Adherence to the Mediterranean Diet and the Risk of Frailty in Old People: A Systematic Review and Meta-Analysis

Abstract

Background

Frailty is a common geriatric syndrome in old people. It remains controversial whether Mediterranean diet could prevent old people from developing into frailty. The aim of this study is to summarize the relevant studies and assess the effectiveness of adherence to Mediterranean diet on frailty in old people.

Method

A systematic search of MEDLINE, EMBASE and Cochrane Central Register of Controlled Trials was conducted to identify all relevant studies up to Oct 2017. We included studies regarding the associations between adherence to Mediterranean diet and risk of frailty among elders. A meta-analysis was performed to explore the effects of Mediterranean diet on frailty.

Results

Six studies matched the inclusion criteria, of which five were prospective and one was cross-sectional. A total of 10,210 participants from the five prospective cohort studies were included to perform the meta-analyses. In comparison with lowest adherence to Mediterranean diet, elders with highest adherence to Mediterranean diet were significantly associated with lower risk of frailty in the future (RR= 0.56, 95% CI=0.36-0.89, p=0.015). Furthermore, the pooled estimates from four studies performed among participants in western countries (European and North American) showed that higher adherence to Mediterranean diet was associated with a 52% reduced risk of frailty (RR= 0.48, 95% CI=0.32-0.72, p<0.001). However, one study showed no association between Mediterranean diet and frailty among Asian elders (RR=1.06, 95% CI=0.83-1.36, p=0.638).

Conclusion

A higher adherence to Mediterranean diet is associated with a lower risk of frailty in old people. Meanwhile, the benefits may be more obvious among elders from western countries.

Introduction

Frailty refers to a state of decreased physiologic reserve and resistance to stressors, resulting from an age-related cumulative decline in many systems during a lifetime (1), which is a common geriatric syndrome in old people and displaying a soundless epidemic event in primary care (2). Frailty is a strong predictor of a variety of adverse outcomes including falls, disability, cognitive decline, hospitalization, institutionalization and even mortality (3). Moreover, frailty also increases healthcare costs and the financial burden to families and societies (4), and becoming an emerging public health priority (5). As population aging, a higher prevalence of frailty will appear and these situations will be worsen.

In fact, frailty is a potentially reversible and transitional state in the dynamic progression from robustness to functional decline (6). Therefore, interventions for frailty could be practical, have profound benefits of improving the quality of life of elderly people and reduce expenses for their care. Up to now, various intervention measures for frailty have been explored, among which the effects of nutritional supplement, especially for the protein intake and certain micronutrients, on the functional improvement of frail elderly people are still controversial due to a healthy diet habit combined with a physically active lifestyle is the basis of successful ageing (7). The investigation of dietary patterns can get the overall influence of diet, synergy of various nutrients, on frailty and overcome the limitations of single nutrient supplement (8). Previous studies have explored the impacts of some dietary patterns on frailty (9, 10) and Mediterranean diet (MD) has received wide attention (11, 12, 13). MD is one of the most important healthy dietary pattern in the world, characterized by a high amounts of olive oil and olives, fruits, vegetables, unrefined cereals, legumes, and nuts, moderate amounts of fish and dairy products and low quantities of meat and meat products, and often wine in moderation if possible (14). It was reported that MD was related to improvement of grip strength (15), decreased decline of walking speed (16), lower odds of sarcopenia (17) and fracture (18) among elders, all of which indicated the protective role of MD on frailty. Recently, an eight-year follow-up cohort study showed higher adherence to MD can lower the risk of developing frailty in future (11). However, another cohort study on 2724 elders failed to find any significant associations between MD and frailty (19). Considering the controversial conclusions and lack of any summary of these studies in this respect, we conducted this systematic review and meta-analysis to assess the effectiveness of adherence to MD on frailty in old people.

Methods

This systematic review was performed in accordance with PRISMA (Preferred Reporting Items for Systematic Review and Meta-Analyses) statement (20).

Literature search

A thorough search using multiple electronic database up to Oct 2017 was conducted, including MEDLINE, EMBASE and Cochrane Central Register of Controlled Trials. The search strategies conformed to the requirements of specific database. The following key words were used: frail elderly, frailty, Mediterranean, diet, besides a range of synonyms and related words were also applicable (Supplementary 1). No limitation was applied during the search. In order to avoid overlooking additional related studies, the references of the selected articles were reviewed and grey literatures (http://www.opengrey.eu/) were searched.

Study selection

Two reviewers (LS, YL) independently selected relevant articles and assessed their eligibility according to the inclusion and exclusion criteria. Disagreements were solved by discussion. Literatures that met the following criteria were eligible for inclusion: 1), all original studies conducted on humans that investigated the relationship between MD and risk of frailty; 2) the participants were with an average age of 60 or above; 3) given that there are still wide debates on operative definition of frailty (21), in this study, we included studies reported the outcomes involved frailty defined by any validated assessment methods, such as frailty phenotype, Rockwood Frailty Index, FRAIL scale, the Study of Osteoporotic Fractures scale. Frailty phenotype defines frailty as meeting three or more of the five criteria—weight loss, exhaustion, low energy expenditure, slowness, and weakness. Based on frailty phenotype, researchers developed some simple assessment methods such as FRAIL scale and Study of Osteoporotic Fractures scale. Rockwood Frailty Index describes frailty as a state caused by the accumulation of health deficits during the life course and frailty can be measured by counting the number of deficits accumulated. These tools were most commonly used in previous studies (22). 4) Various tools can be used to measure the degree of adherence to the Mediterranean diet. In this review, studies with a validated evaluation for adherence to a Mediterranean-style diet were included, such as Mediterranean Diet Score (MDS) constructed by Trichopoulou et al (14) and Mediterranean diet score (aMED) proposed by Panagiotakos et al (23). A value of 0 or 1 is assigned to each component in MDS using sex-specific median values as cutoff and the total MDS ranged from 0 (minimal adherence to a MD) to 9 (maximal adherence to a MD). While calculating aMED, participants are asked to rate their intakes of each food item using a score ranging from 0 (less adherence) to 5 (better adherence) and the total possible score ranges from 0 to 55, with higher values representing higher adherence to a MD.

Letters, comments, reviews, meta-analyses, ecological studies were excluded and we also excluded studies meeting the exclusion criteria: 1) did not have a clear definition of frailty in the study; (2) studies that examined dietary patterns but without a definition of MD; (3) studies that examined food items rather than the whole MD.

Quality Assessment

The Newcastle-Ottawa Scale (NOS) was used to assess the quality of selected observational studies (24). The NOS assigns a maximum of 9 points based on three quality parameters: selection, comparability and outcome (24).

Data Abstraction

The following information was extracted from each included study: 1), the basic information of the articles, i.e. authors, publication year; 2), country; 3), study design; 4), the number of participants; 5), participants characteristics, i.e. mean age, number of women; 6), Follow-up period; 7), the method used to collect diet information; 8), assessment of adherence to MD; 9), assessment of frailty; 10), main outcomes.

Statistical analysis

The STATA 11.0 (Stata Corp, College Station, TX, USA) was used for all analyses. To compute a pooled RR with 95% CI, we used the adjusted ORs, or hazard risks and their 95% CIs of the highest adherence to MD compared with the lowest adherence to MD [25]. Random effects model was used to calculate pooled effect estimates. Homogeneity was tested with Cochran's Q statistic using chi-square and I-squared. I2 value of 50 % or below represented acceptable heterogeneity. A funnel plot and Beggar's test was used to explore publication bias if 10 or above studies were included. Sensitivity analysis was performed to evaluate the impact of individual studies on the overall results. We also conducted subgroup analyses according to different population and frailty criteria.

Results

Search results

Figure 1 summarized the process for literature selection. The systematic search yielded 319 citations. We removed 126 duplicates and excluded 173 citations through title and abstract screening. After that, 20 studies were left for full-text review, of which 14 studies were excluded because five articles reported physical function but without a definition for frailty, seven articles were without a description for MD, one was repeated research report and one was review. The remaining six studies matched the inclusion criteria. No additional results were found through references screening and grey literatures searching.

Figure 1.

Preferred Reporting Items for Systematic Reviews and Meta- Analyses (PRISMA) flow diagram for the study selection process

Characteristics of studies

All studies were published between 2012 and 2017. Four studies were conducted in European countries (12, 13, 26, 27), one in China (19) and one in North American (11). The six studies included five prospective cohort studies (11, 12, 13, 19, 27)and one cross-sectional study (26). Duration of follow up among the prospective cohort studies was between two to eight years.

Mean age of the included studies ranges from 61.2 to 83 years old and all were conducted in both genders (Table 1). Five studies used Mediterranean diet score (MDS) developed by Trichopoulou et al (12, 13, 19, 26, 27) to assess the adherence to MD among the subjects and one study used a different score of Mediterranean diet (aMED) proposed by Panagiotakos et al (11). Most of the studies (11, 12, 13, 19, 26)used validated food frequency questionnaire (FFQ) to collect data of food consumption except the study from León-Muñoz et al (27) (which used a validated computerized diet history in the preceding year (Table 1). In terms of the assessment of frailty, four studies used Fried criteria (12, 13, 26, 27), one study used FRAIL scale (19) and one study used the SOF (Study of Osteoporotic Fracture) index (11) (Table 1).

Table 1.

Characteristics of selected studies that reported the relationship between adherence to Mediterranean diet and frailty

Author/Origin publication year	Study Design, N, women (%)	Country	Mean age† (years)	Duration follow-up(years)	Frailty assessment	Collection of food consumption	Assessment of adherence to MD	Adjusted HR/OR (95%CI)	Adjustment	Quality assessment
Talegawkar et al. /2012 (13)	Cohort study 690, 51.7%	Italy	73.0	6	Fried criteria	FFQ	MDS	0.30 (0.14, 0.66)	Age, sex, energy intake, status of frailty at previous examinations, BMI, education, MMSE score, current smoker, and presence of chronic diseases.	7
León-Muñoz et al /2014 (27)	Cohort study 1815, not clear	Spain	68	3.5	Fried criteria	1-year diet history	MDS	0.48 (0.30,0.77)	Age, sex, educational level, tobacco, BMI, energy intake, cardiovascular disease, diabetes mellitus, cancer, asthma or chronic bronchitis, osteo-muscular disease, depression requiring treatment, and number of drug treatments.	8
Chan et al /2015 (19)	Cohort Study 2724 , 50.3%	China	71.8	4	FRAIL Scale	FFQ	MDS	1.06 (0.83–1.36)	Age, sex, BMI, energy intake, PASE, education level, smoking status, alcohol use, GDS category, CSID category, living alone and marital status at baseline.	8
Rahi et al /2017 (12)	Cohort study 560 62.5%	France	81.7	2	Fried criteria	FFQ	MDS	0.32 (0.14, 0.72)	Age, sex, marital status, education, BMI, diabetes, cardiovascular history, taking ≥ 5 drugs/d, hypertension, MMSE score and CES-D Scale.	8
Veronese et al /2017 (11)	Cohort study 4421 62.5%	North American	61.2	8	SOF index	FFQ	aMED	0.71 (0.50, 0.99)	Age, sex, race, BMI, education, smoking habits, yearly income, PASE score, Charlson comorbidity index, daily energy intake.	8
Bollwein et al /2013 (26)	Cross-sectional study 192, 64.6%	German	83	–	Fried criteria	FFQ	MDS	0.19 (0.05–0.82)	Age, sex, energy intake, comorbidity, and educational level.	6

Note: HR: hazard ratio; OR: odds ratio; MDS: Mediterranean Diet Score developed by Trichopoulou et al; aMED: Mediterranean diet score (aMED) proposed by Panagiotakos et al; FFQ: food frequency questionnaire; SOF: Study of Osteoporotic Fracture; BMI: body mass index; MMSE: mini-mental State examination; PASE: Physical Activity Scale of the Elderly; GDS: Geriatric Depression Scale; CSID: Cognitive Screening Instrument for Dementia; CES-D: Center for Epidemiologic Studies Depression; † : The highest adherence to MD compared with the lowest adherence to MD.

Mediterranean diet and risk of frailty

Higher adherence to MD was associated with a reduce risk of frailty in five studies (11-13, 26, 27), however, one prospective study (19) failed to found any significant associations between MD and frailty. All the five prospective studies were considered to be adequate for meta-analysis after quality assessment (Table 1) (11, 12, 13, 19, 27). A total of 10,210 elders were included in the analysis of RR for risk of frailty. In comparison with lowest adherence to MD, elders with highest adherence to MD were significantly associated with lower risk of frailty in the future (RR= 0.56, 95% CI=0.36-0.89, p=0.015, Fig. 2) The level of heterogeneity between these studies was high-I² =80.6% P<0.001, Fig. 2). A significant reduced heterogeneity was observed after performing a subgroup analysis by population (European and American vs. Asian) and frailty criteria (Fig. 2, Fig. 3). The pooled estimates from four studies performed among participants in western countries (European and North American) (11, 12, 13, 27) showed that higher adherence to MD was associated with a 52% reduced risk of frailty (RR= 0.48, 95% CI=0.32-0.72, p<0.001, Fig. 2). The level of heterogeneity between these studies was moderate (I² = 53.7%, P= 0.090 Fig. 2). However, one study showed no association between MD and frailty among Asian elders (RR=1.06, 95% CI=0.83-1.36, p=0.638, Fig.2)[19]. Furthermore, pooled estimates from three studies with a same definition for frailty (Fried criteria) (12, 13, 27) also showed the protective effects of adherence to MD on frailty (RR= 0.40, 95% CI=0.28- 0.57, p<0.001, Fig. 3) .The level of heterogeneity between these studies was low (I² =0.0%, P= 0.499, Fig. 3). We then conducted a sensitivity analysis to detect the influence of each study on the pooled estimates for the outcomes. After the exclusion of studies conducted by Talegawkar et al. (13) and Rahi et al (12), the results showed a similar trend but no statistical significance was shown (Fig. 4).

Figure 2.

Forest plot of the RR estimates for the association between MD and frailty, and subgroup analysis according to the population

Figure 3.

Subgroup analysis according to different definition for frailty

Figure 4.

Sensitivity analysis of all studies Our study searched all relevant studies and provided

Discussion

Among several factors that might influence the frail status, dietary intakes are of great importance. Previous studies have confirmed that specific dietary pattern can significantly affect frailty status among elders (9). Therefore, to find out a healthy diet that can prevent elders from developing into frailty could provide valuable clue to intervene frail elderly more efficiently. To the best of our knowledge, this is the first meta-analysis to explore the effectiveness of a healthy diet pattern, Mediterranean diet, on frailty and we found that a high adherence to MD was significantly associated with a 44% decreased risk of frailty in the future (RR= 0.56, 95% CI=0.36- 0.89, p=0.015, Fig. 3).

Our results were consistent with previous studies which demonstrated the long-term effects of the MD-associated benefits on maintaining physical function in older people (15, 16, 28). The benefits of Mediterranean diet in reducing risk of frailty might be mediated by several factors. Significant prevention of chronic diseases (such as metabolic syndrome (29), pulmonary diseases (30), cardiovascular diseases (31) and cognitive impairment (32)) supports the favorable effect sof MD on frailty, which is generally thought to be caused by multifactorial factors. Furthermore, a greater adherence to MD was associated with higher intake of total vegetal protein, antioxidant micronutrients (including vitamin C, vitamin E and carotenoids) and lower intake of total saturated fat (14)]. It has been proved that this specific dietary intake pattern can decrease oxidative stress and inflammation (33), both of which play an important role in the pathogenesis of frailty (1, 3).

The meta-analysis of the five prospective cohort studies showed relatively high I-squared suggesting a significant heterogeneity across these studies. A decreased heterogeneity measured by I-squared was observed after the subgroup analysis by population and frailty definition, from 80.2% to 53.7% and 0.0% respectively. It suggested that population and the criteria of frailty might play an important role in producing the heterogeneity. More interestingly, it seemed that the longterm benefit from adherence to MD was more obvious among elders from Europe and North America than that among elders from Asia (Fig.2). Adherence to MD was associated with a 52% reduced risk of frailty (RR= 0.48, 95% CI=0.32-0.72, p<0.001, Fig.2) among elders from western countries while no significant associations were found among elders from Asia (RR=1.06, 95% CI=0.83-1.36, p=0.638, Fig.2).The difference in eating habits between western and eastern countries may contribute to this phenomenon. European and North American used to follow a ‘western diet' (WD) characterized by high intake of red meat and animal fat, and low intake of fresh fruits and vegetables, contains a high amount of highly processed foods and is high in saturated and trans-fats (34), which is almost contrary to a Mediterranean-style diet. Since a Western diet is related to increased levels of inflammation, higher risk of chronic diseases (metabolic diseases, cardiovascular diseases and etc.) (35) and less healthy behaviors (10), the participants who follow a WD are more likely to be frail compared to those follow a MD (10, 36). However, Chinese diet is similar to the Mediterranean diet in that both are characterized by high vegetable and fruit consumption and low meat consumption (26). Therefore, the benefit of MD can be more significant in western countries than that in China. In fact, the impact of MD on frailty in overall population remains uncertain. In the subsequent sensitive analysis, the association showed no statistical significance after exclusion of studies from Talegawkar et al. (13) and Rahi et al (12) (Fig. 4). Even though the small number of the included studies may explain the instability, these results can also inform us that it is necessary to take the local dietary culture into account when trying to encourage old people to follow a Mediterranean diet with the goal of reducing risk of frailty.

Valuable evidence about the impact of MD on frailty in old people. However, some limitations of this study should be pointed out. First, the number of included studies is small thus influence the overall quality of the systematic review. Second, all the included studies were observational studies, which may induce some bias. However, it is difficult to conduct clinical trials on this topic. Furthermore, we did not identify any appropriate clinical trial in this review. Thirdly, in most developing countries, they used 60 years as cut off age for old people and we proposed that the mean age over 60 years meant most participants aged 60 years or above. Thus, we chose the mean age cut-off of 60 years in the inclusion criteria. In this way, we could include any potential studies involved elders in our review. However, this may result in including some studies with some participants aged below 60 years old (eg. Participants aged 45 to 79 years in Veronese et al). Therefore, our results must be interpreted with caution regarding this limitation. Due to the variation in study design, the method of food consumption assessment, diagnostic criteria for frailty and adjustments across studies, a significant heterogeneity was found across the five prospective studies in the meta-analysis. But we did subgroup analyses to try to find the origin of heterogeneity. In order to ensure maximization of the statistical power, ORs and HRs were considered to yield similar estimates of RR and all RR estimates were pooled together (25). In addition, we did not detect the publication bias because of less than ten studies were included.

In conclusion, there is a significant association between a high adherence to Mediterranean Diet and frailty in old people. A higher adherence to Mediterranean diet could lower the risk of frailty in the future. Meanwhile, the benefits may be more obvious among elders from western countries.

Declaration of conflicting interest

None.

Electronic supplementary material

Supplementary material is available for this article at https://doi.org/10.1007/s12603-018-1020-x and is accessible for authorized users.

Supplementary 1: searching strategies