Background
Recently, although life expectancies in the developed countries, including Japan, have been increasing, the number of older people with functional disabilities who need assistance from others is also on rise. Extending healthy life expectancy is an urgent task for the gerontologists.
Frailty is a state in which an older person becomes vulnerable to the external stresses due to declining age-related physiological reserve and can lead to disabilities, falls, fractures, and death [
1,
2]. Frailty is a reversible condition because physical and nutritional intervention can improve a person’s physical condition. The concept of multidimensional frailty based on a comprehensive geriatric assessment has been proposed because cognitive and social frailties, as well as physical frailty, have a major effect on disability and mortality. Thus, it is essential to screen for frailty and cognitive deficits in the older people to prevent deterioration of their functional ability.
The prevalence of frailty has been reported to be approximately10% in the general population of older inhabitants. Although cardiometabolic diseases [diabetes mellitus (DM), hypertension (HT), dyslipidemia (DL), and heart failure] have been associated with the prevalence of frailty in epidemiological studies, this prevalence in the individuals treated in the cardiology and diabetes specialty outpatient clinics remains unknown.
However, it is difficult to complete the multidimensional assessment of frailty during the routine visits in the outpatient clinics. Therefore, we recently established a frailty clinic and identified a cohort group of patients in the clinic for inclusion in a 3-year prospective longitudinal study.
The aim of this prospective study was to answer the following questions: first, what is the prevalence and incidence of frailty in the specialized frailty clinic? what are the associations, if any, between frailty status and clinical outcomes of fall, cardiovascular disease, dementia, hospitalization, functional disability, and death? and what are the most useful indices for predicting these outcomes in evaluating frailty status?
In this article, we describe our frailty clinic and the baseline characteristics of the patients in a cohort for the prospective longitudinal study.
Discussion
In this study, we described our recently established frailty clinic, which mainly treats patients with cardiometabolic diseases. At present, there is an urgent need to assess frailty and cognition in older individuals with cardiometabolic disease because diabetes and cardiovascular disease are associated with the aging process and with frailty and cognitive impairment [
19]. We performed a comprehensive geriatric assessment for these patients to evaluate frailty, suspected dementia, cognitive impairment, and sarcopenia and found that the prevalence of all of these conditions increased with increasing age (Fig.
2).
The prevalence of frailty diagnosed according to the mCHS criteria in our study population was about twice as high as that of the recently reported community-dwelling older persons, and in most of those, the prevalence was approximately 10% [
20‐
22]. One reason for the discrepancy was the difference in the diagnostic criteria. We modified the cutoff points for SMI, grip strength, and walking speed, and we substituted some questions from the KCL for the original ones since they were easier to obtain answers from the subjects; however, the major reasons for the discrepancy was that our study populations included higher rate of patients aged over 80 and that they comprised outpatients, especially those who had cardiometabolic diseases, since most of them were recruited from the departments of cardiology or diabetes.
In this report, we attempted to evaluate frailty status using several different scales, but the prevalence of frailty varied from 24 to 34% depending on the scale. The highest prevalence was observed in the KCL criteria, which was 10% higher than that in the mCHS criteria, perhaps because the KCL questionnaire assesses multidimensional aspects of frailty, including physical, cognitive, and social frailty; malnutrition; poor oral health; and depression. The high prevalence of cardiometabolic diseases could explain this discrepancy, particularly with DM, which is frequently associated with physical frailty as well as mental disorders, as is the case with the TMIG-IC criteria, which includes questionnaires regarding intellectual activity and social roles. The CFS criterion showed a high prevalence of frailty when we defined it using a cutoff value of ≥4 instead of the original cutoff value of > 5 (the prevalence rate was reduced to 8.6% when we diagnosed frailty using the ≥5 cutoff). In a prospective study, we believe that all of these results will be useful in clarifying the most appropriate diagnostic criteria for predicting functional disability or mortality in patients with cardiovascular risk factors.
Although a few patients were suspected of having dementia, there were a substantial number of patients with mildly impaired cognitive function, and the frequency of the MoCA-J ≤ 25 was revealed to be ≥80%. It has been reported that mild cognitive impairment (MCI) could already be a significant risk for progression of disabilities in older persons [
23], and screening these patients is vital. Similar to frailty, the prevalence of cognitive impairment was significantly different among diagnostic criteria. It is known that MoCA is more sensitive than the MMSE for detecting MCI because it assesses cognitive domain impairment, including executive functioning, attention and concentration, visuospatial skills, and memory. In a report by Trzepacz et al., a score of 25 for MoCA was equivalent to a score of 29 for the MMSE and a score of 26–30 for MoCA was equivalent to a score of 30 for the MMSE [
24]. Thus, MoCA may be more sensitive than the MMSE in detecting cognitive impairment. In fact, reports have shown the superiority of MoCA over the MMSE in screening for MCI in patients with DM [
25] and heart failure [
26].
The prevalence of sarcopenia in our population was considerably higher than that of community-dwelling people in the Asian countries (males, 7.1%; females, 19.8%) [
14]. This is natural because most of our subjects were affected by various chronic diseases. Indeed, both DM and heart failure are known to be risk factors for skeletal muscle mass reduction [
27,
28]. Similar to our study, Han et al. have reported that the prevalence of sarcopenia in China increased with the accumulation of cardiovascular risk factors, DM, HT, and DL using the AWGS criteria [
29]. However, the prevalence of sarcopenia in individuals with these diseases was 11.1–22.2%, which was low compared with that in our subjects [
29]; this could be explained by the difference in age of the study subjects as well as the exclusion criteria regarding patients with previous cardiovascular diseases. Notably, among the diagnostic items of sarcopenia, the majority of patients met the criterion of low muscle mass and muscle weakness, whereas almost all did not meet that of walking speed (Table
3). In the AWGS criteria, the cutoff points for SMI and grip strength were slightly lower than those of the European Working Group on Sarcopenia in Older People [
30]; however, the cutoff point for walking speed remained unchanged. It is suspected that SMI of the Japanese is considerably smaller than that of the European people reflecting their small body size, whereas in contrast, walking speed in the Japanese older people is comparatively faster [
31]. Considering these specific characteristics of Japanese, it might be necessary to produce a more appropriate diagnostic criterion for sarcopenia in the Japanese older persons. In addition to the items needed for the diagnosis of sarcopenia, we performed the one-leg standing and TUG tests. It has been reported that both indices are associated with instrumental ADL status [
13,
32] and falls [
33,
34].
In this study, we also evaluated the depressive mood, nutritional status, and social support network since several reports have revealed that depressive mood and malnutrition could be risks for frailty [
35,
36], and a recent report revealed that the older persons living alone are susceptible to becoming frail [
37], which indicates that the lack of social support also could be a crucial risk factor for frailty.
We found that the prevalence of frailty, cognitive impairment, and sarcopenia increased with advancing age; however, the prevalence of sarcopenia plateaued in the subjects > 80 years of age. Few studies have investigated sarcopenia in very old subjects (≥85 years). Our ceiling effect could be accounted for by selection bias. Because our study was held mainly in an outpatient clinic, those who registered and were > 85 years old were relatively healthy and did not represent the general population of the same age. In the Newcastle 85+ study, the authors mentioned that low BMI (≤18.5) was a significant risk factor for the prevalence and incidence of sarcopenia in this age group [
38]. In our patients ≥85 years; however, the median BMI was 21.6, and most of the patients had normal nutrition.
It has been reported that DM are associated with high prevalence and incidence of frailty [
39,
40], and that HT is also related to prevalent frailty [
41,
42], our results stratified by age revealed almost no significant difference in the prevalence of frailty by DM or HT. It is also known that DM is associated with cognitive dysfunction [
43] and sarcopenia [
44], but no difference was observed except for the prevalence of sarcopenia in the youngest group. This result may also be due to a selection bias in the specialty clinic. Although these subjects became stable, they might have been referred from general practitioners because they had poor control of glucose or blood pressure and multi-morbidities, such as CAD, stroke, or heart failure. These backgrounds of the subjects could have diluted the effects of each single disease, especially in those in the older age groups. For example, it has been reported that chronic heart failure was associated with frailty [
45] and cognitive impairment [
46]. However, even when considering the bias above, the prevalence of frailty, sarcopenia, and cognitive impairment in the youngest DM group appears to be high, suggesting the importance of taking all possible measures to prevent frailty from occurring at an earlier stage in DM patients.
The prevalence of frailty in the oldest DM group (aged ≥85 years) showed a trend of reduction compared to that in the non-DM group; however, this was not the case for HT, and the reason for this discrepancy remains unclear. Perhaps, some selection or survivorship bias might have influenced these results. Nevertheless, it is unclear why the prevalence of sarcopenia was low in HT subjects aged 80–84 years.
It has been reported in the Japanese population that frailty is associated with both sarcopenia and cognitive decline [
47]; however, our results provide valuable new information about how these conditions related to aging overlap with each other. Although the prevalence of cognitive impairment was high, it is noteworthy that almost all of the frail and sarcopenic subjects were cognitively impaired. Alternatively, the prevalence of suspected dementia among the frail and sarcopenic subjects were relatively small. The factors that determine the coincidence of progression of physical function and cognitive decline should further be elucidated by observing this cohort longitudinally.
The strength of our study was that this is the first study to describe the establishment of a frailty clinic with the unique unprecedented backgrounds of our clinic’s patients. Several studies have evaluated frailty status in their own frailty clinics; however, their patients’ backgrounds were quite different from ours. Tavasson et al. reported that the prevalence of frailty in those who visited their original frailty clinic was 54.5%, which was considerably higher than ours [
48]. Their registration criterion was “those considered as frail by their physician” so that they could include several patients with functional disabilities; this was evidenced by the low mean gait speed of their participants of 0.78 m/s. The prevalence of frailty in other studies assessed in “geriatric outpatient clinics” were approximately 35% [
49,
50], which were close to the prevalence in our study; however, the sample numbers were small (
n < 200), the studies were conducted in the US or Canada, and although the prevalence of hypertension was higher in one of the studies, there appeared to be few patients with metabolic diseases. Our study appears to be the first that was conducted in Japanese patients with mainly cardiometabolic diseases who were self-supported but at high risk of becoming frail.
Another strength of our study is that we evaluated frailty and cognitive status by using multiple test modalities, including the CHS, CFS, KCL, and TMIG-IC for frailty and the MMSE, HDS-R, MoCA-J, and DASC-21 for cognitive impairment. This study characteristic is also unprecedented. Using these precise datasets, we could determine the index for frailty and cognitive function that is most associated with and most appropriate to predict a specific outcome. The comprehensive assessment using patients with cardiometabolic disease at baseline will help us to explore risk factors for the progression of frailty and cognitive decline in an ongoing 3-year longitudinal prospective observational study concerning frailty in patients with diabetes or heart diseases.
Our study had some limitations. First, the study was conducted with a relatively small sample size to detect a difference in the age- and disease-stratified analyses. Nevertheless, our results have clarified for the first time the prevalence of frailty, sarcopenia, and cognitive impairment in patients with a wide range of age who presented with cardiometabolic diseases. Second, this study was performed only in one Japanese institution. Our results should be confirmed in large multicenter and multiracial studies. In addition, the heterogeneity of our subjects’ backgrounds may make it difficult to apply the results to the general population. We plan to expand the subject samples to include a wide variety of diseases. Third, as this analysis was performed in a cross-sectional study design, the causal associations between cardiometabolic diseases and frailty, sarcopenia, or cognitive impairment remain unknown. To clarify the exact associations, further longitudinal studies are warranted.