In this study, we found that male sex, lower BMI, lower MMSE, lower ADL, frailty, polypharmacy, and higher CCI were risk factors for long-term mortality. We discovered that mortality risks were significantly increased, especially in male patients with frailty. Moreover, polypharmacy was a major prognostic determinant in the male, non-frail group. Therefore, to improve long-term patient survival, it is prudent to identify and manage polypharmacy appropriately, especially in male patients without frailty.
The prevalence of frailty could vary by assessment tools, region, and different clinical settings. A systematic review study showed that the frailty prevalence in the community setting measured using physical frailty was 12% compared to 24% for the deficit accumulation model[5]. Using the frailty phenotype as the assessment tool, Tobias et al showed that the frailty prevalence in the outpatient setting was 17.8%[24]. However, another study conducted at Ain Shams University Hospitals, Singapore showed a much higher prevalence: 48%, using the clinical frailty scale as the assessment tool[25]. In our study, the prevalence of frailty in the outpatient setting of a tertiary referral hospital was 44%. Although we used the CHS phenotypic criteria to identify frailty, our result is similar to that of the study by Hagar et al[25]. In addition, Chen et al indicated that the prevalence of frailty, prefrailty, and non-frailty in community-dwelling Taiwanese were 4.9%, 40.0%, and 55.1%[26]. To our knowledge, this is the first study to report the prevalence of frailty in the older Taiwanese patients in outpatient clinics. The results of our study provide insight into sex-specific mortality risks with regard to frailty and polypharmacy to enable the development of potential strategies for managing frailty in this vulnerable population.
Many studies have shown that the prevalence of frailty is higher in females than in males, even in the same age group. However, the mortality rate of older men is higher than that of older women[7, 8], which is known as the “sex-frailty paradox”[8, 27]. Furthermore, our findings support the abovementioned paradox. Sex differences in genetic, biological, or psychosocial factors as well as in physiological reserve, burden of disease, and disability might contribute to this paradox[8, 27]. However, our data suggested that a potentially unrecognized factor – polypharmacy – might contribute differentially to mortality in the male and female groups. Further studies are needed to confirm our findings.
In addition to frailty, the geriatric syndrome is associated with mortality in the older population[28–30]. A longitudinal cohort study conducted at the outpatient geriatric clinic of Amsterdam University Medical Center classified the common geriatric syndrome into 5 domains (somatic, mental, nutritional, physical, and social) and the result showed that the geriatric syndrome was cumulatively associated with mortality[13]. These results raise an intriguing possibility that the interaction between the geriatric syndrome and frailty may be partly responsible for the sex-frailty paradox. Our study found that polypharmacy is an important determinant of mortality and confers different effects between both sexes as well as between groups with and without frailty. We speculated that polypharmacy may be one of the causes of the sex-frailty paradox. Polypharmacy is highly prevalent in the older population and is as high as 46.6% of the geriatric population. The higher the number of drugs that are used, the higher the risk of hospitalization and death[17]. A systemic review of 25 observational studies demonstrated a significant association between polypharmacy and frailty[31]. Hagar et al demonstrated that the prevalence of polypharmacy in the frail group was significantly higher than that in the non-frail and pre-frail groups[25]. Moreover, Beatriz et al revealed thatpolypharmacy and frailty status are interrelated, and their interaction determines the frequency of health-related adverse events. Polypharmacy and frailty were associated with a significantly higher incidence of incident disability, hospitalization, and mortality, compared to non-frail individuals without polypharmacy [32]. The results from the European dataset, Survey of Health, Ageing, and Retirement in Europe ( SHARE )cohort showed that polypharmacy contributes more to the long-term mortality in the non-frail group compared with that in the frail group[33]. Moreover, this finding is supported by the results of another population-based cohort study that was conducted by using Taiwan’s National Health Insurance Reimbursement Database[34]. Within the same polypharmacy category, the grade of frailty status was associated with an adverse effect on mortality. Interestingly, the dose–response association between polypharmacy and mortality was only observed in the fit and mild-frail participants[34]. Our findings that polypharmacy modulates the 5-year mortality risk in the non-frail, but not in the frail, group supports the results reported from the abovementioned studies. Nevertheless, the presence of a sex difference in the interaction between polypharmacy and frailty was previously unknown. We found that polypharmacy might contribute to the long-term survival outcome only in the male and non-frail older adult group. Our data support the assumption that avoiding polypharmacy and inappropriate medication might improve the risk of long-term mortality, especially in male, non-frail groups.
There are some limitations of our study. First, our study population was recruited from the geriatric clinics of a single hospital in Taiwan. Our result might not be generalizable to the general older outpatient and patients of non-Han ethnic group. Second, the medication categories and pill burden of polypharmacy were not analyzed. Therefore, we could not confirm which drug classes might have a deleterious impact on long-term mortality. Third, this study did not analyze the cause of death. Thus, the presence of a causal relationship between polypharmacy, frailty, and death needs to be ascertained. However, this study provided valuable information that polypharmacy might contribute significantly to mortality in the non-frail, male older adult group. Further prospective studies are needed to determine whether a reduction of polypharmacy offers long-term survival benefits to geriatric patients, with or without frailty.