Background
The number of patients with diabetes is increasing and is expected to reach 300 million globally by 2025 [
1,
2]. The primary aim of diabetes treatment is to prevent vascular complications and maintain quality of life (QOL) [
3]. Moreover, this treatment is extremely important to improve prognosis and prevent the onset of cardiovascular diseases [
4]. Furthermore, hospitalization of patients with diabetes due to complications and severe hypoglycemia has subsequently increased medical expenses [
5,
6], indicating that controlling medical expenses related to hospitalization will become increasingly crucial in the future.
Recently, frailty has gained attention in the field of diabetology. Frailty is defined as a condition in which physical and mental activities reduce with age, leading to physical and mental weakness; however, in frailty, activities of daily living and QOL can be maintained through appropriate intervention [
7,
8]. In the literature, varying incidence of frailty in middle-aged to elderly patients with diabetes has been reported ranging from 32 to 48% [
9]. The prevalence of frailty among community-dwelling elderly is 5–10% [
9‐
11]. In patients with diabetes, chronic inflammation, increased oxidative stress, and insulin resistance cause loss of musculoskeletal mass and muscle weakness, which may increase the incidence of frailty [
9,
12]. Furthermore, it is thought that frailty causes chronic inflammation and insulin resistance, which are believed to be closely related to vascular complications and mortality [
13]. In patients with diabetes, death and cardiovascular diseases are attributed to classical risk factors such as hypertension, dyslipidemia, and smoking habit in approximately 60% of the patients [
14], whereas these events are attributed to frailty in the remaining patients [
14‐
16]. Moreover, frailty is reportedly associated with hospitalization and higher medical expenses [
17,
18], which is considered a problem in terms of medical economics. Frailty is thought to improve with appropriate intervention [
19‐
21]; thus, its early detection for early intervention is considered important.
As mentioned above, examination of the relationship of frailty with mortality, hospitalization, and cardiovascular diseases in patients with diabetes is important both clinically and in terms of medical economics. Therefore, the aim of the present study was to comprehensively analyze the relationship between frailty and mortality, hospitalization, and cardiovascular diseases in patients with diabetes through a systematic review and meta-analysis.
Discussion
The present study examined the relationship between frailty and mortality, hospitalization, and cardiovascular diseases in patients with diabetes using a meta-analysis. As a result, prefrailty and frailty were found to have a significant relationship with mortality and hospitalization. Regarding cardiovascular diseases, although a relationship was found with prefrailty and frailty, only 1 study was included, thereby not providing robust results.
According to the meta-analysis of previous studies that examined the relationship between frailty and mortality in community-dwelling individuals, the pooled HR of prefrailty related to mortality was 1.75 (95% CI 1.14–2.70) [
33]; when calculated for frailty, the risk of mortality increased by 1.8–2.3-fold [
34]. In the present study, while a significant relationship was observed between the pooled HR of prefrailty and frailty related to mortality, we thought that the pooled HR was lower than that reported in previous studies. Among the studies included in our meta-analysis, in the study by Chao et al. [
31], the participant sample size was larger than that included in other studies; moreover, the patients were in their 50 s, which is relatively young. The impact of frailty on mortality increases with age [
35,
36]; therefore, it is possible that the pooled HR of frailty related to mortality was underestimated in the present study. In contrast, the relationship between frailty and mortality was observed only in patients aged < 60 years in the subgroup analysis. A previous study involving patients with type 2 diabetes indicated that the impact of diabetes on mortality was higher in middle-aged patients than in elderly patients [
37]. This might be caused by higher smoking and obesity rates as well as lower prescription rates of statins in middle-aged patients with diabetes than in elderly patients with diabetes [
37]. Middle-aged patients with diabetes with an increased risk of death possibly have a huge impact of frailty on their mortality. However, given the small sample size of patients aged ≥ 65 years, the results of the present study may be underpowered. Thus, further studies are required to examine the impact of frailty on mortality according to age. When analyzing the comprehensive relationship of prefrailty and frailty with mortality using pooled ORs, no significant difference was noted. It was inferred that a relationship was not observed because of the small sample size and statistical power.
In a meta-analysis on the relationship between frailty and hospitalization in community-dwelling individuals, the pooled OR of prefrailty and frailty related to hospitalization was 1.26 (95% CI 1.18–1.33) and 1.90 (95% CI 1.74–2.07), respectively [
17]. In the present study, the pooled OR of prefrailty and frailty related to hospitalization was higher than that reported in previous studies, thus suggesting that frailty contributes to hospitalization in patients with diabetes. Furthermore,
I2 was 0% in our analysis with no heterogeneity; thus, it was thought that the results were robust to a certain degree. Although the subgroup analysis for age showed a relationship between frailty and hospitalization regardless of age, a relationship between prefrailty and hospitalization was observed only in patients aged ≥ 60 years. Some previous studies [
27,
30,
32] have indicated that age possibly strongly influences prefrailty related to hospitalization, consistent with our results. Therefore, caution for hospitalization is clinically important in prefrail elderly patients with diabetes. When analyzing the comprehensive relationship between prefrailty and frailty related to hospitalization using pooled HR, we assumed that HR was relatively small, although a significant relationship was observed between them. It was inferred that the young age of the study participants included in this analysis could have affected the results.
In the present meta-analysis, very few included studies examined the relationship between frailty and cardiovascular diseases; as a result, robust results were not obtained. In previous studies on community-dwelling individuals, it was reported that prefrailty and frailty are the independent risk factors for cardiovascular diseases [
38]. In the present meta-analysis, there was only one study [
31] that examined the relationship between frailty and cardiovascular diseases in patients with diabetes; therefore, we believe that further analysis using more studies is warranted.
Although the mechanism underlying the relationship of frailty with mortality, hospitalization, and cardiovascular diseases in patients with diabetes remains largely unclear, the following mechanism is considered. Frailty is closely associated with reduced physical and/or cognitive function [
39], which leads to poor vital prognosis [
40,
41]. It is possible that performing less physical activity along with reduced cognitive function will contribute to the relationship between frailty and prognosis. Furthermore, as another mechanism, it is suggested that hypoglycemia is involved. In previous studies, hypoglycemia has been found to be associated with a risk of increased mortality and cardiovascular diseases [
42,
43]. Reportedly, the prevalence of hypoglycemia increases with frailty [
44], and hypoglycemia may contribute to the relationship of frailty with mortality and cardiovascular diseases. Furthermore, in the present meta-analysis, a particularly robust relationship was observed between frailty and hospitalization. Previous studies [
45,
46] have suggested that accidental falls are involved as the mechanism linking frailty and hospitalization. It is believed that falls are common among patients with diabetes [
47], and it is possible that falls contribute even more to the relationship between frailty and hospitalization. In addition, severe hypoglycemia and its complications are closely associated with hospitalization in patients with diabetes [
5,
6]. In patients who are frail and have diabetes, the accumulation of factors such as falls, severe hypoglycemia, and its complications may contribute to hospitalization. However, in the present meta-analysis, the reason for hospitalization was not determined; thus, further examination of these mechanisms is needed.
In patients with diabetes, death and cardiovascular diseases are attributed to classical risk factors such as hypertension, dyslipidemia, and smoking in approximately 60% of the patients [
14], and the contributing factor for the remaining 40% is frailty [
14‐
16]. A previous study reported that frailty is a prognostic factor for mortality independent of diabetes-related complications [
25]. In patients with diabetes, frailty is now considered an important predictor of vital prognosis [
48], and the importance of medical care that takes frailty into consideration has been proposed [
49]. Nutrition for frailty [
19], exercise [
20,
21], and avoidance of hypoglycemia [
39] may prevent the exacerbation of or improve frailty, and it is thought that early detection of frailty and early intervention are important against frailty. In the future, further examination is warranted to assess the effect of therapeutic intervention on vital prognosis and hospitalization for patients with diabetes and frailty.
The present meta-analysis has several limitations. First, we cannot eliminate the possibility of relevant studies in the databases that we do not use for literature search in our meta-analysis, which may have affected the results. Second, our meta-analysis includes some studies wherein the adjustment for confounding factors is considered inadequate, which may have caused a bias. Third, the definition of frailty used in the included studies differs among the studies, which may have affected the results. Fourth, heterogeneity is particularly high in the analysis involving mortality and cardiovascular diseases as the outcomes, which may have also affected the results. Lastly, relatively few studies were included in our meta-analysis, and a subgroup analysis could not be performed. We believe that re-examination to overcome these limitations is needed in the future wherein more studies can be included.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.