Association of Frailty and Cardiometabolic Risk Among Community-Dwelling Middle-Aged and Older People: Results from the I-Lan Longitudinal Aging Study

Abstract

The aim of this study was to evaluate the association of cardiometabolic risk and frailty through a community-based aging cohort in Taiwan In total, 1839 participants (men, 47.5%; mean age 63.9 ± 9.3 years) from the first wave of the I-Lan longitudinal cohort study, recruited between August of 2011 and August of 2013, were retrieved for the analysis of this cross-sectional study. Frailty was defined by Cardiovascular Health Study (CHS) criteria, encompassing un-intentional weight loss, slow walk speed, poor grip strength, exhaustion, and low activity. Comparisons between cardiometabolic risk and frailty status were performed after adjustment for age, hormone parameters, functional measurements, and skeletal muscle mass. Independent association of cardiometabolic risk and frailty status was identified through the multivariate logistic regression model. We found that the prevalence of frailty and pre-frial were 6.8% and 40.5%, respectively. Adjustments for age, blood pressure, low-density lipoprotein cholesterol (LDL-C), uric acid, creatinine, and carotid intima media thickness were not significantly associated with frailty. However, lower total cholesterol and high-density lipoprotein cholesterol (HDL-C), higher high-sensitivity C-reactive protein (hsCRP) and glycemia profiles were significantly associated with frailty. For hormone parameters, dehydroepiandrosterone sulfate (DHEA-S), insulin-like growth factor-1 (IGF-1), and free androgen index were not significantly associated with frailty after age adjustment. In a multivariate logistic regression model, abdominal obesity, homeostasis model assessment of insulin resistance (HOMA-IR), and hsCRP were significantly associated with frailty. The odds ratio (OR) for frailty was 3.57 (95% confidence interval [CI] 1.88–6.78, p < 0.001), 1.30 (95% CI 1.02–1.66, p = 0.032), and 1.66 (95% CI 1.10–2.49, p = 0.016), respectively, in a fully adjusted model. Conversely, higher total cholesterol was associated with a lower prevalence of frailty (OR = 0.44, 95% CI 0.22–0.89, p = 0.023) in the final model. In this study, abdominal obesity, insulin resistance, and inflammation were significantly associated with frailty, and the effect was independent of functional measurement and decline of skeletal muscle mass. An integrated approach targeted at cardiometabolic aging and frailty is needed in clinical practice.

Introduction

Frailty is a biological syndrome characterized by a decline of physiological reserve across multiple systems and increased vulnerability to physical and/or mental stress during the process of aging.¹ It is also often characterized by deterioration of cognitive function, muscle and nerve systems, as well as cardiopulmonary fitness.² Frailty is described as the pre-disability state, which may result in multiple negative health outcomes, such as low quality of life, hospitalization, and death.^3,4 More importantly, frailty is considered to be a potentially reversible state that has attracted extensive research attention internationally.^5,6 Recognition of frailty has been of increasing importance for cardiovascular health because an association between frailty and increased risk of cardiovascular disease has been recently reported,⁷ and frailty also adds extra prognostic significance to existing risk models.⁸ Frail older people are at significantly higher risk of mortality and morbidity than non-frail older people across various kinds of cardiovascular diseases,⁹ encompassing stable coronary heart disease (CHD),¹⁰ acute coronary syndromes,¹¹ heart failure,¹² and surgical interventions.¹³ The importance of frailty for cardiovascular diseases is not only its prognostic impact but also the benefit of therapeutic interventions for frailty to improve cardiovascular outcomes.¹⁴

Previous literature has suggested that frailty may lead to cardiovascular diseases.⁷ On the other hand, mid-life vascular risk factors, such as low physical activity¹⁵ and CHD, have also been reported to increase incidence of frailty.^8,16,17 It is noteworthy that the current consensus for hypertension, hyperlipidemia, and diabetes management all support modifying the treatment goal for older people, especially among those with impaired activities of daily living and/or cognition, as well as frail older people.^18

–21 Several studies have been done to evaluate the association between cardiovascular risk factors and frailty, but these associations have not been reported in Asian populations.^22,23 Moreover, it remained unclear whether the association between frailty and cardiovascular risk was independent of mood, nutritional status, cognitive function, and hormone parameters. Therefore, the main aim of this study was to evaluate the association between frailty and cardiometabolic risk among community-dwelling middle-aged and older Taiwanese people after adjustment for various potential confounders that may have been overlooked in previous studies.

Materials and Methods

Study participants

The I-Lan Longitudinal Aging Study (ILAS) is a population-based aging cohort study in I-Lan County, Taiwan. The main purpose of ILAS was to investigate the complex inter-relationship between aging, cognitive decline, sarcopenia, and frailty. ILAS randomly enrolled community-dwelling residents aged 50 years and older in the Yuanshan Township of I-Lan County. Subjects who met one of the following criteria were excluded from the study: (1) Unable to communicate to complete an interview; (2) functionally dependent, i.e., unable to walk for 6 meters within a reasonable period of time; (3) limited life expectancy (less than 6 months) due to major illnesses; (4) currently institutionalized; and (5) contraindicated for brain magnetic resonance imaging. The data were retrieved from the first wave of the cohort, collected from August of 2011 to August of 2013. The entire study was approved by the Institutional Review Board of National Yang Ming University.

Demography and physical assessment

A questionnaire encompassing information about demographic characteristics, including educational level, smoking and alcohol consumption, medical history, and medication currently taken, was completed by trained interviewers. Hypertension, diabetes mellitus, and dyslipidemia were defined as self-reported accompanied by use of related medications. CHD was defined as history of re-vascularization or self-reported history of myocardial infarction or angina accompanied by use of anti-anginal medications. Multi-morbidity was assessed by using Charlson's Co-Morbidity Index (CCI). Cognitive function, nutritional status, and depressive symptoms were evaluated by the Mini-Mental State Examination (MMSE), Mini-Nutrition Assessment (MNA), and 5-item Taiwan Geriatric Depression Scale (TGDS-5), respectively. A TGDS-5 scoring of 2 or higher was defined as positive for depressive symptoms.²⁴

All participants also underwent anthropometric measurements by trained research nurses, including height, body weight, and abdominal circumferences. Body mass index (BMI) was calculated as weight (kg) divided by height squared (m²). When measuring blood pressure, participants were seated quietly for at least 5 min in a chair with the arm supported at heart level. An appropriate-sized cuff was applied to ensure accuracy. The systolic (SBP) and diastolic blood pressure (DBP) values of two measurements were recorded.²⁵ For skeletal muscle mass measurement, a whole body dual-energy X-ray absorptiometry (DXA) scan was performed for each participant to measure lean body mass (LBM) using a Lunar Prodigy instrument (GE Healthcare, Madison, WI). Appendicular skeletal muscle mass (ASM) was defined as the sum of the lean body mass of four limbs and was adjusted by square of height, and ASM/height² (ASM/ht², kg/m²) was used for further analysis.²⁶

Cardiometabolic risk and hormone profile

The venipuncture was done after a 10-hr overnight fast for all study subjects. Serum concentrations of glucose (mg/dL), total cholesterol (TC; mg/dL), triglycerides (TGs; mg/dL), low-density lipoprotein cholesterol (LDL-C; mg/dL), high-density lipoprotein cholesterol (HDL-C; mg/dL), uric acid (mg/dL), creatinine (mg/dL), and high-sensitivity C-reactive protein (hsCRP; mg/dL) were measured by using an automatic analyzer (ADVIA 1800, Siemens, Malvern, PA). Whole-blood glycated hemoglobin A1c (HbA1c) was measured by the enzymatic method using the Tosoh G8 HPLC Analyzer (Tosoh Bioscience, Inc., San Francisco, CA). The serum level of insulin (μIU/mL) was measured by the chemiluminescence immunoassay analyzer (ADVIA Centaur, Siemens, Malvern, PA). Homeostatic model assessment of insulin resistance (HOMA-IR) was calculated as [glucose (mg/dL) × insulin (uIU/mL)]/405.

Abnormality of cardiometabolic profiles were defined as follows: (1) High blood pressure, with SBP ≥140 mmHg or DBP ≥90 mmHg, (2) high TC, with TC ≥160 mg/dL, (3) low HDL-C, with serum HDL-C <40 mg/dL in men or <50 mg/dL in women, (4) high TGs, with serum TGs level ≥150 mg/dL, (5) high LDL-C, with serum LDL-C level ≥130 mg/dL, and (6) abdominal obesity, with abdominal circumference ≥90 cm in men and ≥80 cm in women. The serum levels of total testosterone (ng/dL) were measured using a chemiluminescence immunoassay analyzer (ADVIA Centaur, Siemens). Dehydroepiandrosterone sulfate (DHEA-S, μg/dL) and sex hormone-binding globulin (SHBG; nM/L) were measured by using electrochemiluminometry (Roche Elecsys e411; Roche, Indianapolis, MI). The free androgen index (FAI) was defined as total testosterone (nM/L)/SHBG (nM/L). Insulin-like growth factor-1 (IGF-1; ng/mL) was measured by chemiluminescence immunoassay analyzer (DPC Immulite 2000, SIEMENS, Malvern, PA).

Carotid intima media thickness (IMT; mm) was measured by using high-resolution ultrasonography (GE LOGIQ 400 PRO; GE, Cleveland, OH) by experienced technicians. IMT was defined as the distance from the edge of the lumen–intima interface to the edge of the collagen-containing upper layer of the adventitia of the bilateral common carotid artery in longitudinal view. The mean value of the right and left IMT was used for analysis.

Definition of frailty

The diagnosis of frailty was based on the Cardiovascular Health Study (CHS) criteria, which consisted of five components—weight loss, weakness, slowness, low physical activity, and exhaustion.²⁷ (1) Weight loss was defined as un-intentional weight loss of 5% or more in the last year. (2) Weakness was defined as the gender-specific lowest quintile of maximal strength on the dominant hand. Strength was measured with a digital dynamometer (Smedlay's Dynamo Meter; TTM, Tokyo, Japan). (3) Slowness was defined as the gender-specific lowest quintile of gait speed in the timed 6-meter walk test. (4) Low physical activity was defined as the lowest quintile of International Physical Activity Questionnaire–Short Form (IPAQ-SF) score, which was used to assess weekly energy expenditure in kcal based on self-reported physical activities and exercises performed. (5) Exhaustion was evaluated by self-report of fatigue as indicated by two questions from the Center for Epidemiologic Studies Depression Scale (CES-D), including “I felt everything I did was an effort” and “I could not get going.” Exhaustion was defined when a positive answer was provided to either of the two questions with a frequency of more than 3–4 days per week. Participants with three or more of these characteristics were defined as frail and those with one or two characteristics as pre-frail, whereas those without any of the characteristics were robust.

Statistical analysis

In this study, we categorized the participants according to degree of frailty (robust, pre-frail, frail) and to investigate the relationship between frailty with cardiometabolic and hormone parameters and function measurements cross-sectionally. Continuous variables in the text and tables were reported in mean values and standard deviation (SD) and categorical variables were reported in percentage. For continuous variables, analysis of variance (ANOVA) was used followed by analysis of co-variance (ANCOVA) after adjustment for age, whereas the chi-squared test was applied to compare categorical variables when appropriate.

To evaluate the association between adverse cardiometabolic risk profiles and degree of frailty, multiple logistic regression was employed. Model 1 was adjusted for age and gender, and model 2 was further adjusted for significant variables in functional and co-morbidity assessment with p value less than 0.1. Because skeletal muscle mass plays an important role in muscle performance, including gait speed and grip strength, model 3 included all of the adjustments of model 2 plus ASM/ht². The results are expressed as ORs and 95% CIs. A p value less than 0.05 (two-sided) was considered statistically significant. All statistical analyses were performed by statistical software (SPSS 20.0, Chicago, IL).

Resutls

Overall, data of 1839 participants were retrieved for analysis. A total of 47.5% of the participants were men and the mean age was 63.9 years (range, 50.0–92.2 years; median 62.1 years, interquartile range 15.1 years; SD = 9.3 years). Table 1 shows baseline characteristics, functional measurements, and cardiometabolic and hormone parameters according to their status of frailty. The prevalence of frailty among study participants was 6.8% (n = 125). The most common components among the pre-frail and frail groups were slowness (42.5% and 94.4% individually), followed by weakness (38.6% and 88.8% individually) and low activity (35.1% and 84.8%, respectively), whereas exhaustion and weight loss were much less common. Compared to the robust group, frail participants had a higher prevalence of smoking and a history of hypertension, diabetes, and hyperlipidemia, and depressive symptoms (p value 0.028, <0.001, <0.001, 0.025, <0.001, respectively).

Table 1.

Participant Characteristics Categorized by Degree of Frailty

	Robust (n = 970, 52.7%)	Pre-frail (n = 744, 40.5%)	Frail (n = 125, 6.8%)	p value
Age	60.7 ± 7.5	66.3 ± 9.3	74.6 ± 9.2	<0.001
Men (n, %)	444 (45.8)	369 (49.6)	60 (48.0)	0.289
Frailty phenotype
Slowness (n, %)	0	316 (42.5)	118 (94.4)	<0.001
Weakness (n, %)	0	287 (38.6)	111 (88.8)	<0.001
Weight loss (n, %)	0	77 (10.3)	20 (16.0)	<0.001
Exhaustion (n, %)	0	35 (4.7)	38 (30.4)	<0.001
Low activity (n, %)	0	261 (35.1)	106 (84.8)	<0.001
Personal hstory
Alcohol consumption (n, %)				<0.001
Current	355 (36.6)	232 (31.2)	20 (16.0)
Quitted	70 (7.2)	63 (8.5)	17 (13.6)
Smoking (n, %)				0.028
Current	161 (16.6)	150 (20.2)	25 (20.0)
Quitted	103 (10.6)	103 (13.8)	19 (15.2)
CAD history (n, %)	43 (4.4)	43 (5.8)	9 (7.2)	0.260
HTN history (n, %)	336 (34.6)	354 (47.6)	72 (57.6)	<0.001
DM history (n, %)	130 (13.4)	139 (18.7)	40 (32.0)	<0.001
Hyperlipidemia history (n, %)	60 (6.2)	70 (9.4)	13 (10.4)	0.025
Depressive symptoms (n, %)TGDS-5 ≥ 2	8 (0.8)	22 (3.0)	18 (14.4)	<0.001

CAD, coronary artery disease; HTN, hypertension; DM, diabetes mellitus; TGDS, Taiwan Geriatric Depression Scale.

Before age adjustment, we found that traditional cardiovascular risk factors, such as SBP, glucose level, or insulin resistance, had a significantly positive association with status of frailty (p values all less than 0.001 for SBP, HbA1c, and HOMA-IR). In contrast, lipid profiles such as TC, LDL-C, and HDL-C levels were decreased along with status of frailty (p values <0.001, 0.022, and 0.001, respectively). BMI was not significantly associated with frailty in this study, whereas abdominal obesity was significantly and positively associated with frailty (p value <0.001). Carotid IMT and inflammatory markers such as homocysteine and hsCRP were increased with status of frailty as well (p values both <0.001). For hormone parameters, frailty was associated with lower IGF-1, DHEA-S in both genders, and FAI in men (p values all <0.001) (data not shown)

Table 2 shows the association between frailty with cardiometabolic and hormone parameters after adjustment for age. Frail subjects have significantly lower MNA and MMSE scores (p values both <0.001), lower skeletal muscle mass (p < 0.001 in men and 0.013 in women), and higher CCI values (p < 0.001). Adjusted for age, we found that blood pressure, LDL-C, uric acid, creatinine, and carotid IMT were no longer significantly associated with frailty. However, frailty was significantly associated with lower TC and HDL-C and higher hsCRP (p values 0.02, 0.022, 0.02, respectively). Glucose metabolism profiles, such as fasting blood glucose, HbA1c, and HOMA-IR, maintained a positive association with frailty (p values 0.006, <0.001, <0.001, respectively). For hormone parameters, the association of frailty with DHEA-S, IGF-1, and FAI became insignificant after age adjustment.

Table 2.

Cardiometabolic and Hormone Parameters Categorized by Degree of Frailty, Adjusted for Age

	Robust	Pre-frail	Frail	p value
Function measurements
MNA	27.4 ± 0.06	27.1 ± 0.06	25.5 ± 0.16	<0.001
MMSE	26.2 ± 0.1	25.4 ± 0.1	22.9 ± 0.3	<0.001
Education year	6.6 ± 0.1	5.9 ± 0.2	5.4 ± 0.4	0.001
Skeletal muscle mass
ASM/(height)²
Men	8.1 ± 0.04	7.8 ± 0.04	7.6 ± 0.11	<0.001
Women	6.4 ± 0.03	6.3 ± 0.04	6.1 ± 0.10	0.013
CCI	0.9 ± 0.04	1.1 ± 0.04	1.4 ± 0.1	<0.001
Cardiometabolic parameters
BMI (kg/m²)	24.8 ± 0.1	25.0 ± 0.1	24.9 ± 0.3	0.571
Abdominal circumference (cm)	84.4 ± 0.3	85.1 ± 0.4	86.2 ± 0.9	0.145
SBP (mmHg)	130.3 ± 0.6	132.0 ± 0.6	130.2 ± 1.6	0.119
DBP (mmHg)	79.9 ± 0.4	80.0 ± 0.5	80.1 ± 1.2	0.978
Fasting blood glucose (mg/dL)	100.5 ± 1.0	104.9 ± 1.1	107.4 ± 2.9	0.006
HbA1c (%)	6.0 ± 0.03	6.1 ± 0.04	6.4 ± 0.10	<0.001
HOMA-IR	1.5 ± 0.03	1.7 ± 0.04	1.8 ± 0.09	<0.001
TC (mg/dL)	197.3 ± 1.2	193.0 ± 1.3	190.0 ± 3.3	0.020
TGs (mg/dL)	120.5 ± 3.2	132.3 ± 3.6	125.5 ± 9.1	0.056
HDL-C (mg/dL)	55.6 ± 0.5	53.7 ± 0.5	53.3 ± 1.3	0.022
LDL-C (mg/dL)	118.8 ± 1.1	118.4 ± 1.2	117.7 ± 3.1	0.937
Uric acid (mg/dL)	5.9 ± 0.1	5.9 ± 0.1	5.8 ± 0.1	0.575
Creatinine (mg/dL)	0.8 ± 0.01	0.8 ± 0.01	0.8 ± 0.03	0.672
Homocysteine (mmol/L)	13.1 ± 0.2	13.8 ± 0.2	13.1 ± 0.6	0.066
hsCRP (mg/dL)	0.2 ± 0.02	0.2 ± 0.02	0.4 ± 0.05	0.020
Carotid IMT (mm)	0.7 ± 0.004	0.7 ± 0.005	0.7 ± 0.013	0.216
Hormone-related parameters
Total testosterone (ng/dL)
Men	483.4 ± 8.8	483.5 ± 9.4	460.6 ± 23.9	0.651
Women	25.5 ± 1.7	28.1 ± 2.0	20.1 ± 5.0	0.244
Free androgen index
Men	33.9 ± 0.5	35.1 ± 0.6	33.4 ± 1.5	0.224
Women	1.7 ± 0.1	2.0 ± 0.1	1.2 ± 0.3	0.072
DHEA-S (μg/dL)
Men	133.1 ± 3.4	132.6 ± 3.7	123.1 ± 9.3	0.605
Women	82.4 ± 2.2	80.6 ± 2.6	83.0 ± 6.6	0.855
IGF-1 (ng/mL)
Men	143.8 ± 2.7	137.9 ± 2.9	135.0 ± 7.4	0.287
Women	134.0 ± 2.3	129.1 ± 2.7	124.6 ± 6.8	0.267

MNA, Mini-Nutrition Assessment; MMSE, Mini-Mental State Exam; ASM, appendicular skeletal muscle mass; CCI, Charlson's Co-Morbidity Index; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; HbA1c, glycated hemoglobin; HOMA-IR, homeostasis model assessment of insulin resistance; TC, total cholesterol; TGs, triglycerides; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; hsCRP, high-sensitivity C-reactive protein; carotid IMT, carotid intima media thickness; DHEA-S, dehydroepiandrosterone sulfate; IGF-1, insulin-like growth factor.

The adjusted OR for pre-frailty and frailty by cardiovascular risk factors is summarized in Table 3. For personal history, current smoking status was associated with pre-frailty in a fully adjusted model (OR = 1.60, 95% CI 1.17–2.20, p = 0.003), but not frailty. A medical history of hypertension, diabetes mellitus (DM), hyperlipidemia, or coronary artery disease was not significantly associated with the status of frailty in this study. For cardiometabolic parameters, abdominal obesity, HOMA-IR, and hsCRP were associated with frailty consistently across three models; the OR for frailty was 3.57 (95% CI 1.88–6.78, p < 0.001), 1.30 (95% CI 1.02–1.66, p = 0.032), and 1.66 (95% CI 1.10–2.49, p = 0.016), respectively, in a fully adjusted model. These associations were even stronger in an adjusted model 2 compared to model 1, and the trend was more evident in the frail group compared to the pre-frail group. Higher TC was associated with lower prevalence of frailty consistently (OR = 0.44, 95% CI 0.22–0.89, p = 0.023 in model 3) as well, whereas other adverse lipid profiles did not have significant correlations with frailty.

Table 3.

Multiple Logistic Regression on Degree of Frailty

	Model 1 ^a				Model 2 ^b				Model 3 ^c
	Pre-frail OR (95% CI)	p value	Frail OR (95% CI)	p value	Pre-frail OR (95% CI)	p value	Frail OR (95% CI)	p value	Pre-frail OR (95% CI)	p value	Frail OR (95% CI)	p value
Personal history
Hypertension history	1.22 (0.95–1.56)	0.127	1.06 (0.64–1.76)	0.824	1.20 (0.93–1.55	0.165	1.16 (0.66–2.05)	0.600	1.22 (0.94–1.58)	0.131	1.26 (0.70–2.26)	0.446
DM history	0.91 (0.63–1.32)	0.627	1.54 (0.81–2.91)	0.188	0.79 (0.53–1.18)	0.251	0.96 (0.45–2.04)	0.913	0.85 (0.57–1.27)	0.417	1.01 (0.47–2.20)	0.971
Hyperlipidemia history	1.23 (0.80–1.88)	0.346	0.83 (0.36–1.93)	0.664	1.19 (0.77–1.83)	0.446	0.73 (0.30–1.82)	0.504	1.12 (0.72–1.74)	0.614	0.70 (0.27–1.79)	0.453
CAD history	0.95 (0.58–1.58)	0.849	0.89 (0.37–2.11)	0.787	0.86 (0.51–1.43)	0.552	0.64 (0.25–1.65)	0.354	0.87 (0.52–1.46)	0.596	0.69 (0.26–1.81)	0.452
Current smoking	1.43 (1.06–1.93)	0.019^*	1.43 (0.76–2.77)	0.270	1.15 (1.01–1.96)	0.016^*	1.58 (0.78–3.21)	0.209	1.60 (1.17–2.20)	0.003^**	1.65 (0.75–3.60)	0.210
Current alcohol consumption	0.85 (0.67–1.08)	0.191	0.37 (0.20–0.69)	0.002^**	0.92 (0.71–1.17)	0.483	0.51 (0.25–1.01)	0.054	0.96 (0.74–1.25)	0.784	0.51 (0.24–1.07)	0.073
Cardiometabolic parameters
Higher abdominal circumference	1.18 (0.94–1.49)	0.156	1.77 (1.07–2.93)	0.026^*	1.26 (0.99–1.60)	0.062	3.80 (2.03–7.11)	<0.001^***	1.21 (0.95–1.55)	0.123	3.57 (1.88–6.78)	<0.001^***
High SBP	1.20 (0.93–1.54)	0.155	1.04 (0.64–1.71)	0.867	1.18 (0.92–1.52)	0.199	0.98 (0.57–1.71)	0.954	1.18 (0.91–1.53)	0.205	0.88 (0.50–1.55)	0.652
High TC	0.80 (0.56–1.12)	0.192	0.44 (0.23–0.82)	0.010^*	0.82 (0.58–1.16)	0.268	0.45 (0.23–0.90)	0.023^*	0.81 (0.57–1.15)	0.241	0.44 (0.22–0.89)	0.023^*
High LDL–C	1.02 (0.80–1.31)	0.860	1.60 (0.93–2.75)	0.087	1.02 (0.80–1.31)	0.875	1.19 (0.65–2.20)	0.578	0.97 (0.75–1.25)	0.810	1.22 (0.65–2.30)	0.533
Low HDL–C	0.96 (0.74–1.26)	0.774	0.73 (0.42–1.27)	0.262	0.93 (0.71–1.22)	0.586	0.56 (0.30–1.02)	0.057	0.95 (0.72–1.25)	0.712	0.55 (0.29–1.04)	0.066
High TGs	1.23 (0.94–1.62)	0.129	1.28 (0.72–2.29)	0.404	1.26 (0.96–1.66	0.098	1.47 (0.77–2.78)	0.241	1.31 (0.99–1.73)	0.062	1.36 (0.69–2.68)	0.379
HOMA-IR	1.15 (1.00–1.32)	0.047^*	1.25 (1.00–1.57)	0.051	1.16 (1.00–1.34)	0.045^*	1.33 (1.05–1.68)	0.017^*	1.14 (0.99–1.31)	0.080	1.30 (1.02–1.66)	0.032^*
Homocysteine	1.02 (0.99–1.04)	0.132	0.98 (0.94–1.03)	0.377	1.02 (1.00–1.04)	0.111	0.99 (0.95–1.03)	0.632	1.02 (0.99–1.04)	0.104	0.99 (0.94–1.04)	0.686
hsCRP	1.02 (0.78–1.32)	0.905	1.47 (0.99–2.20)	0.059	1.01 (0.78–1.31)	0.942	1.71 (1.15–2.56	0.009^**	1.00 (0.76–1.30)	0.971	1.66 (1.10–2.49)	0.016^*

Model 1, adjusted age and gender

Model 2, adjusted age, gender, MMSE, MNA, depressive symptoms, and CCI.

Model 3, adjusted age, gender, MMSE, MNA, depressive symptoms, CCI, and ASM/ht².

p value <0.05.

p value <0.01.

***

p value <0.001.

OR, odds ratio; CI, confidence interval; CAD, coronary artery disease; SBP, systolic blood pressure; TC, total cholesterol; LDL-C, low-density lipoprotein cholesterol; HDL-C, high-density lipoprotein cholesterol; TGs, triglycerides; HOMA-IR, homeostasis model assessment of insulin resistance; hsCRP, high-sensitivity C-reactive protein; MMSE, Mini-Mental State Exam; MNA, Mini-Nutrition Assessment; CCI, Charlson's Co-Morbidity Index; ASM, appendicular skeletal muscle mass.

Discussion

Previous studies have shown the association of frailty with traditional cardiometabolic parameters, such as obesity, high blood pressure, poor renal function, and carotid IMT.^22,28
–30 In the current study, these factors and high LDL-C levels were no longer associated with frailty after adjustment for age, whereas abdominal obesity, higher insulin resistance, and inflammation remained strongly associated with the status of frailty. The relationship between BMI and frailty gained extensive research interest in the past. Longitudinal studies have demonstrated mid-life overweight or obesity increased the risk of frailty in old age.^31,32 Nevertheless, in daily practice, in which the trajectory of weight was often unclear, the relationship between BMI and frailty in older people became more complicated. Hubbard et al. have reported the U-shaped relationship between BMI and frailty in the England Longitudinal Study of Ageing.³³ Sheehan et al.³⁴ have proposed a linear relationship between BMI and frailty. In our analysis, although higher BMI was not associated with the status of frailty, abdominal obesity strongly increased the prevalence of frailty. The impact was not attenuated by adjusting skeletal muscle mass, implying the independent effect of abdominal adiposity on frailty. Woo et al. and Zoico et al. have reported similar findings, but they only examined the association of obesity and physical function instead of frailty.^35,36 Although overweight or central obesity have been demonstrated to be a survival benefit among the older people,^37
–39 the benefit disappeared when earlier weight trajectories taken into account.⁴⁰

Higher TC or LDL-C levels are well-known risk factors for cardiovascular disease and mortality,⁴¹ but the effect was lessened in older people, especially after adjustment for potential confounding factors.^42,43 In particular, low TC levels eventually increased overall mortality in the oldest old.⁴⁴ In this study, the LDL-C level was not associated with frailty after adjustment for age whereas higher TC levels were associated with a lower likelihood for frailty. Though TC levels lower than 160 mg/dL have been considered a marker of malnutrition, the effect was only lessened slightly after adjustment for MNA score. The results also echoed the debate of statin use in the older people, whereas current evidence did not support routine use of statins for those who were over 75 without clinical atherosclerotic cardiovascular disease.¹⁹ It is possible that frailty and functional status played a mediator role between low TC levels and higher overall mortality in older people.⁴⁵

Among the cardiometabolic factors evaluated in this study, HOMA-IR was significantly associated with pre-frailty and frailty, suggesting the close relationship of insulin resistance in the pathophysiology of frailty, and was compatible with the previous report.⁴⁶ We further demonstrated the effect was only slightly attenuated after adjustment for skeletal muscle mass, although low muscle mass and myokines have been regarded to be associated with impaired insulin sensitivity.^47,48 It has been proposed that the correlation between insulin resistance and frailty was only seen in centrally obese participants⁴⁹ and hyperglycemia increased risk of dementia in non-DM and DM patients.⁵⁰ However, the association of insulin resistance and frailty status in this study remained unchanged after adjustment for abdominal circumference and MMSE. Therefore, results of this study suggested that HOMA-IR was associated with frailty independent of body composition and cognitive function.

Inflammatory markers have been reported to be closely associated with cardiovascular aging⁵¹ and frailty.^52
–54 The association between higher hsCRP and frailty in this study was even stronger after adjusting for cognitive status, nutrition, depressive symptoms, and multi-morbidities. However, homocysteine, a postulated risk factor for cardiovascular disease,⁵⁵ was not correlated with frailty. Notably, high hsCRP was significantly associated with frailty but not pre-frailty, which implied the inflammation was more significantly increased from pre-frailty to frailty comparing to robust to pre-frailty. The relationship between insulin resistance, inflammation, and abdominal obesity with frailty were still significant after adjustment for other cardiometabolic markers, implying that the relationship between these factors and frailty were independent of each other rather than a single biological phenomenon.

Despite all the efforts that went into the study, there were several limitations. First, the study has a cross-sectional design, so no causal relationship can be established. Second, the participants from our community cohort were younger, relatively physically active, and had a lower prevalence of cardiovascular disease compared to other studies,^22,46 which may underestimate the effect on frailty. Third, the diagnosis of medical history was based on self-report plus related medication instead of medical record review; nevertheless, it has been validated in collecting medical diagnosis in previous work.⁵⁶

In conclusion, we clarified that most cardiometabolic and hormonal parameters were not associated with frailty after adjustment of age. Abdominal obesity, insulin resistance, and inflammation were still highly correlated with degree of frailty, and the effect was independent of functional measurement and decline of skeletal muscle mass. An integrated approach targeting cardiometabolic aging and development of frailty is of great importance in clinical practice.

Footnotes

Acknowledgments

We express our gratitude to the staff in Center for Geriatrics and Gerontology of Taipei Veterans General Hospital and Department of Family Medicine of Taipei Veterans General Hospital Yuanshan Branch and to all the participants for their assistance. This study was supported by the Aging and Health Research Center, National Yang Ming University; Center for Geriatrics and Gerontology, Taipei Veterans General Hospital; as well as the Ministry of Science and Technology of Taiwan (MOST 103-2633-B-400-002; and MOST 101-2314-B-010-008).

Author Disclosure Statement

No competing financial interests exist.

References

Bergman

, Ferrucci

, Guralnik

, Hogan

, Hummel

, Karunananthan

, Wolfson

. Frailty: An emerging research and clinical paradigm—issues and controversies. J Gerontol A Biol Sci Med Sci, 2007; 62:731–737.

Abellan van Kan

, Rolland

, Morley

, Vellas

. Frailty: Toward a clinical definition. J Am Med Dir Assoc, 2008; 9:71–72.

Bagshaw

, Stelfox

, Johnson

, McDermid

, Rolfson

, Tsuyuki

, Ibrahim

, Majumdar

. Long-term association between frailty and health-related quality-of-life among survivors of critical illness: A prospective multicenter cohort study. Crit Care Med, 2015; 43:973–982.

Rouge Bugat

, Cestac

, Oustric

, Vellas

, Nourhashemi

. Detecting frailty in primary care: A major challenge for primary care physicians. J Am Med Dir Assoc, 2012; 13:669–672.

Chen

, Rockwood

. Planning for frailty. J Clin Gerontol Geriatr, 2012; 3:3–4.

Lee

, Lee

, Chan

. Interventions targeting geriatric frailty: A systemic review. J Clin Gerontol Geriatr, 2012; 3:47–52.

Sergi

, Veronese

, Fontana

, De Rui

, Bolzetta

, Zambon

, Corti

, Baggio

, Toffanello

, Crepaldi

, Perissinotto

, Manzato

. Pre-frailty and risk of cardiovascular disease in older people men and women: The Pro.V.A. Study. J Am Coll Cardiol, 2015; 65:976–983.

Afilalo

, Karunananthan

, Eisenberg

, Alexander

, Bergman

. Role of frailty in patients with cardiovascular disease. Am J Cardiol, 2009; 103:1616–1621.

Afilalo

, Alexander

, Mack

, Maurer

, Green

, Allen

, Popma

, Ferrucci

, Forman

. Frailty assessment in the cardiovascular care of older adults. J Am Coll Cardiol, 2014; 63:747–762.

10.

Corti

, Salive

, Guralnik

. Serum albumin and physical function as predictors of coronary heart disease mortality and incidence in older persons. J Clin Epidemiol, 1996; 49:519–526.

11.

Sujino

, Tanno

, Nakano

, Funada

, Hosoi

, Senbonmatsu

, Nishimura

. Impact of hypoalbuminemia, frailty, and body mass index on early prognosis in older patients (≥85 years) with ST-elevation myocardial infarction. J Cardiol, 2015; 66:263–268.

12.

Lupon

, Gonzalez

, Santaeugenia

, Altimir

, Urrutia

, Mas

, Diez

, Pascual

, Cano

, Valle

. Prognostic implication of frailty and depressive symptoms in an outpatient population with heart failure. Rev Esp Cardiol, 2008; 61:835–842.

13.

Lee

, Buth

, Martin

, Yip

, Hirsch

. Frail patients are at increased risk for mortality and prolonged institutional care after cardiac surgery. Circulation, 2010; 121:973–978.

14.

Fiatarone

, O'Neill

, Ryan

, Clements

, Solares

, Nelson

, Roberts

, Kehayias

, Lipsitz

, Evans

. Exercise training and nutritional supplementation for physical frailty in very elderly people. N Engl J Med, 1994; 330:1769–1775.

15.

Savela

, Koistinen

, Stenholm

, Tilvis

, Strandberg

, Pitkala

, Salomaa

, Strandberg

. Leisure-time physical activity in midlife is related to old age frailty. J Gerontol A Biol Sci Med Sci, 2013; 68:1433–1438.

16.

Newman

, Gottdiener

, McBurnie

, Hirsch

, Kop

, Tracy

, Walston

, Fried

, Cardiovascular Health Study Research G. Associations of subclinical cardiovascular disease with frailty. J Gerontol A Biol Sci Med Sci, 2001; 56:M158–M166.

17.

Strandberg

, Pitkala

, Tilvis

, O'Neill

, Erkinjuntti

. Geriatric syndromes—vascular disorders?. Ann Med, 2013; 45:265–273.

18.

James

, Oparil

, Carter

, Cushman

, Dennison-Himmelfarb

, Handler

, Lackland

, LeFevre

, MacKenzie

, Ogedegbe

, Smith

Jr ., Svetkey

, Taler

, Townsend

, Wright

Jr ., Narva

, Ortiz

2014 evidence-based guideline for the management of high blood pressure in adults: Report from the panel members appointed to the Eighth Joint National Committee (JNC 8). JAMA, 2014; 311:507–520.

19.

Stone

, Robinson

, Lichtenstein

, Bairey Merz

, Blum

, Eckel

, Goldberg

, Gordon

, Levy

, Lloyd-Jones

, McBride

, Schwartz

, Shero

, Smith

Jr ., Watson

, Wilson

, American College of Cardiology/American Heart Association Task Force on Practice G. 2013 ACC/AHA guideline on the treatment of blood cholesterol to reduce atherosclerotic cardiovascular risk in adults: A report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. J Am Coll Cardiol, 2014; 63:2889–2934.

20.

Kirkman

, Briscoe

, Clark

, Florez

, Haas

, Halter

, Huang

, Korytkowski

, Munshi

, Odegard

, Pratley

, Swift

, Consensus Development Conference on Diabetes and Older Adults. Diabetes in older adults: A consensus report. J Am Geriatr Soc, 2012; 60:2342–2356.

21.

Muller

, Smulders

, de Leeuw

, Stehouwer

. Treatment of hypertension in the oldest old: A critical role for frailty?. Hypertension. 2014; 63:433–441.

22.

Ramsay

, Arianayagam

, Whincup

, Lennon

, Cryer

, Papacosta

, Iliffe

, Wannamethee

. Cardiovascular risk profile and frailty in a population-based study of older British men. Heart, 2014; doi:10.1136/heartjnl-2014-306472.

23.

Ricci

, Pessoa

, Ferriolli

, Dias

, Perracini

. Frailty and cardiovascular risk in community-dwelling older people: A population-based study. Clin Interv Aging, 2014; 9:1677–1685.

24.

Rinaldi

, Mecocci

, Benedetti

, Ercolani

, Bregnocchi

, Menculini

, Catani

, Senin

, Cherubini

. Validation of the five-item geriatric depression scale in older people subjects in three different settings. J Am Geriatr Soc, 2003; 51:694–698.

25.

National High Blood Pressure Education Program.. The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure. Report No.: 04-5230. Bethesda, MD: National Heart, Lung, and Blood Institute, 2004.

26.

Chen

, Liu

, Woo

, Assantachai

, Auyeung

, Bahyah

, Chou

, Chen

, Hsu

, Krairit

, Lee

, Liang

, Limpawattana

, Lin

, Peng

, Satake

, Suzuki

, Won

, Wu

, Zhang

, Zeng

, Akishita

, Arai

. Sarcopenia in Asia: Consensus report of the Asian Working Group for Sarcopenia. J Am Med Dir Assoc, 2014; 15:95–101.

27.

Fried

, Tangen

, Walston

, Newman

, Hirsch

, Gottdiener

, Seeman

, Tracy

, Kop

, Burke

, McBurnie

, Cardiovascular Health Study Collaborative Research G. Frailty in older adults: Evidence for a phenotype. J Gerontol A Biol Sci Med Sci, 2001; 56:M146–M156.

28.

Fattori

, Santimaria

, Alves

, Guariento

, Neri

. Influence of blood pressure profile on frailty phenotype in community-dwelling elders in Brazil—FIBRA study. Arch Gerontol Geriatr, 2013; 56:343–349.

29.

Bastos-Barbosa

, Ferriolli

, Coelho

, Moriguti

, Nobre

, Lima

. Association of frailty syndrome in the older people with higher blood pressure and other cardiovascular risk factors. Am J Hypertens, 2012; 25:1156–1161.

30.

Avila-Funes

, Meillon

, Gonzalez-Colaco Harmand

, Tzourio

, Dartigues

, Amieva

. Association between frailty and carotid central structure changes: The Three-City Study. J Am Geriatr Soc, 2014; 62:1906–1911.

31.

Strandberg

, Sirola

, Pitkala

, Tilvis

, Strandberg

, Stenholm

. Association of midlife obesity and cardiovascular risk with old age frailty: A 26-year follow-up of initially healthy men. Int J Obes (Lond), 2012; 36:1153–1157.

32.

Stenholm

, Strandberg

, Pitkala

, Sainio

, Heliovaara

, Koskinen

. Midlife obesity and risk of frailty in old age during a 22-year follow-up in men and women: The Mini-Finland Follow-up Survey. J Gerontol A Biol Sci Med Sci, 2014; 69:73–78.

33.

Hubbard

, Lang

, Llewellyn

, Rockwood

. Frailty, body mass index, and abdominal obesity in older people. J Gerontol A Biol Sci Med Sci, 2010; 65:377–381.

34.

Sheehan

, O'Connell

, Cunningham

, Crosby

, Kenny

. The relationship between increased body mass index and frailty on falls in community dwelling older adults. BMC Geriatr, 2013; 13:132.

35.

Woo

, Leung

, Kwok

. BMI, body composition, and physical functioning in older adults. Obesity (Silver Spring), 2007; 15:1886–1894.

36.

Zoico

, Di Francesco

, Mazzali

, Zivelonghi

, Volpato

, Bortolani

, Dioli

, Coin

, Bosello

, Zamboni

. High baseline values of fat mass, independently of appendicular skeletal mass, predict 2-year onset of disability in older people subjects at the high end of the functional spectrum. Aging Clin Exp Res, 2007; 19:154–159.

37.

Chen

, Peng

, Liu

, Lin

, Chen

, Lan

, Chang

. Body mass index, health status, and mortality of older Taiwanese men: Overweight good, underweight bad, obesity neutral. J Am Geriatr Soc, 2013; 61:2233–2234.

38.

Flicker

, McCaul

, Hankey

, Jamrozik

, Brown

, Byles

, Almeida

. Body mass index and survival in men and women aged 70 to 75. J Am Geriatr Soc, 2010; 58:234–241.

39.

Auyeung

, Lee

, Leung

, Kwok

, Leung

, Woo

. Survival in older men may benefit from being slightly overweight and centrally obese—a 5-year follow-up study in 4,000 older adults using DXA. J Gerontol A Biol Sci Med Sci, 2010; 65:99–104.

40.

Strandberg

, Stenholm

, Strandberg

, Salomaa

, Pitkala

, Tilvis

. The “obesity paradox,” frailty, disability, and mortality in older men: A prospective, longitudinal cohort study. Am J Epidemiol, 2013; 178:1452–1460.

41.

Anderson

, Odell

, Wilson

, Kannel

. Cardiovascular disease risk profiles. Am Heart J, 1991; 121(1 Pt 2):293–298.

42.

Krumholz

, Seeman

, Merrill

, Mendes de Leon

, Vaccarino

, Silverman

, Tsukahara

, Ostfeld

, Berkman

. Lack of association between cholesterol and coronary heart disease mortality and morbidity and all-cause mortality in persons older than 70 years. JAMA, 1994; 272:1335–1340.

43.

Schatz

, Masaki

, Yano

, Chen

, Rodriguez

, Curb

. Cholesterol and all-cause mortality in older people people from the Honolulu Heart Program: A cohort study. Lancet, 2001; 358:351–355.

44.

Weverling-Rijnsburger

, Blauw

, Lagaay

, Knook

, Meinders

, Westendorp

. Total cholesterol and risk of mortality in the oldest old. Lancet, 1997; 350:1119–1123.

45.

Tilvis

, Valvanne

, Strandberg

, Miettinen

. Prognostic significance of serum cholesterol, lathosterol, and sitosterol in old age; a 17-year population study. Ann Med, 2011; 43:292–301.

46.

Barzilay

, Blaum

, Moore

, Xue

, Hirsch

, Walston

, Fried

. Insulin resistance and inflammation as precursors of frailty: The Cardiovascular Health Study. Arch Intern Med, 2007; 167:635–641.

47.

Eckardt

, Gorgens

, Raschke

, Eckel

. Myokines in insulin resistance and type 2 diabetes. Diabetologia, 2014; 57:1087–1099.

48.

Volpi

, Mittendorfer

, Rasmussen

, Wolfe

. The response of muscle protein anabolism to combined hyperaminoacidemia and glucose-induced hyperinsulinemia is impaired in the older people. J Clin Endocrinol Metab, 2000; 85:4481–4490.

49.

Goulet

, Hassaine

, Dionne

, Gaudreau

, Khalil

, Fulop

, Shatenstein

, Tessier

, Morais

. Frailty in the older people is associated with insulin resistance of glucose metabolism in the postabsorptive state only in the presence of increased abdominal fat. Exp Gerontol, 2009; 44:740–744.

50.

Crane

, Walker

, Hubbard

, Li

, Nathan

, Zheng

, Haneuse

, Craft

, Montine

, Kahn

, McCormick

, McCurry

, Bowen

, Larson

. Glucose levels and risk of dementia. N Engl J Med, 2013; 369:540–548.

51.

, Xia

, Kalionis

, Wan

, Sun

. The role of oxidative stress and inflammation in cardiovascular aging. Biomed Res Int, 2014; 2014:615312.

52.

Puts

, Visser

, Twisk

, Deeg

, Lips

. Endocrine and inflammatory markers as predictors of frailty. Clin Endocrinol (Oxf), 2005; 63:403–411.

53.

Gale

, Baylis

, Cooper

, Sayer

. Inflammatory markers and incident frailty in men and women: The English Longitudinal Study of Ageing. Age (Dordr). 2013; 35:2493–2501.

54.

Reiner

, Aragaki

, Gray

, Wactawski-Wende

, Cauley

, Cochrane

, Kooperberg

, Woods

, LaCroix

. Inflammation and thrombosis biomarkers and incident frailty in postmenopausal women. Am J Med, 2009; 122:947–954.

55.

Wierzbicki

AS.

Homocysteine and cardiovascular disease: A review of the evidence. Diab Vasc Dis Res, 2007; 4:143–150.

56.

Rolland

, Lauwers-Cances

, Cesari

, Vellas

, Pahor

, Grandjean

. Physical performance measures as predictors of mortality in a cohort of community-dwelling older French women. Eur J Epidemiol, 2006; 21:113–122.