Function Impairment  MEDAS  Continuous per 2-Point Increment in the MEDAS  Tertile 1  Tertile 2  Tertile 3  p Trend  Impairment in agility, n  134  112  97    343   Model 1  1.00  0.98 (0.73; 1.33)  0.62 (0.46; 0.84)  .002  0.80 (0.70; 0.91)   Model 2  1.00  1.03 (0.75; 1.42)  0.69 (0.50; 0.96)  .03  0.84 (0.73; 0.97)   Model 3  1.00  1.02 (0.73; 1.42)  0.67 (0.48; 0.94)  .02  0.83 (0.71; 0.96)   Model 4  1.00  1.07 (0.76; 1.50)  0.70 (0.49; 0.98)  .04  0.84 (0.72; 0.98)  Impairment in mobility, n  89  64  59    212   Model 1  1.00  0.84 (0.59; 1.21)  0.59 (0.41; 0.84)  .004  0.78 (0.66; 0.92)   Model 2  1.00  0.86 (0.59; 1.25)  0.61 (0.42; 0.89)  .01  0.80 (0.67; 0.95)   Model 3  1.00  0.83 (0.56; 1.22)  0.69 (0.40; 0.88)  .01  0.79 (0.66; 0.94)   Model 4  1.00  0.87 (0.58; 1.29)  0.62 (0.42; 0.92)  .02  0.80 (0.67; 0.96)  Impairment in overall physical functioninga,b, n  177  153  127    457   Model 1  1.00  0.82 (0.63; 1.07)  0.60 (0.46; 0.79)  <.001  0.83 (0.73; 0.93)   Model 2  1.00  0.81 (0.62; 1.07)  0.61 (0.46; 0.81)  <.001  0.83 (0.74; 0.94)   Model 3  1.00  0.80 (0.60; 1.06)  0.60 (0.45; 0.79)  <.001  0.83 (0.73; 0.94)   Model 4  1.00  0.82 (0.62; 1.08)  0.61 (0.45; 0.81)  <.001  0.84 (0.74; 0.95)  Impairment in any of the above domains, n  246  215  202    663   Model 1  1.00  0.90 (0.70; 1.16)  0.69 (0.54; 0.89)  .004  0.85 (0.76; 0.95)   Model 2  1.00  0.91 (0.70; 1.18)  0.72 (0.55; 0.93)  .01  0.87 (0.77; 0.97)   Model 3  1.00  0.90 (0.69; 1.18)  0.70 (0.54; 0.92)  .01  0.87 (0.77; 0.97)   Model 4  1.00  0.92 (0.70; 1.20)  0.72 (0.55; 0.94)  .01  0.87 (0.77; 0.98)  Physical Function Impairment  MEDAS  Continuous per 2-Point Increment in the MEDAS  Tertile 1  Tertile 2  Tertile 3  p Trend  Impairment in agility, n  134  112  97    343   Model 1  1.00  0.98 (0.73; 1.33)  0.62 (0.46; 0.84)  .002  0.80 (0.70; 0.91)   Model 2  1.00  1.03 (0.75; 1.42)  0.69 (0.50; 0.96)  .03  0.84 (0.73; 0.97)   Model 3  1.00  1.02 (0.73; 1.42)  0.67 (0.48; 0.94)  .02  0.83 (0.71; 0.96)   Model 4  1.00  1.07 (0.76; 1.50)  0.70 (0.49; 0.98)  .04  0.84 (0.72; 0.98)  Impairment in mobility, n  89  64  59    212   Model 1  1.00  0.84 (0.59; 1.21)  0.59 (0.41; 0.84)  .004  0.78 (0.66; 0.92)   Model 2  1.00  0.86 (0.59; 1.25)  0.61 (0.42; 0.89)  .01  0.80 (0.67; 0.95)   Model 3  1.00  0.83 (0.56; 1.22)  0.69 (0.40; 0.88)  .01  0.79 (0.66; 0.94)   Model 4  1.00  0.87 (0.58; 1.29)  0.62 (0.42; 0.92)  .02  0.80 (0.67; 0.96)  Impairment in overall physical functioninga,b, n  177  153  127    457   Model 1  1.00  0.82 (0.63; 1.07)  0.60 (0.46; 0.79)  <.001  0.83 (0.73; 0.93)   Model 2  1.00  0.81 (0.62; 1.07)  0.61 (0.46; 0.81)  <.001  0.83 (0.74; 0.94)   Model 3  1.00  0.80 (0.60; 1.06)  0.60 (0.45; 0.79)  <.001  0.83 (0.73; 0.94)   Model 4  1.00  0.82 (0.62; 1.08)  0.61 (0.45; 0.81)  <.001  0.84 (0.74; 0.95)  Impairment in any of the above domains, n  246  215  202    663   Model 1  1.00  0.90 (0.70; 1.16)  0.69 (0.54; 0.89)  .004  0.85 (0.76; 0.95)   Model 2  1.00  0.91 (0.70; 1.18)  0.72 (0.55; 0.93)  .01  0.87 (0.77; 0.97)   Model 3  1.00  0.90 (0.69; 1.18)  0.70 (0.54; 0.92)  .01  0.87 (0.77; 0.97)   Model 4  1.00  0.92 (0.70; 1.20)  0.72 (0.55; 0.94)  .01  0.87 (0.77; 0.98)  Note: BMI = body mass index; MEDAS = Mediterranean Diet Adherence Screener; MET = metabolic equivalent; SF-12 = 12-Item Short-Form Health Survey. Model 1: logistic model adjusted for age and sex. Model 2: logistic model adjusted as in Model 1 and for educational level (≤primary, secondary, university), smoking status (never smoker, former smoker, current smoker), time spent watching television (quintiles of h/wk), energy intake (quintiles of kcal/d), and BMI. Model 3: logistic model adjusted as in Model 2 and for osteomuscular disease, cardiovascular disease, cancer, chronic lung disease, and depression requiring treatment. Model 4: logistic model adjusted as in Model 3 and for physical activity (quintiles of MET-h/wk) and cognitive impairment. a≥5-point decrease in the SF-12 physical component summary score from baseline to follow-up. bAdditionally adjusted for the SF-12 function score summary at baseline. View Large We performed sensitivity analyses in the oldest old individuals (≥70 years) and in subgroups of subjects at high risk of impaired physical function, compared with subjects at low risk (Supplementary Table 3). Although statistical significance was lost in some cases due to smaller sample size, the analyses consistently showed a tendency toward reduced likelihood of impaired physical function associated with a higher MEDAS score in both high risk and low risk individuals. We did not observe a sex interaction in any of the studied associations except for the MEDAS score and impaired mobility; the risk of mobility limitation associated with a 2-point increment in MEDAS score was stronger for men (odds ratio: 0.59, 95% confidence interval: 0.43; 0.81) than for women (odds ratio: 0.91, 95% confidence interval: 0.72; 1.14), although the direction remained the same. Moreover, when we removed from the analyses the individuals hospitalized during follow-up, no substantial change in the results was observed. Table 3 shows the association between each MEDAS component and physical function impairment. As expected, there was a tendency toward reduced likelihood of impaired physical function associated with many components of the diet; moreover, achievement of the target for nut consumption was significantly associated with lower odds of impaired agility and of decreased overall functioning. Also the achievement of target for fruits, fish or seafood, and preference of white over red meat was associated with a lower odds of decreased overall functioning. Table 3. Odds Ratios (95% Confidence Interval) for the Association Between Achievement of the Targets of the Mediterranean Diet Adherence Screener and Physical Function Impairment During a 3.5-Year Follow-up of Older Adults (N = 1,630) Targets of the MEDAS  Impairment in Agility  Impairment in Mobility  Impairment in Overall Physical Functioninga,b  Olive oil for cooking (yes)  1.45 (0.71; 2.95)  1.18 (0.52; 2.65)  1.87 (1.00; 3.48)  Olive oil (≥4 tablespoons/d)  0.64 (0.37; 1.11)  0.68 (0.34; 1.34)  0.80 (0.52; 1.22)  Vegetables (≥2 servings/d)  0.88 (0.59; 1.31)  0.98 (0.61; 1.56)  1.19 (0.86; 1.65)  Fruits (≥3 servings/d)  1.05 (0.76; 1.44)  0.78 (0.53; 1.13)  0.66 (0.49; 0.87)  Red meat, hamburger, or sausage (<1 serving/d)  1.10 (0.71; 1.71)  1.17 (0.68; 2.01)  0.95 (0.67; 1.36)  Butter, margarine, or cream (<1 serving/d)  0.91 (0.56; 1.48)  1.37 (0.76; 2.49)  1.36 (0.87; 2.12)  Sugar-sweetened or carbonated beverages (<1 time/d)  0.87 (0.52; 1.45)  1.19 (0.63; 2.24)  0.88 (0.57; 1.37)  Wine (≥1 serving/d)  1.02 (0.73; 1.44)  0.78 (0.51; 1.19)  0.96 (0.73; 1.27)  Legumes (≥3 times/wk)  1.02 (0.71; 1.47)  0.70 (0.44; 1.12)  0.86 (0.63; 1.17)  Fish or seafood (≥3 times/wk)  0.82 (0.61; 1.09)  0.76 (0.55; 1.06)  0.78 (0.61; 1.00)  Commercial baked goods (<2 times/wk)  0.89 (0.67; 1.18)  0.84 (0.60; 1.17)  0.96 (0.75; 1.22)  Nuts (≥3 times/wk)  0.68 (0.49; 0.94)  0.71 (0.48; 1.05)  0.72 (0.54; 0.95)  Preference for white over red meat  0.92 (0.67; 1.26)  1.10 (0.76; 1.58)  0.73 (0.55; 0.96)  Foods with “sofrito” (≥2 times/wk)  0.74 (0.43; 1.29)  0.75 (0.40; 1.41)  0.79 (0.50; 1.27)  Targets of the MEDAS  Impairment in Agility  Impairment in Mobility  Impairment in Overall Physical Functioninga,b  Olive oil for cooking (yes)  1.45 (0.71; 2.95)  1.18 (0.52; 2.65)  1.87 (1.00; 3.48)  Olive oil (≥4 tablespoons/d)  0.64 (0.37; 1.11)  0.68 (0.34; 1.34)  0.80 (0.52; 1.22)  Vegetables (≥2 servings/d)  0.88 (0.59; 1.31)  0.98 (0.61; 1.56)  1.19 (0.86; 1.65)  Fruits (≥3 servings/d)  1.05 (0.76; 1.44)  0.78 (0.53; 1.13)  0.66 (0.49; 0.87)  Red meat, hamburger, or sausage (<1 serving/d)  1.10 (0.71; 1.71)  1.17 (0.68; 2.01)  0.95 (0.67; 1.36)  Butter, margarine, or cream (<1 serving/d)  0.91 (0.56; 1.48)  1.37 (0.76; 2.49)  1.36 (0.87; 2.12)  Sugar-sweetened or carbonated beverages (<1 time/d)  0.87 (0.52; 1.45)  1.19 (0.63; 2.24)  0.88 (0.57; 1.37)  Wine (≥1 serving/d)  1.02 (0.73; 1.44)  0.78 (0.51; 1.19)  0.96 (0.73; 1.27)  Legumes (≥3 times/wk)  1.02 (0.71; 1.47)  0.70 (0.44; 1.12)  0.86 (0.63; 1.17)  Fish or seafood (≥3 times/wk)  0.82 (0.61; 1.09)  0.76 (0.55; 1.06)  0.78 (0.61; 1.00)  Commercial baked goods (<2 times/wk)  0.89 (0.67; 1.18)  0.84 (0.60; 1.17)  0.96 (0.75; 1.22)  Nuts (≥3 times/wk)  0.68 (0.49; 0.94)  0.71 (0.48; 1.05)  0.72 (0.54; 0.95)  Preference for white over red meat  0.92 (0.67; 1.26)  1.10 (0.76; 1.58)  0.73 (0.55; 0.96)  Foods with “sofrito” (≥2 times/wk)  0.74 (0.43; 1.29)  0.75 (0.40; 1.41)  0.79 (0.50; 1.27)  Note: MEDAS = Mediterranean Diet Adherence Screener; SF-12 = 12-Item Short-Form Health Survey. Analysis adjusted as in Model 3 in Tables 2 and 3. a≥5-point decrease in the SF-12 physical component summary score from baseline to follow-up. bAdditionally adjusted for the SF-12 physical function score at baseline. View Large Discussion In this study, among older adults, a higher accordance with a Mediterranean dietary pattern, as measured by the MEDAS score, was associated with a lower likelihood of impairment in agility, mobility, and overall physical functioning. By contrast, accordance with the MDS was only associated with overall physical functioning. Our finding of better functioning associated with the MEDAS but to a much lesser extent, the MDS highlights the complexity of defining a universal Mediterranean dietary pattern. Because operational definitions of this diet have been derived from specific Mediterranean regions and time periods, many different scores have been used in the literature, and there is an ongoing debate about which components should be considered in each score (35), including the type of fat, specification of grain type, definition of moderate alcohol intake, and the presence of dairy products, nuts, and fish. For our analyses, we chose one of the most widely used sample-based scores (MDS) and the normative score (MEDAS) developed for Spain. Although both patterns focused on the consumption of fruits, vegetables, fish, and unsaturated fat, they differ in the weight assigned to nuts, olive oil, and white meat, and also in the items considered unhealthy (dairy and all types of meat in the MDS vs red meat, butter, sugar-sweetened beverages, and baked goods in the MEDAS). In addition, the number of components of each score (9 in the MDS vs 14 in the MEDAS) and the method of scoring (data-driven vs arbitrary cutoff values) may also have influenced the results. In addition, the number of components and the use of cutoff values in the MEDAS resulted in larger variations in scoring across the study participants and, thus, may have provided greater power to observe an association. To our knowledge, a few population studies have investigated the association between the Mediterranean diet and impaired physical function. Among the cross-sectional analyses, in the Nurses’ Health Study, a North American cohort, a higher adherence to the MDS was related to a greater likelihood of no major limitations in physical function, measured based on the mobility questions from the SF-36 (18). Also, in another Spanish population, higher adherence to the MDS was associated with higher scores in the physical component of the SF-12 in men, but not in women (17). Moreover, in participants from the U.S. NHANES and Israeli MABAT ZAHAV studies, a higher adherence to the MDS was associated with significant higher walking speed. This association was attenuated after adjustment for cognitive decline (19). Finally, a study with an Australian cohort found an association between higher adherence to the MDS and better scores in the physical function score of the SF-12 (20), which is in line with our study. In the InCHIANTI cohort, a prospective longitudinal study that assessed this association among 935 Italian participants, the authors found that higher adherence to the MDS was associated with better lower body performance after 9 years of follow-up (21). Finally, in the Health ABC cohort, with North American participants, walking speed over 8 years was faster among those with higher adherence to a Mediterranean diet score (22). As regards the association between the Mediterranean diet and disability, which is a common consequence of impaired physical function, a prospective analysis of the French Three City Study found that women in the highest category of the MDS had a 50% risk reduction in incident disability in activities of daily living, compared to women in the lowest category; however, no association was found for men (36). In a previous analysis of the Seniors-ENRICA cohort, we also found that a higher score on either the MDS or the MEDAS was associated with lower risk of the frailty syndrome, defined as having three out of the following five criteria: exhaustion, weight loss, low grip strength, slow walking speed, or low physical activity (37); however, when weight loss was excluded from the frailty definition, only the MEDAS score was significantly associated with a lower risk of frailty. Our present results suggest that the beneficial effect of the Mediterranean diet on disability and frailty in older adults could be mediated through better physical functioning. In this study, impairment in overall physical functioning was defined as an at least 5-point decrease in the physical function score of the SF-12 from baseline to follow-up. In the IQOLA project, which examined the impact of chronic conditions (arthritis, chronic lung disease, and congestive heart failure) on quality of life in 15 countries, those conditions with a major impact on quality of life reduced the score on the physical component of the SF-12 less than 5 points (38). Therefore, our definition of impaired functioning seems appropriate and of clinical relevance. Our main findings were robust. The inverse associations between the MEDAS score and impaired agility or mobility were similar or slightly stronger in subgroups of participants defined by their higher risk of functional limitations. In addition, when the analyses were adjusted for physical activity and cognitive impairment, which are closely related to the outcome, the main results still held. Also, adjusting for chronic diseases, which might lie in the causal pathway, did not materially change the results. Several mechanisms, such as reduced oxidation and inflammation, might play a role in the association between Mediterranean diet and physical function. Specifically, accordance with the Mediterranean diet has been associated with decreased oxidative damage and lower serum levels of inflammatory markers (39,40). Also, augmented oxidative damage and inflammatory markers have previously been found to predict the onset of physical limitations (41,42). In addition, homocysteine, a sensitive marker for a deficiency in B vitamins, such as folic acid, has been shown to be independently associated with physical function (14). Thus, B vitamins, provided by the Mediterranean diet, may influence physical function through homocysteine levels. Strengths of this study are its prospective design, collection of food consumption with a validated diet history, and adjustment for many potential confounders. Some limitations should also be acknowledged. Equal weights were assigned to each component of the Mediterranean diet scores even though their health effects might be different. Also, certain misclassification of food consumption cannot be ruled out, despite excluding participants with an implausibly high or low energy intake. Another limitation was the use of self-reported information as a proxy for mobility, agility, and overall physical performance; however, these measurements have been widely used in clinical practice and in population studies. There were a high percentage of persons excluded from the analyses because many participants did not complete the Mini-Mental State Examination. Although results were similar when we included persons without this information, we still decided to exclude participants with unknown cognitive status from the analytical sample because the impact of cognitive function on physical function is unclear. Therefore, the studied participants are not a reflection of the national structure of the Spanish older population. Moreover, functional impairment was evaluated at the end of the follow-up, so that temporality and development of impairments during the interval period could not be fully ascertained. Finally, as in any observational study, some residual confounding may persist. In conclusion, a Mediterranean-style diet was associated with a lower likelihood of impaired physical function in Spanish older adults. The weaker association of the MDS with this outcome suggests that a more precise definition of this diet is needed to be able to find consistent health effects. Supplementary Material Please visit the article online at http://gerontologist.oxfordjournals.org/ to view supplementary material. Funding This work was supported by grants from the Instituto de Salud Carlos III, State Secretary of R+D+I of Spain, and FEDER/FSE (FIS 12/1166 and 13/0288) and the European Union (FP7-HEALTH-2012 Proposal No 305483-2, FRAILOMIC Initiative). Conflict of Interest None of the authors has a conflict of interest related to the subject of this work. References 1. Wang H Dwyer-Lindgren L Lofgren KTet al.  . Age-specific and sex-specific mortality in 187 countries, 1970-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet . 2012; 380: 2071– 2094. doi: 10.1016/S0140-6736(12)61719-X Google Scholar CrossRef Search ADS PubMed  2. Murray CJ Vos T Lozano Ret al.  . Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet . 2012; 380: 2197– 2223. doi: 10.1016/S0140-6736(12)61689-4 Google Scholar CrossRef Search ADS PubMed  3. Chatterji S Byles J Cutler D Seeman T Verdes E. Health, functioning, and disability in older adults-present status and future implications. Lancet . 2014; 386: 563– 565. doi: 10.1016/S0140-6736(14)61462-8 4. Verbrugge LM Jette AM. The disablement process. Soc Sci Med . 1994; 38: 1– 14. Google Scholar CrossRef Search ADS PubMed  5. Jette AM. Toward a common language for function, disability, and health. Phys Ther . 2006; 86: 726– 734. Google Scholar CrossRef Search ADS PubMed  6. Fried LP Ferrucci L Darer J Williamson JD Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol A Biol Sci Med Sci . 2004; 59: 255– 263. Google Scholar CrossRef Search ADS PubMed  7. Studenski S Perera S Patel Ket al.  . Gait speed and survival in older adults. JAMA . 2011; 305: 50– 58. doi: 10.1001/jama.2010.1923 Google Scholar CrossRef Search ADS PubMed  8. Guralnik JM Simonsick EM Ferrucci Let al.  . A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol . 1994; 49: M85– M94. Google Scholar CrossRef Search ADS PubMed  9. Guallar-Castillón P Sagardui-Villamor J Banegas JRet al.  . Waist circumference as a predictor of disability among older adults. Obesity . 2007; 15: 233. doi: 10.1038/oby.2007.532 Google Scholar CrossRef Search ADS PubMed  10. Artaud F Dugravot A Sabia S Singh-Manoux A Tzourio C Elbaz A. Unhealthy behaviours and disability in older adults: three-City Dijon cohort study. BMJ . 2013; 347: f4240. doi: 10.1136/bmj.f4240 Google Scholar CrossRef Search ADS PubMed  11. Vercambre MN Boutron-Ruault MC Ritchie K Clavel-Chapelon F Berr C. Long-term association of food and nutrient intakes with cognitive and functional decline: a 13-year follow-up study of elderly French women. Br J Nutr . 2009; 102: 419– 427. doi: 10.1017/S0007114508201959 Google Scholar CrossRef Search ADS PubMed  12. Houston DK Stevens J Cai J Haines PS. Dairy, fruit, and vegetable intakes and functional limitations and disability in a biracial cohort: the Atherosclerosis Risk in Communities Study. Am J Clin Nutr . 2005; 81: 515– 522. doi: 10.1016/s1073-5437(08)70273-0 Google Scholar CrossRef Search ADS PubMed  13. Cesari M Pahor M Bartali Bet al.  . Antioxidants and physical performance in elderly persons: the Invecchiare in Chianti (InCHIANTI) study. Am J Clin Nutr . 2004; 79: 289– 294. Google Scholar CrossRef Search ADS PubMed  14. Ng TP Aung KC Feng L Scherer SC Yap KB. Homocysteine, folate, vitamin B-12, and physical function in older adults: cross-sectional findings from the Singapore Longitudinal Ageing Study. Am J Clin Nutr . 2012; 96: 1362– 1368. doi: 10.3945/ajcn.112.035741 Google Scholar CrossRef Search ADS PubMed  15. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol . 2002; 13: 3– 9. Google Scholar CrossRef Search ADS PubMed  16. Xu B Houston DK Locher JLet al.  . Higher Healthy Eating Index-2005 scores are associated with better physical performance. J Gerontol A Biol Sci Med Sci . 2012; 67: 93– 99. doi: 10.1093/gerona/glr159 Google Scholar CrossRef Search ADS PubMed  17. Muñoz MA Fíto M Marrugat J Covas MI Schröder H; REGICOR and HERMES Investigators. Adherence to the Mediterranean diet is associated with better mental and physical health. Br J Nutr . 2009; 101: 1821– 1827. doi: 10.1017/S0007114508143598 Google Scholar CrossRef Search ADS PubMed  18. Samieri C Sun Q Townsend MKet al.  . The association between dietary patterns at midlife and health in aging: an observational study. Ann Intern Med . 2013; 159: 584– 591. doi: 10.7326/0003-4819-159-9-201311050-00004 Google Scholar CrossRef Search ADS PubMed  19. Zbeida M Goldsmith R Shimony T Vardi H Naggan L Shahar DR. Mediterranean diet and functional indicators among older adults in non-Mediterranean and Mediterranean countries. J Nutr Health Aging . 2014; 18: 411– 418. doi: 10.1007/s12603-014-0003-9 Google Scholar CrossRef Search ADS PubMed  20. Milte CM Thorpe MG Crawford D Ball K McNaughton SA. Associations of diet quality with health-related quality of life in older Australian men and women. Exp Gerontol . 2015; 64: 8– 16. doi: 10.1016/j.exger.2015.01.047 Google Scholar CrossRef Search ADS PubMed  21. Milaneschi Y Bandinelli S Corsi AMet al.  . Mediterranean diet and mobility decline in older persons. Exp Gerontol . 2011; 46: 303– 308. doi: 10.1016/j.exger.2010.11.030 Google Scholar CrossRef Search ADS PubMed  22. Shahar DR Houston DK Hue TFet al.  . Adherence to Mediterranean diet and decline in walking speed over 8 years in community-dwelling older adults. J Am Geriatr Soc . 2012; 60: 1881– 1888. doi: 10.1111/j.1532-5415.2012.04167.x Google Scholar CrossRef Search ADS PubMed  23. Rodríguez-Artalejo F Graciani A Guallar-Castillón Pet al.  . Rationale and methods of the study on nutrition and cardiovascular risk in Spain (ENRICA). Rev Esp Cardiol . 2011; 64: 876– 882. doi: 10.1016/j.recesp.2011.05.019 Google Scholar CrossRef Search ADS PubMed  24. León-Muñoz LM García-Esquinas E López-García E Banegas JR Rodríguez-Artalejo F. Major dietary patterns and risk of frailty in older adults: a prospective cohort study. BMC Med . 2015; 13: 11. doi: 10.1186/s12916-014-0255-6 Google Scholar CrossRef Search ADS PubMed  25. Guallar-Castillón P Sagardui-Villamor J Balboa-Castillo Tet al.  . Validity and reproducibility of a Spanish dietary history. PLoS One . 2014; 9: e86074. doi: 10.1371/journal.pone.0086074 Google Scholar CrossRef Search ADS PubMed  26. Trichopoulou A Kouris-Blazos A Wahlqvist MLet al.  . Diet and overall survival in elderly people. BMJ . 1995; 311: 1457– 1460. Google Scholar CrossRef Search ADS PubMed  27. 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– 2608. doi: 10.1056/NEJMoa025039 Google Scholar CrossRef Search ADS PubMed  28. Struijk EA Beulens JW May AMet al.  . Dietary patterns in relation to disease burden expressed in disability-adjusted life years. Am J Clin Nutr . 2014; 100: 1158– 1165. doi: 10.3945/ajcn.113.082032 Google Scholar CrossRef Search ADS PubMed  29. Schröder H Fitó M Estruch Ret al.  . A short screener is valid for assessing Mediterranean diet adherence among older Spanish men and women. J Nutr . 2011; 141: 1140– 1145. doi: 10.3945/jn.110.135566 Google Scholar CrossRef Search ADS PubMed  30. Rosow I Breslau N. A Guttman health scale for the aged. J Gerontol . 1966; 21: 556– 559. Google Scholar CrossRef Search ADS PubMed  31. Vilagut G Valderas JM Ferrer M Garin O López-García E Alonso J. Interpretation of SF-36 and SF-12 questionnaires in Spain: physical and mental components. Med Clin . 2008; 130: 726– 735. Google Scholar CrossRef Search ADS   32. Pols MA Peeters PH Ocké MC Slimani N Bueno-de-Mesquita HB Collette HJ. Estimation of reproducibility and relative validity of the questions included in the EPIC Physical Activity Questionnaire. Int J Epidemiol . 1997; 26( suppl 1): S181– S189. Google Scholar CrossRef Search ADS PubMed  33. Graciani A Banegas JR Guallar-Castillón P Domínguez-Rojas V Rodríguez-Artalejo F. Cognitive assessment of the non-demented elderly community dwellers in Spain. Dement Geriatr Cogn Disord . 2006; 21: 104– 112. doi: 10.1159/000090509 Google Scholar CrossRef Search ADS PubMed  34. Glymour MM Weuve J Berkman LF Kawachi I Robins JM. When is baseline adjustment useful in analyses of change? An example with education and cognitive change. Am J Epidemiol . 2005; 162: 267– 278. doi: 10.1093/aje/kwi187 Google Scholar CrossRef Search ADS PubMed  35. Bach A Serra-Majem L Carrasco JLet al.  . The use of indexes evaluating the adherence to the Mediterranean diet in epidemiological studies: a review. Public Health Nutr . 2006; 9: 132– 146. Google Scholar CrossRef Search ADS PubMed  36. Féart C Pérès K Samieri C Letenneur L Dartigues JF Barberger-Gateau P. Adherence to a Mediterranean diet and onset of disability in older persons. Eur J Epidemiol . 2011; 26: 747– 756. doi: 10.1007/s10654-011-9611-4 Google Scholar CrossRef Search ADS PubMed  37. León-Muñoz LM Guallar-Castillón P López-García E Rodríguez-Artalejo F. Mediterranean diet and risk of frailty in community-dwelling older adults. J Am Med Dir Assoc . 2014; 15: 899– 903. doi: 10.1016/j.jamda.2014.06.013. Google Scholar CrossRef Search ADS PubMed  38. Alonso J Ferrer M Gandek Bet al.  ; IQOLA Project Group. Health-related quality of life associated with chronic conditions in eight countries: results from the International Quality of Life Assessment (IQOLA) Project. Qual Life Res . 2004; 13: 283– 298. Google Scholar CrossRef Search ADS PubMed  39. Fitó M Guxens M Corella Det al.  ; PREDIMED Study Investigators. Effect of a traditional Mediterranean diet on lipoprotein oxidation: a randomized controlled trial. Arch Intern Med . 2007; 167: 1195– 1203. doi: 10.1001/archinte.167.11.1195 Google Scholar CrossRef Search ADS PubMed  40. Esposito K Marfella R Ciotola Met al.  . Effect of a Mediterranean-style diet on endothelial dysfunction and markers of vascular inflammation in the metabolic syndrome: a randomized trial. JAMA . 2004; 292: 1440– 1446. doi: 10.1001/jama.292.12.1440 Google Scholar CrossRef Search ADS PubMed  41. Cesari M Kritchevsky SB Nicklas BJet al.  . Lipoprotein peroxidation and mobility limitation: results from the Health, Aging, and Body Composition Study. Arch Intern Med . 2005; 165: 2148– 2154. doi: 10.1001/archinte.165.18.2148 Google Scholar CrossRef Search ADS PubMed  42. Penninx BW Kritchevsky SB Newman ABet al.  . Inflammatory markers and incident mobility limitation in the elderly. J Am Geriatr Soc . 2004; 52: 1105– 1113. doi: 10.1111/j.1532-5415.2004.52308.x Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2016. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The Journals of Gerontology Series A: Biomedical Sciences and Medical Sciences Oxford University Press

Loading next page...
 
/lp/ou_press/mediterranean-dietary-patterns-and-impaired-physical-function-in-older-26bNJ2OEp8
Publisher
Oxford University Press
Copyright
© The Author(s) 2016. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.
ISSN
1079-5006
eISSN
1758-535X
D.O.I.
10.1093/gerona/glw208
Publisher site
See Article on Publisher Site

Abstract

Abstract Background Information about nutritional risk factors of functional limitation is scarce. The aim of this study was to examine the association between the Mediterranean diet and risk of physical function impairment in older adults. Methods We used data from 1,630 participants in the Seniors-ENRICA cohort aged ≥60 years. In 2008–2010, adherence to the Mediterranean diet pattern was measured with the Mediterranean Diet Score (MDS) and the Mediterranean Diet Adherence Screener (MEDAS). Study participants were followed up through 2012 to assess incident impairment in agility and mobility as well as impairment in overall physical functioning, defined as a ≥5-point decrease from baseline to follow-up in the physical component summary of the 12-Item Short-Form Health Survey. Results Over a median follow-up of 3.5 years, we identified 343 individuals with agility limitation, 212 with mobility limitation, and 457 with decreased overall physical functioning. No association was found between the MDS score and the likelihood of impaired agility or mobility, although a 2-point increment in the MDS score was marginally associated with lower likelihood for decreased overall physical function. Compared to individuals in the lowest tertile of the MEDAS score, those in the highest tertile showed a lower odds of agility limitation (odds ratio: 0.67, 95% confidence interval: 0.48; 0.94, p trend = .02), mobility limitation (odds ratio: 0.69, 95% confidence interval: 0.40; 0.88, p trend = .01), and decreased overall physical functioning (odds ratio: 0.60, 95% confidence interval: 0.45; 0.79, p trend < .001). Conclusions In this prospective cohort study, a Mediterranean-style dietary pattern, especially when measured with the MEDAS, was associated with a lower likelihood of physical function impairment in older adults. MEDAS, MDS, Mobility, Agility Life expectancy in Spain is among the highest in Europe and has increased from 69.0 to 78.4 years in men and from 74.6 to 84.2 years in women in the period between 1970 and 2010 (1). Nevertheless, persons who live longer are not necessarily in good health (2,3). In fact, many older adults suffer from multimorbidity that, in combination with the ageing-related decline in many biological systems, may result in impairments of physical functioning (4,5). In turn, these functional limitations can lead to disability, defined as a difficulty or dependency in carrying out activities essential to independent living (6). Moreover, functional limitations have also been associated with increased risk of institutionalization and death (7,8). Several modifiable behavioral factors have been associated with limitations in physical functioning in older adults, including obesity and low physical activity (9,10). In addition, previous research suggests that certain nutrients (antioxidants and B vitamins) and food products (fruit, vegetables, and dairy) play a role in maintaining physical functioning in old age (10–14). One limitation of investigating the effect of individual dietary factors is that nutrients or food products tend to correlate or even interact with each other (15). Also, the effects of individual dietary factors may be too small to be detected. Thus, the use of dietary patterns could overcome these limitations (15). So far, only a few studies have investigated the association of dietary patterns with impaired physical function (16–21); their results suggested a protective effect of the Mediterranean diet (17–21). However, the majority of these studies were limited by their cross-sectional design (17–20), and the previous longitudinal studies performed focused on only one domain of physical performance (21,22). Therefore, the objective of this study was to examine the prospective association between a Mediterranean-style diet and the impairment of physical function, specifically of agility, mobility, and overall physical functioning, in a large population of older adults from Spain. Methods Study Design and Participants Data were taken from the Seniors-ENRICA cohort, whose methods have been reported elsewhere (23,24). In brief, the cohort was derived from the ENRICA study, a survey conducted in 2008–2010 among 12,948 individuals representative of the noninstitutionalized adult population of Spain. The study participants aged 60 years or older (n = 2,614) from the ENRICA study comprised the Seniors-ENRICA cohort. At baseline, information on sociodemographic variables, lifestyle, health status, and morbidity was collected through a phone interview. In two subsequent home visits, appropriately trained research staff collected dietary information, conducted a physical exam, and obtained blood and urine samples. Participants were followed up through 2012, when a second wave of data collection was performed to update information from baseline. Ninety-five participants (3.6%) died during follow-up. From the 2,519 Seniors-ENRICA cohort participants alive, we excluded 9 participants with baseline medical diagnosis of dementia or Alzheimer’s disease and 441 participants without information about their cognitive function. We also excluded 8 individuals without diet information or an implausibly high or low energy intake (outside the range of 800–5,000 kcal/d for men and 500–4,000 kcal/d for women) and 160 participants who lacked data on mobility (n = 139), agility (n = 2), or overall physical function score (n = 19). Individuals excluded were less educated and reported a higher prevalence of diagnosed diabetes, osteomuscular diseases, cardiovascular diseases, and cancer than individuals without missing values. Additionally, we excluded participants with baseline limitations in physical functioning: 186 with impaired mobility and 85 with impaired agility. Thus, the analyses were conducted on 1,630 individuals. Study participants gave written informed consent. The study was approved by the Clinical Research Ethics Committee of the La Paz University Hospital in Madrid. Study Variables Diet At baseline, information on food consumption was obtained through a validated computer-assisted face-to-face diet history, which was developed from that used in the EPIC cohort study in Spain (25). Additional information about validation of this diet history has been provided in the online Supplementary Material. Two scores were used to measure accordance with a Mediterranean-style diet, the Mediterranean Diet Score (MDS) and the Mediterranean Diet Adherence Screener (MEDAS) (26–28). The MDS was developed by Trichopoulou and colleagues and has been widely used in Mediterranean as well as non-Mediterranean countries (26–28). The MDS is based on nine different items (vegetables, fruit and nuts, legumes, grains, fish and seafood, the ratio of unsaturated fatty acids to saturated fatty acids, meat and poultry, dairy, and alcohol) (27). A value of 1 was assigned to moderate alcohol intake (5–25 g/d for women and 10–50 g/d for men) and to an intake above the sex-specific median for all other items, except for meat and poultry and for dairy. For these food groups, considered detrimental for health, 1 point was assigned to an intake below the sex-specific median. Scores for all nine items were summed, resulting in a range from 0 to 9, whereby a higher score reflects better accordance with the Mediterranean diet. Due to the use of sample medians, this score is highly dependent on the sample characteristics. The MEDAS score was developed to assess compliance with the dietary intervention of the PREDIMED trial and uses defined cutoff values for its 14 components (29). The MEDAS score includes two questions on food intake habits and 12 questions on food consumption. One point was given for each of the following components: using olive oil as the principal source of fat for cooking; preferring white meat over red meat; ≥4 tablespoons of olive oil/d; ≥2 servings of vegetables/d; ≥3 pieces of fruit/d; <1 serving of red meat, a hamburger or sausage/d; <1 serving of butter, margarine, or cream/d; <1 time of sugar-sweetened or carbonated beverages/d; ≥1 servings of red wine/d; ≥3 servings of legumes/wk; ≥3 servings of fish or seafood/wk; <2 commercial baked goods/wk; ≥3 servings of nuts/wk; and ≥2 servings/wk of a dish with a traditional sauce of tomatoes, garlic, onion, or leeks sautéed in olive oil (sofrito). The MEDAS score ranges from 0 to 14, with a higher score indicating better accordance with the Mediterranean diet. Correlation between both diet scores was 0.41, which indicates a modest concordance. Physical function We assessed three different domains of physical function: agility, mobility, and overall physical functioning. Persons were defined as having impaired agility when they answered “a lot” to the following question from the Rosow and Breslau scale (30): “On an average day with your current health, would you be limited in bending and kneeling?”, whose categories of response were “yes, a lot,” “yes, a little,” and “not at all,” In the same way, impairment in mobility was defined as answering “a lot” to any of the following questions from the Rosow and Breslau scale: “On an average day with your current health, would you be limited in the following activities: 1) picking up or carrying a shopping bag?; 2) climbing one flight of stairs?; 3) walking several city blocks (a few hundred meters)?” Lastly, a limitation in overall physical function was deemed to exist when the score on the physical function score of the 12-Item Short-Form Health Survey (SF-12) decreased ≥5 points from baseline to follow-up (31). Other variables At baseline, we obtained information on sociodemographic variables, lifestyle, anthropometrics, and disease history (32,33). Weight and height were measured in each subject under standardized conditions. Body mass index was calculated as weight (kilogram) divided by square height (square meter). Physical activity during leisure time (metabolic equivalent hours per week) was ascertained with the EPIC cohort questionnaire, validated in Spain (32). Sedentary behavior was approximated by the time (hours per week) spent watching TV. Total energy intake (kilocalories per day) was estimated with standard composition tables of foods in Spain. Blood pressure was measured with a validated sphygmomanometer using standardized procedures, and hypertension was defined as systolic blood pressure ≥ 140 mm Hg, diastolic blood pressure ≥ 90 mm Hg, or being on antihypertensive drug treatment. Twelve-hour fasting serum glucose was centrally measured with standard techniques, and diabetes mellitus was defined as glucose ≥126 mg/dL or being treated with oral drugs or insulin. Cognitive function was assessed with the Mini-Mental State Examination (MMSE), and cognitive impairment was defined as a MMSE score of <23 (33). Participants also reported the following physician-diagnosed diseases: osteomuscular disease (osteoarthritis, arthritis, and hip fracture), cardiovascular disease (ischemic heart disease, stroke, and heart failure), cancer, chronic lung disease (asthma and chronic bronchitis), and depression requiring treatment. Statistical Analysis We used logistic regression to estimate the odds ratios and the 95% confidence interval of the association between the Mediterranean dietary patterns and incident limitation in physical function. Participants were categorized into sex-specific tertiles of each dietary pattern, and the first tertile (lowest accordance with the Mediterranean diet) was used as the reference in the analyses. To investigate the linear dose–response relationship, we modeled the tertiles of the diet scores as a continuous variable. In addition, we calculated the risk of physical impairment associated with a 2-point increase in the scores (corresponding to approximately 1 SD). The measures to define physical function were agility, mobility, and overall physical functioning as well as impairment in any of these outcomes. Several logistic models were built: the first one adjusting for age and sex; a second model with additional adjustment for educational level, smoking status, sedentary behavior, energy intake, and body mass index; a third model, further adjusting for chronic diseases, to understand their impact on the studied association; and a forth model by further adjusting the analyses for time spent on leisure physical activity and for cognitive impairment, which are closely related to the outcome, to understand their contribution to the studied association. For the models that included the change in overall physical functioning as the outcome, additional adjustment for the baseline value of the SF-12 physical function score was performed and compared with the same model without this adjustment (data not shown) because it is unclear whether analyses of change should be adjusted for baseline values (34). We conducted several sensitivity analyses to assess the robustness of the results. Thus, analyses were replicated in subgroups of individuals at high risk of impaired physical function: age ≥ 70 years, low physical activity, current or former smoking, body mass index ≥ 30 kg/m2, and important morbidity, compared to individuals at low risk. A possible modifying effect of sex was also tested. Differences between subgroups were tested by using likelihood-ratio tests, which compared models with and without cross-product interaction terms. We also repeated the analyses among participants who were not hospitalized during the follow-up to observe if the effect of diet on physical function would be stronger when participants with severe incident diseases were excluded. Finally, we examined the independent association of individual components of the Mediterranean diet scores with impaired physical function. Analyses were conducted using the SAS software, version 9.2 (SAS Institute). Results Supplementary Table 1 and Table 1 show the characteristics of study participants according to the sex-specific tertiles of the MEDAS and MDS scores. Compared to those in the lowest tertile of MDS, those in the highest tertile showed a lower frequency of current smoking, obesity, and diabetes. Those with a higher MEDAS score were also less often current smokers, showed less sedentary behavior, and had a lower frequency of obesity, diabetes, cognitive impairment, and depression. Table 1. Population Characteristics of the Study Participants Across the Tertiles of the Mediterranean Diet (N = 1,630)   MEDAS   Population Characteristics  Tertile 1  Tertile 2   Tertile 3   p Trend   Score, men/women  0–6/0–6  7–8/7  9–14/8–14    Participants, n  545  541  544    Age, y  67.9 ± 6.4  68.1 ± 6.0  68.2 ± 5.5  .34  Educational level, %   ≤Primary  50.1  50.3  50.2  .79   Secondary  25.9  25.9  27.0     University  24.0  23.8  22.8    Smoking status, %   Current smoker  14.9  13.6  9.0  .01   Former smoker  30.0  34.6  31.1     Never smoker  55.1  51.8  59.9    Leisure-time physical activity, MET-h/wk  21.1 ± 15.2  24.5 ± 15.8  22.9 ± 15.3  .06  Time spent watching TV, h/wk  17.9 ± 11.8  17.9 ± 10.5  16.4 ± 9.8  .03  Energy intake, kcal/d  2,117 ± 610  2,049 ± 553  2,009 ± 536  .002  BMI, kg/m2  28.4 ± 4.2  28.2 ± 3.7  27.7 ± 4.0  .004  Diagnosed diseases, %   Hypertension  63.5  68.7  61.7  .54   Diabetes  17.4  12.5  12.3  .02   Cognitive impairmenta  3.1  1.9  1.3  .03   Osteomuscular diseaseb  45.7  42.5  48.4  .30   Cardiovascular diseasec  4.8  5.4  4.8  .99   Cancer  2.4  3.3  3.1  .47   Chronic lung disease  9.9  8.1  9.4  .76   Depression  9.2  6.7  4.6  .003    MEDAS   Population Characteristics  Tertile 1  Tertile 2   Tertile 3   p Trend   Score, men/women  0–6/0–6  7–8/7  9–14/8–14    Participants, n  545  541  544    Age, y  67.9 ± 6.4  68.1 ± 6.0  68.2 ± 5.5  .34  Educational level, %   ≤Primary  50.1  50.3  50.2  .79   Secondary  25.9  25.9  27.0     University  24.0  23.8  22.8    Smoking status, %   Current smoker  14.9  13.6  9.0  .01   Former smoker  30.0  34.6  31.1     Never smoker  55.1  51.8  59.9    Leisure-time physical activity, MET-h/wk  21.1 ± 15.2  24.5 ± 15.8  22.9 ± 15.3  .06  Time spent watching TV, h/wk  17.9 ± 11.8  17.9 ± 10.5  16.4 ± 9.8  .03  Energy intake, kcal/d  2,117 ± 610  2,049 ± 553  2,009 ± 536  .002  BMI, kg/m2  28.4 ± 4.2  28.2 ± 3.7  27.7 ± 4.0  .004  Diagnosed diseases, %   Hypertension  63.5  68.7  61.7  .54   Diabetes  17.4  12.5  12.3  .02   Cognitive impairmenta  3.1  1.9  1.3  .03   Osteomuscular diseaseb  45.7  42.5  48.4  .30   Cardiovascular diseasec  4.8  5.4  4.8  .99   Cancer  2.4  3.3  3.1  .47   Chronic lung disease  9.9  8.1  9.4  .76   Depression  9.2  6.7  4.6  .003  Note: BMI = body mass index; MEDAS = Mediterranean Diet Adherence Screener; MET = metabolic equivalent; MMSE, Mini-Mental State Examination. For continuous variables, mean and standard deviation are reported. aCognitive impairment is defined as a MMSE score < 23. bOsteoarthritis, arthritis, and hip fracture. cIschemic heart disease, stroke, and heart failure. View Large Over a median follow-up of 3.5 years, we identified 663 individuals (40.7%) with incident physical impairment in any of the three domains considered: 408 (25.0%) had limitation in one domain, 161 (9.9%) in two domains, and 94 (5.8%) in the three of them. Specifically, 343 participants had incident impairment in agility, 212 had impaired mobility, and 457 showed a decline in overall physical function. We found no association between the MDS and the risk of impaired agility or mobility, although a 2-point increment in the MDS was marginally associated with lower likelihood for decreased overall physical function (Supplementary Table 2). By contrast, compared to persons in the lowest tertile of the MEDAS score, those in the highest tertile showed a lower odds of impaired agility, impaired mobility, and decreased overall physical functioning, in the fully adjusted analyses (Table 2). Results for impairment in overall physical functioning did not materially change when we removed the adjustment for the baseline SF-12 physical function score. A 2-point increment of the MEDAS score was associated with a 17% lower likelihood for impaired agility, a 21% lower likelihood for impaired mobility, a 17% lower likelihood for decreased overall function, and a 13% lower likelihood of impairment in any of the above domains. Repeating the analyses additionally adjusting for both physical activity and cognitive impairment did not materially change the results (Table 2). Neither did the results change when physical activity and cognitive impairment were entered separately (data not shown). Table 2. Odds Ratios (95% Confidence Interval) for the Association Between the Mediterranean Diet Adherence Screener and Physical Function Impairment During a 3.5-Year Follow-up of Older Adults (N = 1,630) Physical Function Impairment  MEDAS  Continuous per 2-Point Increment in the MEDAS  Tertile 1  Tertile 2  Tertile 3  p Trend  Impairment in agility, n  134  112  97    343   Model 1  1.00  0.98 (0.73; 1.33)  0.62 (0.46; 0.84)  .002  0.80 (0.70; 0.91)   Model 2  1.00  1.03 (0.75; 1.42)  0.69 (0.50; 0.96)  .03  0.84 (0.73; 0.97)   Model 3  1.00  1.02 (0.73; 1.42)  0.67 (0.48; 0.94)  .02  0.83 (0.71; 0.96)   Model 4  1.00  1.07 (0.76; 1.50)  0.70 (0.49; 0.98)  .04  0.84 (0.72; 0.98)  Impairment in mobility, n  89  64  59    212   Model 1  1.00  0.84 (0.59; 1.21)  0.59 (0.41; 0.84)  .004  0.78 (0.66; 0.92)   Model 2  1.00  0.86 (0.59; 1.25)  0.61 (0.42; 0.89)  .01  0.80 (0.67; 0.95)   Model 3  1.00  0.83 (0.56; 1.22)  0.69 (0.40; 0.88)  .01  0.79 (0.66; 0.94)   Model 4  1.00  0.87 (0.58; 1.29)  0.62 (0.42; 0.92)  .02  0.80 (0.67; 0.96)  Impairment in overall physical functioninga,b, n  177  153  127    457   Model 1  1.00  0.82 (0.63; 1.07)  0.60 (0.46; 0.79)  <.001  0.83 (0.73; 0.93)   Model 2  1.00  0.81 (0.62; 1.07)  0.61 (0.46; 0.81)  <.001  0.83 (0.74; 0.94)   Model 3  1.00  0.80 (0.60; 1.06)  0.60 (0.45; 0.79)  <.001  0.83 (0.73; 0.94)   Model 4  1.00  0.82 (0.62; 1.08)  0.61 (0.45; 0.81)  <.001  0.84 (0.74; 0.95)  Impairment in any of the above domains, n  246  215  202    663   Model 1  1.00  0.90 (0.70; 1.16)  0.69 (0.54; 0.89)  .004  0.85 (0.76; 0.95)   Model 2  1.00  0.91 (0.70; 1.18)  0.72 (0.55; 0.93)  .01  0.87 (0.77; 0.97)   Model 3  1.00  0.90 (0.69; 1.18)  0.70 (0.54; 0.92)  .01  0.87 (0.77; 0.97)   Model 4  1.00  0.92 (0.70; 1.20)  0.72 (0.55; 0.94)  .01  0.87 (0.77; 0.98)  Physical Function Impairment  MEDAS  Continuous per 2-Point Increment in the MEDAS  Tertile 1  Tertile 2  Tertile 3  p Trend  Impairment in agility, n  134  112  97    343   Model 1  1.00  0.98 (0.73; 1.33)  0.62 (0.46; 0.84)  .002  0.80 (0.70; 0.91)   Model 2  1.00  1.03 (0.75; 1.42)  0.69 (0.50; 0.96)  .03  0.84 (0.73; 0.97)   Model 3  1.00  1.02 (0.73; 1.42)  0.67 (0.48; 0.94)  .02  0.83 (0.71; 0.96)   Model 4  1.00  1.07 (0.76; 1.50)  0.70 (0.49; 0.98)  .04  0.84 (0.72; 0.98)  Impairment in mobility, n  89  64  59    212   Model 1  1.00  0.84 (0.59; 1.21)  0.59 (0.41; 0.84)  .004  0.78 (0.66; 0.92)   Model 2  1.00  0.86 (0.59; 1.25)  0.61 (0.42; 0.89)  .01  0.80 (0.67; 0.95)   Model 3  1.00  0.83 (0.56; 1.22)  0.69 (0.40; 0.88)  .01  0.79 (0.66; 0.94)   Model 4  1.00  0.87 (0.58; 1.29)  0.62 (0.42; 0.92)  .02  0.80 (0.67; 0.96)  Impairment in overall physical functioninga,b, n  177  153  127    457   Model 1  1.00  0.82 (0.63; 1.07)  0.60 (0.46; 0.79)  <.001  0.83 (0.73; 0.93)   Model 2  1.00  0.81 (0.62; 1.07)  0.61 (0.46; 0.81)  <.001  0.83 (0.74; 0.94)   Model 3  1.00  0.80 (0.60; 1.06)  0.60 (0.45; 0.79)  <.001  0.83 (0.73; 0.94)   Model 4  1.00  0.82 (0.62; 1.08)  0.61 (0.45; 0.81)  <.001  0.84 (0.74; 0.95)  Impairment in any of the above domains, n  246  215  202    663   Model 1  1.00  0.90 (0.70; 1.16)  0.69 (0.54; 0.89)  .004  0.85 (0.76; 0.95)   Model 2  1.00  0.91 (0.70; 1.18)  0.72 (0.55; 0.93)  .01  0.87 (0.77; 0.97)   Model 3  1.00  0.90 (0.69; 1.18)  0.70 (0.54; 0.92)  .01  0.87 (0.77; 0.97)   Model 4  1.00  0.92 (0.70; 1.20)  0.72 (0.55; 0.94)  .01  0.87 (0.77; 0.98)  Note: BMI = body mass index; MEDAS = Mediterranean Diet Adherence Screener; MET = metabolic equivalent; SF-12 = 12-Item Short-Form Health Survey. Model 1: logistic model adjusted for age and sex. Model 2: logistic model adjusted as in Model 1 and for educational level (≤primary, secondary, university), smoking status (never smoker, former smoker, current smoker), time spent watching television (quintiles of h/wk), energy intake (quintiles of kcal/d), and BMI. Model 3: logistic model adjusted as in Model 2 and for osteomuscular disease, cardiovascular disease, cancer, chronic lung disease, and depression requiring treatment. Model 4: logistic model adjusted as in Model 3 and for physical activity (quintiles of MET-h/wk) and cognitive impairment. a≥5-point decrease in the SF-12 physical component summary score from baseline to follow-up. bAdditionally adjusted for the SF-12 function score summary at baseline. View Large We performed sensitivity analyses in the oldest old individuals (≥70 years) and in subgroups of subjects at high risk of impaired physical function, compared with subjects at low risk (Supplementary Table 3). Although statistical significance was lost in some cases due to smaller sample size, the analyses consistently showed a tendency toward reduced likelihood of impaired physical function associated with a higher MEDAS score in both high risk and low risk individuals. We did not observe a sex interaction in any of the studied associations except for the MEDAS score and impaired mobility; the risk of mobility limitation associated with a 2-point increment in MEDAS score was stronger for men (odds ratio: 0.59, 95% confidence interval: 0.43; 0.81) than for women (odds ratio: 0.91, 95% confidence interval: 0.72; 1.14), although the direction remained the same. Moreover, when we removed from the analyses the individuals hospitalized during follow-up, no substantial change in the results was observed. Table 3 shows the association between each MEDAS component and physical function impairment. As expected, there was a tendency toward reduced likelihood of impaired physical function associated with many components of the diet; moreover, achievement of the target for nut consumption was significantly associated with lower odds of impaired agility and of decreased overall functioning. Also the achievement of target for fruits, fish or seafood, and preference of white over red meat was associated with a lower odds of decreased overall functioning. Table 3. Odds Ratios (95% Confidence Interval) for the Association Between Achievement of the Targets of the Mediterranean Diet Adherence Screener and Physical Function Impairment During a 3.5-Year Follow-up of Older Adults (N = 1,630) Targets of the MEDAS  Impairment in Agility  Impairment in Mobility  Impairment in Overall Physical Functioninga,b  Olive oil for cooking (yes)  1.45 (0.71; 2.95)  1.18 (0.52; 2.65)  1.87 (1.00; 3.48)  Olive oil (≥4 tablespoons/d)  0.64 (0.37; 1.11)  0.68 (0.34; 1.34)  0.80 (0.52; 1.22)  Vegetables (≥2 servings/d)  0.88 (0.59; 1.31)  0.98 (0.61; 1.56)  1.19 (0.86; 1.65)  Fruits (≥3 servings/d)  1.05 (0.76; 1.44)  0.78 (0.53; 1.13)  0.66 (0.49; 0.87)  Red meat, hamburger, or sausage (<1 serving/d)  1.10 (0.71; 1.71)  1.17 (0.68; 2.01)  0.95 (0.67; 1.36)  Butter, margarine, or cream (<1 serving/d)  0.91 (0.56; 1.48)  1.37 (0.76; 2.49)  1.36 (0.87; 2.12)  Sugar-sweetened or carbonated beverages (<1 time/d)  0.87 (0.52; 1.45)  1.19 (0.63; 2.24)  0.88 (0.57; 1.37)  Wine (≥1 serving/d)  1.02 (0.73; 1.44)  0.78 (0.51; 1.19)  0.96 (0.73; 1.27)  Legumes (≥3 times/wk)  1.02 (0.71; 1.47)  0.70 (0.44; 1.12)  0.86 (0.63; 1.17)  Fish or seafood (≥3 times/wk)  0.82 (0.61; 1.09)  0.76 (0.55; 1.06)  0.78 (0.61; 1.00)  Commercial baked goods (<2 times/wk)  0.89 (0.67; 1.18)  0.84 (0.60; 1.17)  0.96 (0.75; 1.22)  Nuts (≥3 times/wk)  0.68 (0.49; 0.94)  0.71 (0.48; 1.05)  0.72 (0.54; 0.95)  Preference for white over red meat  0.92 (0.67; 1.26)  1.10 (0.76; 1.58)  0.73 (0.55; 0.96)  Foods with “sofrito” (≥2 times/wk)  0.74 (0.43; 1.29)  0.75 (0.40; 1.41)  0.79 (0.50; 1.27)  Targets of the MEDAS  Impairment in Agility  Impairment in Mobility  Impairment in Overall Physical Functioninga,b  Olive oil for cooking (yes)  1.45 (0.71; 2.95)  1.18 (0.52; 2.65)  1.87 (1.00; 3.48)  Olive oil (≥4 tablespoons/d)  0.64 (0.37; 1.11)  0.68 (0.34; 1.34)  0.80 (0.52; 1.22)  Vegetables (≥2 servings/d)  0.88 (0.59; 1.31)  0.98 (0.61; 1.56)  1.19 (0.86; 1.65)  Fruits (≥3 servings/d)  1.05 (0.76; 1.44)  0.78 (0.53; 1.13)  0.66 (0.49; 0.87)  Red meat, hamburger, or sausage (<1 serving/d)  1.10 (0.71; 1.71)  1.17 (0.68; 2.01)  0.95 (0.67; 1.36)  Butter, margarine, or cream (<1 serving/d)  0.91 (0.56; 1.48)  1.37 (0.76; 2.49)  1.36 (0.87; 2.12)  Sugar-sweetened or carbonated beverages (<1 time/d)  0.87 (0.52; 1.45)  1.19 (0.63; 2.24)  0.88 (0.57; 1.37)  Wine (≥1 serving/d)  1.02 (0.73; 1.44)  0.78 (0.51; 1.19)  0.96 (0.73; 1.27)  Legumes (≥3 times/wk)  1.02 (0.71; 1.47)  0.70 (0.44; 1.12)  0.86 (0.63; 1.17)  Fish or seafood (≥3 times/wk)  0.82 (0.61; 1.09)  0.76 (0.55; 1.06)  0.78 (0.61; 1.00)  Commercial baked goods (<2 times/wk)  0.89 (0.67; 1.18)  0.84 (0.60; 1.17)  0.96 (0.75; 1.22)  Nuts (≥3 times/wk)  0.68 (0.49; 0.94)  0.71 (0.48; 1.05)  0.72 (0.54; 0.95)  Preference for white over red meat  0.92 (0.67; 1.26)  1.10 (0.76; 1.58)  0.73 (0.55; 0.96)  Foods with “sofrito” (≥2 times/wk)  0.74 (0.43; 1.29)  0.75 (0.40; 1.41)  0.79 (0.50; 1.27)  Note: MEDAS = Mediterranean Diet Adherence Screener; SF-12 = 12-Item Short-Form Health Survey. Analysis adjusted as in Model 3 in Tables 2 and 3. a≥5-point decrease in the SF-12 physical component summary score from baseline to follow-up. bAdditionally adjusted for the SF-12 physical function score at baseline. View Large Discussion In this study, among older adults, a higher accordance with a Mediterranean dietary pattern, as measured by the MEDAS score, was associated with a lower likelihood of impairment in agility, mobility, and overall physical functioning. By contrast, accordance with the MDS was only associated with overall physical functioning. Our finding of better functioning associated with the MEDAS but to a much lesser extent, the MDS highlights the complexity of defining a universal Mediterranean dietary pattern. Because operational definitions of this diet have been derived from specific Mediterranean regions and time periods, many different scores have been used in the literature, and there is an ongoing debate about which components should be considered in each score (35), including the type of fat, specification of grain type, definition of moderate alcohol intake, and the presence of dairy products, nuts, and fish. For our analyses, we chose one of the most widely used sample-based scores (MDS) and the normative score (MEDAS) developed for Spain. Although both patterns focused on the consumption of fruits, vegetables, fish, and unsaturated fat, they differ in the weight assigned to nuts, olive oil, and white meat, and also in the items considered unhealthy (dairy and all types of meat in the MDS vs red meat, butter, sugar-sweetened beverages, and baked goods in the MEDAS). In addition, the number of components of each score (9 in the MDS vs 14 in the MEDAS) and the method of scoring (data-driven vs arbitrary cutoff values) may also have influenced the results. In addition, the number of components and the use of cutoff values in the MEDAS resulted in larger variations in scoring across the study participants and, thus, may have provided greater power to observe an association. To our knowledge, a few population studies have investigated the association between the Mediterranean diet and impaired physical function. Among the cross-sectional analyses, in the Nurses’ Health Study, a North American cohort, a higher adherence to the MDS was related to a greater likelihood of no major limitations in physical function, measured based on the mobility questions from the SF-36 (18). Also, in another Spanish population, higher adherence to the MDS was associated with higher scores in the physical component of the SF-12 in men, but not in women (17). Moreover, in participants from the U.S. NHANES and Israeli MABAT ZAHAV studies, a higher adherence to the MDS was associated with significant higher walking speed. This association was attenuated after adjustment for cognitive decline (19). Finally, a study with an Australian cohort found an association between higher adherence to the MDS and better scores in the physical function score of the SF-12 (20), which is in line with our study. In the InCHIANTI cohort, a prospective longitudinal study that assessed this association among 935 Italian participants, the authors found that higher adherence to the MDS was associated with better lower body performance after 9 years of follow-up (21). Finally, in the Health ABC cohort, with North American participants, walking speed over 8 years was faster among those with higher adherence to a Mediterranean diet score (22). As regards the association between the Mediterranean diet and disability, which is a common consequence of impaired physical function, a prospective analysis of the French Three City Study found that women in the highest category of the MDS had a 50% risk reduction in incident disability in activities of daily living, compared to women in the lowest category; however, no association was found for men (36). In a previous analysis of the Seniors-ENRICA cohort, we also found that a higher score on either the MDS or the MEDAS was associated with lower risk of the frailty syndrome, defined as having three out of the following five criteria: exhaustion, weight loss, low grip strength, slow walking speed, or low physical activity (37); however, when weight loss was excluded from the frailty definition, only the MEDAS score was significantly associated with a lower risk of frailty. Our present results suggest that the beneficial effect of the Mediterranean diet on disability and frailty in older adults could be mediated through better physical functioning. In this study, impairment in overall physical functioning was defined as an at least 5-point decrease in the physical function score of the SF-12 from baseline to follow-up. In the IQOLA project, which examined the impact of chronic conditions (arthritis, chronic lung disease, and congestive heart failure) on quality of life in 15 countries, those conditions with a major impact on quality of life reduced the score on the physical component of the SF-12 less than 5 points (38). Therefore, our definition of impaired functioning seems appropriate and of clinical relevance. Our main findings were robust. The inverse associations between the MEDAS score and impaired agility or mobility were similar or slightly stronger in subgroups of participants defined by their higher risk of functional limitations. In addition, when the analyses were adjusted for physical activity and cognitive impairment, which are closely related to the outcome, the main results still held. Also, adjusting for chronic diseases, which might lie in the causal pathway, did not materially change the results. Several mechanisms, such as reduced oxidation and inflammation, might play a role in the association between Mediterranean diet and physical function. Specifically, accordance with the Mediterranean diet has been associated with decreased oxidative damage and lower serum levels of inflammatory markers (39,40). Also, augmented oxidative damage and inflammatory markers have previously been found to predict the onset of physical limitations (41,42). In addition, homocysteine, a sensitive marker for a deficiency in B vitamins, such as folic acid, has been shown to be independently associated with physical function (14). Thus, B vitamins, provided by the Mediterranean diet, may influence physical function through homocysteine levels. Strengths of this study are its prospective design, collection of food consumption with a validated diet history, and adjustment for many potential confounders. Some limitations should also be acknowledged. Equal weights were assigned to each component of the Mediterranean diet scores even though their health effects might be different. Also, certain misclassification of food consumption cannot be ruled out, despite excluding participants with an implausibly high or low energy intake. Another limitation was the use of self-reported information as a proxy for mobility, agility, and overall physical performance; however, these measurements have been widely used in clinical practice and in population studies. There were a high percentage of persons excluded from the analyses because many participants did not complete the Mini-Mental State Examination. Although results were similar when we included persons without this information, we still decided to exclude participants with unknown cognitive status from the analytical sample because the impact of cognitive function on physical function is unclear. Therefore, the studied participants are not a reflection of the national structure of the Spanish older population. Moreover, functional impairment was evaluated at the end of the follow-up, so that temporality and development of impairments during the interval period could not be fully ascertained. Finally, as in any observational study, some residual confounding may persist. In conclusion, a Mediterranean-style diet was associated with a lower likelihood of impaired physical function in Spanish older adults. The weaker association of the MDS with this outcome suggests that a more precise definition of this diet is needed to be able to find consistent health effects. Supplementary Material Please visit the article online at http://gerontologist.oxfordjournals.org/ to view supplementary material. Funding This work was supported by grants from the Instituto de Salud Carlos III, State Secretary of R+D+I of Spain, and FEDER/FSE (FIS 12/1166 and 13/0288) and the European Union (FP7-HEALTH-2012 Proposal No 305483-2, FRAILOMIC Initiative). Conflict of Interest None of the authors has a conflict of interest related to the subject of this work. References 1. Wang H Dwyer-Lindgren L Lofgren KTet al.  . Age-specific and sex-specific mortality in 187 countries, 1970-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet . 2012; 380: 2071– 2094. doi: 10.1016/S0140-6736(12)61719-X Google Scholar CrossRef Search ADS PubMed  2. Murray CJ Vos T Lozano Ret al.  . Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions, 1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. Lancet . 2012; 380: 2197– 2223. doi: 10.1016/S0140-6736(12)61689-4 Google Scholar CrossRef Search ADS PubMed  3. Chatterji S Byles J Cutler D Seeman T Verdes E. Health, functioning, and disability in older adults-present status and future implications. Lancet . 2014; 386: 563– 565. doi: 10.1016/S0140-6736(14)61462-8 4. Verbrugge LM Jette AM. The disablement process. Soc Sci Med . 1994; 38: 1– 14. Google Scholar CrossRef Search ADS PubMed  5. Jette AM. Toward a common language for function, disability, and health. Phys Ther . 2006; 86: 726– 734. Google Scholar CrossRef Search ADS PubMed  6. Fried LP Ferrucci L Darer J Williamson JD Anderson G. Untangling the concepts of disability, frailty, and comorbidity: implications for improved targeting and care. J Gerontol A Biol Sci Med Sci . 2004; 59: 255– 263. Google Scholar CrossRef Search ADS PubMed  7. Studenski S Perera S Patel Ket al.  . Gait speed and survival in older adults. JAMA . 2011; 305: 50– 58. doi: 10.1001/jama.2010.1923 Google Scholar CrossRef Search ADS PubMed  8. Guralnik JM Simonsick EM Ferrucci Let al.  . A short physical performance battery assessing lower extremity function: association with self-reported disability and prediction of mortality and nursing home admission. J Gerontol . 1994; 49: M85– M94. Google Scholar CrossRef Search ADS PubMed  9. Guallar-Castillón P Sagardui-Villamor J Banegas JRet al.  . Waist circumference as a predictor of disability among older adults. Obesity . 2007; 15: 233. doi: 10.1038/oby.2007.532 Google Scholar CrossRef Search ADS PubMed  10. Artaud F Dugravot A Sabia S Singh-Manoux A Tzourio C Elbaz A. Unhealthy behaviours and disability in older adults: three-City Dijon cohort study. BMJ . 2013; 347: f4240. doi: 10.1136/bmj.f4240 Google Scholar CrossRef Search ADS PubMed  11. Vercambre MN Boutron-Ruault MC Ritchie K Clavel-Chapelon F Berr C. Long-term association of food and nutrient intakes with cognitive and functional decline: a 13-year follow-up study of elderly French women. Br J Nutr . 2009; 102: 419– 427. doi: 10.1017/S0007114508201959 Google Scholar CrossRef Search ADS PubMed  12. Houston DK Stevens J Cai J Haines PS. Dairy, fruit, and vegetable intakes and functional limitations and disability in a biracial cohort: the Atherosclerosis Risk in Communities Study. Am J Clin Nutr . 2005; 81: 515– 522. doi: 10.1016/s1073-5437(08)70273-0 Google Scholar CrossRef Search ADS PubMed  13. Cesari M Pahor M Bartali Bet al.  . Antioxidants and physical performance in elderly persons: the Invecchiare in Chianti (InCHIANTI) study. Am J Clin Nutr . 2004; 79: 289– 294. Google Scholar CrossRef Search ADS PubMed  14. Ng TP Aung KC Feng L Scherer SC Yap KB. Homocysteine, folate, vitamin B-12, and physical function in older adults: cross-sectional findings from the Singapore Longitudinal Ageing Study. Am J Clin Nutr . 2012; 96: 1362– 1368. doi: 10.3945/ajcn.112.035741 Google Scholar CrossRef Search ADS PubMed  15. Hu FB. Dietary pattern analysis: a new direction in nutritional epidemiology. Curr Opin Lipidol . 2002; 13: 3– 9. Google Scholar CrossRef Search ADS PubMed  16. Xu B Houston DK Locher JLet al.  . Higher Healthy Eating Index-2005 scores are associated with better physical performance. J Gerontol A Biol Sci Med Sci . 2012; 67: 93– 99. doi: 10.1093/gerona/glr159 Google Scholar CrossRef Search ADS PubMed  17. Muñoz MA Fíto M Marrugat J Covas MI Schröder H; REGICOR and HERMES Investigators. Adherence to the Mediterranean diet is associated with better mental and physical health. Br J Nutr . 2009; 101: 1821– 1827. doi: 10.1017/S0007114508143598 Google Scholar CrossRef Search ADS PubMed  18. Samieri C Sun Q Townsend MKet al.  . The association between dietary patterns at midlife and health in aging: an observational study. Ann Intern Med . 2013; 159: 584– 591. doi: 10.7326/0003-4819-159-9-201311050-00004 Google Scholar CrossRef Search ADS PubMed  19. Zbeida M Goldsmith R Shimony T Vardi H Naggan L Shahar DR. Mediterranean diet and functional indicators among older adults in non-Mediterranean and Mediterranean countries. J Nutr Health Aging . 2014; 18: 411– 418. doi: 10.1007/s12603-014-0003-9 Google Scholar CrossRef Search ADS PubMed  20. Milte CM Thorpe MG Crawford D Ball K McNaughton SA. Associations of diet quality with health-related quality of life in older Australian men and women. Exp Gerontol . 2015; 64: 8– 16. doi: 10.1016/j.exger.2015.01.047 Google Scholar CrossRef Search ADS PubMed  21. Milaneschi Y Bandinelli S Corsi AMet al.  . Mediterranean diet and mobility decline in older persons. Exp Gerontol . 2011; 46: 303– 308. doi: 10.1016/j.exger.2010.11.030 Google Scholar CrossRef Search ADS PubMed  22. Shahar DR Houston DK Hue TFet al.  . Adherence to Mediterranean diet and decline in walking speed over 8 years in community-dwelling older adults. J Am Geriatr Soc . 2012; 60: 1881– 1888. doi: 10.1111/j.1532-5415.2012.04167.x Google Scholar CrossRef Search ADS PubMed  23. Rodríguez-Artalejo F Graciani A Guallar-Castillón Pet al.  . Rationale and methods of the study on nutrition and cardiovascular risk in Spain (ENRICA). Rev Esp Cardiol . 2011; 64: 876– 882. doi: 10.1016/j.recesp.2011.05.019 Google Scholar CrossRef Search ADS PubMed  24. León-Muñoz LM García-Esquinas E López-García E Banegas JR Rodríguez-Artalejo F. Major dietary patterns and risk of frailty in older adults: a prospective cohort study. BMC Med . 2015; 13: 11. doi: 10.1186/s12916-014-0255-6 Google Scholar CrossRef Search ADS PubMed  25. Guallar-Castillón P Sagardui-Villamor J Balboa-Castillo Tet al.  . Validity and reproducibility of a Spanish dietary history. PLoS One . 2014; 9: e86074. doi: 10.1371/journal.pone.0086074 Google Scholar CrossRef Search ADS PubMed  26. Trichopoulou A Kouris-Blazos A Wahlqvist MLet al.  . Diet and overall survival in elderly people. BMJ . 1995; 311: 1457– 1460. Google Scholar CrossRef Search ADS PubMed  27. 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– 2608. doi: 10.1056/NEJMoa025039 Google Scholar CrossRef Search ADS PubMed  28. Struijk EA Beulens JW May AMet al.  . Dietary patterns in relation to disease burden expressed in disability-adjusted life years. Am J Clin Nutr . 2014; 100: 1158– 1165. doi: 10.3945/ajcn.113.082032 Google Scholar CrossRef Search ADS PubMed  29. Schröder H Fitó M Estruch Ret al.  . A short screener is valid for assessing Mediterranean diet adherence among older Spanish men and women. J Nutr . 2011; 141: 1140– 1145. doi: 10.3945/jn.110.135566 Google Scholar CrossRef Search ADS PubMed  30. Rosow I Breslau N. A Guttman health scale for the aged. J Gerontol . 1966; 21: 556– 559. Google Scholar CrossRef Search ADS PubMed  31. Vilagut G Valderas JM Ferrer M Garin O López-García E Alonso J. Interpretation of SF-36 and SF-12 questionnaires in Spain: physical and mental components. Med Clin . 2008; 130: 726– 735. Google Scholar CrossRef Search ADS   32. Pols MA Peeters PH Ocké MC Slimani N Bueno-de-Mesquita HB Collette HJ. Estimation of reproducibility and relative validity of the questions included in the EPIC Physical Activity Questionnaire. Int J Epidemiol . 1997; 26( suppl 1): S181– S189. Google Scholar CrossRef Search ADS PubMed  33. Graciani A Banegas JR Guallar-Castillón P Domínguez-Rojas V Rodríguez-Artalejo F. Cognitive assessment of the non-demented elderly community dwellers in Spain. Dement Geriatr Cogn Disord . 2006; 21: 104– 112. doi: 10.1159/000090509 Google Scholar CrossRef Search ADS PubMed  34. Glymour MM Weuve J Berkman LF Kawachi I Robins JM. When is baseline adjustment useful in analyses of change? An example with education and cognitive change. Am J Epidemiol . 2005; 162: 267– 278. doi: 10.1093/aje/kwi187 Google Scholar CrossRef Search ADS PubMed  35. Bach A Serra-Majem L Carrasco JLet al.  . The use of indexes evaluating the adherence to the Mediterranean diet in epidemiological studies: a review. Public Health Nutr . 2006; 9: 132– 146. Google Scholar CrossRef Search ADS PubMed  36. Féart C Pérès K Samieri C Letenneur L Dartigues JF Barberger-Gateau P. Adherence to a Mediterranean diet and onset of disability in older persons. Eur J Epidemiol . 2011; 26: 747– 756. doi: 10.1007/s10654-011-9611-4 Google Scholar CrossRef Search ADS PubMed  37. León-Muñoz LM Guallar-Castillón P López-García E Rodríguez-Artalejo F. Mediterranean diet and risk of frailty in community-dwelling older adults. J Am Med Dir Assoc . 2014; 15: 899– 903. doi: 10.1016/j.jamda.2014.06.013. Google Scholar CrossRef Search ADS PubMed  38. Alonso J Ferrer M Gandek Bet al.  ; IQOLA Project Group. Health-related quality of life associated with chronic conditions in eight countries: results from the International Quality of Life Assessment (IQOLA) Project. Qual Life Res . 2004; 13: 283– 298. Google Scholar CrossRef Search ADS PubMed  39. Fitó M Guxens M Corella Det al.  ; PREDIMED Study Investigators. Effect of a traditional Mediterranean diet on lipoprotein oxidation: a randomized controlled trial. Arch Intern Med . 2007; 167: 1195– 1203. doi: 10.1001/archinte.167.11.1195 Google Scholar CrossRef Search ADS PubMed  40. Esposito K Marfella R Ciotola Met al.  . Effect of a Mediterranean-style diet on endothelial dysfunction and markers of vascular inflammation in the metabolic syndrome: a randomized trial. JAMA . 2004; 292: 1440– 1446. doi: 10.1001/jama.292.12.1440 Google Scholar CrossRef Search ADS PubMed  41. Cesari M Kritchevsky SB Nicklas BJet al.  . Lipoprotein peroxidation and mobility limitation: results from the Health, Aging, and Body Composition Study. Arch Intern Med . 2005; 165: 2148– 2154. doi: 10.1001/archinte.165.18.2148 Google Scholar CrossRef Search ADS PubMed  42. Penninx BW Kritchevsky SB Newman ABet al.  . Inflammatory markers and incident mobility limitation in the elderly. J Am Geriatr Soc . 2004; 52: 1105– 1113. doi: 10.1111/j.1532-5415.2004.52308.x Google Scholar CrossRef Search ADS PubMed  © The Author(s) 2016. Published by Oxford University Press on behalf of The Gerontological Society of America. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com.

Journal

The Journals of Gerontology Series A: Biomedical Sciences and Medical SciencesOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off