**4. Discussion**

Our study aimed at describing the prevalence of frailty and its associated factors, including dietary patterns, in a national sample of community dwelling older Lebanese individuals. In this rural and urban low socioeconomic sample, the estimated prevalence of frailty was 15% on average, with 10% and 20% in men and women, respectively. In men, poor nutritional status and being older than 75 years were found to be associated with frailty. In women, in addition to these two factors, taking more than five drugs

daily, having at least one age-related condition, having a WHTR > 0.718 and following a Westernized-type DP were found to be independently associated with frailty. In women, the HI-MEDDP showed a significant association with frailty prevalence after adjusting for age, educational level, nutritional status, and WHTR. However, this failed to be significant after further adjusting for polypharmacy and age-related conditions.

Our frailty prevalence can be compared to findings of a meta-analysis conducted in 2018, where pooled prevalence of frailty was 17.4% [42]. A Lebanese national study involving 1200 individuals reported a higher prevalence of frailty, at 36.4%, in an exclusively rural elderly population. In this previous study, 73% of the sample had a monthly income below the minimum wage, whereas in our studied sample, 57.8% had low or insufficient income [19]. The difference with our results can be explained by the variability in study design and tools used for the evaluation of frailty, as well as the heterogeneity between rural and urban settings. Rural areas, being usually poorer and more affected by urban migration of young adults, might witness a higher proportion of frail individuals. [45,54–56].

As shown in previous research, women are more often frail than men [54]. Our results also showed that a low educational level (less than 7 years) was associated with prevalence of frailty, particularly in women. As described in the Longitudinal Aging Study in Amsterdam, the impact of low educational level increased the odds of frailty almost three-fold and this association persisted throughout the 13 years of follow-up [57]. Health related parameters were also found to be associated with frailty, particularly in women; they were frailer if they were taking more than five drugs every day, suffered from two or more diseases, and had at least one age-related condition. These factors have been often linked to frailty in several settings [14,58–62].

WHTR was associated with frailty, in the multivariate analysis, suggesting a possible role of higher abdominal adiposity. Although low BMI is known to be associated with frailty, obesity is also considered as risk a factor of frailty [24]. A meta-analysis showed that overweight individuals (BMI between 25–30 kg/m2) exhibited an increased risk of frailty by 20%, whereas obese (BMI ≥ 30), have an increased frailty risk of 90% [21]. In the longitudinal Doetinchem Study, a BMI < 23 kg/m<sup>2</sup> and ≥30 kg/m<sup>2</sup> was associated with higher incidence of frailty [25]. Similar results were found in the Japanese cohort with a lowest incidence of frailty at a BMI between 21.4 and 25.7 kg/m<sup>2</sup> [22]. As shown by Kim et al., the risk of frailty is higher in obese women, which is mediated by WHTR, but not in obese men [28]. In Spain, two cohort studies showed a parallel change of abdominal obesity and BMI to be associated with an increasing risk of frailty [63].

The level of malnutrition was identified as an independent risk factor associated with frailty, although our sample had a high proportion of obese. Our study showed that malnutrition was associated with a substantially increased prevalence of frailty in the total sample and in both gender groups. The relation between malnutrition and frailty was already clearly established in previous studies, and overweight and obesity often co-exist with frailty [21,22,27,28,63–65].

Among the three dietary patterns identified in our study, the WDP was associated with a higher prevalence of frailty, in both men and women, independently of major confounding factors. This pattern was characterized by a high median sugar intake particularly in in women, and the highest consumption among the three patterns in refined flour products. Several studies reported a link between sugar consumption and the risk of frailty. In the 5-year cohort Seniors-ENRICA study, a high consumption of added sugar, ≥ 36 g/day, compared to <15 g/day, was found to increase the odds of frailty by almost two-fold [66]. Furthermore, in the Nurses' Health Study, after adjustment for confounding factors, consumption of ≥2 servings of sugar-sweetened beverages per day compared to no consumption, increased the risk of frailty by almost 30% over a period of 22 years [67].

Healthy moderate DP and Mediterranean DP were often reported to be inversely associated with frailty [34,36,68,69]. Frailty risk was also found to be inversely associated with consumption of fruits and vegetables [70–72]. Our analysis showed that the Mediterranean dietary pattern with moderate intakes, represented by the MOD-MEDDP in our study, had

the lowest prevalence of frailty. On the other hand, when HI-MEDDP group was compared to MOD-MEDDP in women, the predominance of obesity in this group, with a concomitant high fat intake (the highest among the 3 DPs) and a median protein intake of 0.9 g/kg BW, may have contributed to frailty in this subcategory. This could sugges<sup>t</sup> that these conditions, regardless of the quality of fat and diet, could outweigh the beneficial effect of a Mediterranean diet on the prevalence of frailty. Previous reports suggested that protein intake between 1 and 1.5 g/kg BW were necessary for the prevention of frailty [29–32].

In summary, our study was the first to explore the association between frailty and specific dietary patterns extracted in a posteriori method, in adults over 60 years of age in Lebanon, using specifically validated questionnaires. Despite the difficulties in addressing this specific age group, we succeeded in shedding light on some findings specific to our Mediterranean older adult population. Age, age-related conditions, polypharmacy, and malnutrition, remain the main associated factors related to frailty in low socioeconomic settings. We also showed that malnutrition and abdominal obesity co-exist as risk factors for frailty. Most importantly, we demonstrated that a Westernized-type pattern with high sugar consumption, and to a lesser extent, a Mediterranean high caloric intake pattern, were also linked to frailty, and that a more moderate Mediterranean-like pattern was protective, especially in women. More efforts should target actions that improve modifiable factors to prevent or reverse frailty, such as eating patterns and diets that improve WHTR.

We note, however, some limitations concerning our results in relation to the low number of frail participants adopting the WDP; this consequently implies taking the present findings with caution.

Larger prospective studies are required to further investigate the impact of dietary patterns on risk of frailty with a special emphasis on WHTR.

In conclusion, WDP had the strongest association with frailty in this sample. MOD-MEDDP, in comparison to HI-MEDDP and WDP, was associated with the least prevalence of frailty. In this Mediterranean sample, a diet far from the traditional one appears as a key deleterious determinant of frailty.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/article/10.339 0/nu13072188/s1, Table S1: food groups and food items included in the cluster analysis, Table S2: consumption of predefined food categories according to dietary patterns for men and women separately, Table S3: nutritional characteristics of the dietary patterns.

**Author Contributions:** Conceptualization and methodology, N.Y., C.B.; validation, M.A., C.B., C.F.; formal analysis, C.Y. and N.Y.; data curation, N.Y.; writing—original draft preparation, N.Y.; review and editing, C.B., M.A., and C.F.; supervision, C.B.; project administration and funding acquisition, N.Y., M.A. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Research Council of the Saint Joseph University of Beirut, gran<sup>t</sup> number FPH66.

**Institutional Review Board Statement:** The study was conducted according to the guidelines of the Declaration of Helsinki, and approved by the Ethics Committee of SAINT JOSEPH UNIVERSITY (USJ 2016-99).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study.

**Acknowledgments:** We would like to thank Fernande Abou Haidar, Samar Sleilati, and Marie Elias from the Ministry of Social Affairs/Department of Family Affairs, and Rita Hayeck, & Rafic Baddoura from the National Commission for the Elderly, for their precious collaboration and support.

**Conflicts of Interest:** The authors declare no conflict of interest.
