Table 1.
Studies suggesting AM frail phenotype.
| Study | Patients | Aim to | Main Findings |
|---|---|---|---|
| Chao CT, et al., prospective, Taiwan, 2020 [14] | 2119 patients with and 8432 without hypoglycaemia (control), mean (SD) age 65.9 (14) Y. | Examine whether hypoglycaemia increases the risk of frailty. | Patients with compared to those without hypoglycaemia had: A. Less obesity (1.4% vs. 2.4%, p = 0.006). B. Less hyperlipidaemia (55.5% vs. 63.0%, p < 0.001). C. Higher mean (SD) CCIS, 4.2 (2.4) vs. 3.4 (2.2), p < 0.001. D. Higher mean (SD) aDCSI, 1.3 (1.5) vs. 0.8 (1.2), p < 0.001. D. Less patients using statin (36.4% vs. 44.8%, p < 0.001) or fibrate (12.4% vs. 16.1%, p < 0.001). E. No difference in hypoglycaemic therapy use. |
| Chao CT, et al., prospective, Taiwan, 2018 [15] | 560,795 patients with DM, mean (SD) age 56.4 (13.8) Y. | Examine the association of pre-frailty and frailty with mortality. | Frail (≥3 components in FRAIL scale) compared to non-frail patients had: A. Higher mean (SD) age, 75.1 (11.2) vs. 54.8 (13.2), p < 0.01. B. Higher mean (SD) aDCSI, 1.4 (1.5) vs. 0.2 (0.6), p < 0.01. C. Higher mean (SD) CCIS, 5.4 (2.4) vs. 1.7 (1.4), p < 0.01. D. More prevalence of hypertension, CKD, CLD, COPD, AF, CAD, PVD, gout, OA, malignancy, osteoporosis, mental illness, and CVD, all p < 0.01. E. Less hyperlipidaemia, 31.9% vs. 41.2%, p < 0.01. F. Less obesity, 0.9% vs. 1.6%, p < 0.01. G. More hypoglycaemia, 1.2% vs. 0.1%, p < 0.01. H. Less statin and fibrate use, p < 0.01. I. More insulin users, 7.7% vs. 4.8%, p < 0.01. J. Less oral hypoglycaemic use, p < 0.01. |
| Nguyen TTH, et al., cross-sectional, Vietnam, 2019 [16] | 158 subjects with DM, mean (SD) age 69.52 (6.76) Y. | Assess nutritional status using the Mini-Nutrition Assessment Short Form tool, and describe the relationship among related factors. | Malnourished compared to well-nourished had: A. Older (≥80 Y) subjects, 11.5% vs. 6.6%, p = 0.27. B. More underweight (10.2% vs. 0.0%) and less overweight subjects (30.6 vs. 71.6), p < 0.001. C. More exhausted subjects, 30.8% vs. 11.0%, p < 0.01. D. More cognitive impairment, 53.1% vs. 27.5%, p = 0.002. E. More frail subjects, 20.4% vs. 2.8%, p < 0.001. F. Malnutrition risk increased by frailty (OR 9.06, 95% CI 2.37 to 34.65, p = 0.001) and cognitive impairment, 2.98, 1.48 to 6.0, p = 0.002. |
| Yanagita I, et al., cross-sectional, prospective, Japan, 2018 [17] | 132 hospitalised patients with DM, mean (SD) age 78.3 (8.0) Y. | Evaluate frailty risk factors including HbA1c using CFS. | Frail compared to non-frail had: A. Higher mean (SD) age, 82.78 (8.16) vs. 75.17 (6.20) Y, p < 0.001. B. Lower mean (SD) HbA1c, 6.6 (0.93) vs. 7.27 (1.04)%, p < 0.001). B. Lower mean (SD) albumin, 35.62 (4.82) vs. 41.73 g/L, p < 0.001). C. Lower mean (SD) HDL-cholesterol, 1.23 (0.39) vs. 1.46 9 (0.43) mmol/L, p < 0.01). D. Lower mean (SD) Hb, 117.89 (17.28) vs. 132.38 (17.02) g/L, p < 0.001. E. Lower mean (SD) SBP (KPa) 16.81 (2.12) vs. 17.60 (1.92), p < 0.05). F. Lower mean (SD) body weight, 51.64 (12.72) vs. 59.17 (10.94) Kg, p < 0.001). G. Fewer patients using statin or fibrate (%) 34.54 vs. 54.54, p < 0.05. H. No difference in hypoglycaemic therapy use. |
| McAlister FA, et al., retrospective, UK, 2018 [18] | 292,170 subjects with DM, mean (SD) age 61.7 (15.6) Y. | Examine glycaemic control across health states. | Frail compared to non-frail subjects had: A. Higher mean (SD) age, 72.2 (11.2) vs. 60.8 (15.5), p < 0.0001. B. Higher mean (SD) CCIS, 2.73 (1.6) vs. 1.04 (0.86), p < 0.0001. C. Lower mean (SD) BMI, 26.0 (6.8) vs. 30.17 (6.9), p < 0.0001. D. Lower mean (SD) HbA1c, 7.1 (1.46) vs. 7.38 (1.64)%, p < 0.0001. E. Lower total cholesterol (p = 0.007), triglycerides (p = 0.05) and LDL-cholesterol (0.001). F. More comorbidities and diabetes complications, p < 0.0001. |
| De Decker L, et al., cross-sectional, France, 2017 [19] | 1552 patients with DM, mean (SD) age 86.4 (4.4) Y. | Determine the association between hypoglycaemia and a high burden of comorbidities. | Patients with compared to those without hypoglycaemia had the mean (SD): A. Lower body weight (Kg) 65.80 (14.7) vs. 69.5 (24.5), p = 0.004). B. Lower eGFR (ml/min) 46.1 (20.8) vs. 51.9 (26.1), p < 0.001. C. Higher CCIS 4.7 (2.3) vs. 3.8 (2.1), p < 0.001. C. Greater level of dependency (p < 0.001). D. Higher prevalence of CV disease (%) 68.6% vs. 54.3%, p < 0.001. E. Higher prevalence of dementia (%) 67% vs. 59.3%, p < 0.006. F. More use of insulin, p < 0.001 G. More SMBG (%) 84.6% vs. 67.8, p < 0.001. H. Lower use of SU, metformin, p < 0.001 and glinides, p = 0.002. |
| Cacciatore F, et al., prospective, Italy, 2013 [20] | 1288 subjects, mean (SD) 72.4 (6.3) Y, F/U 12 Y. | Examine the predictive role of frailty on long-term mortality. | Severely frail compared to non-frail patients had: A. Higher mean (SD) age, 78.9 (6.0) vs. 72.1 (4.7), p < 0.001. B. Lower mean (SD) BMI, 26.4 (5.1) vs. 27.0 (3.6), p = 0.087. C. Lower mean (SD) waist circumference, 94.8 (19.2) vs. 98.8 (14.9), p = 0.672. D. Higher mean (SD) CCIS, 4.8 (2.2) vs. 2.2 (1.6), p < 0.001. E. More insulin users, 26.0% vs. 13.4%, p = 0.004. F. More hypoglycaemic drug users 87.2% vs. 73.1, p = 0.008. G. More comorbidities of CHF, CKD, low MMSE, high GDS, disability, and mortality, all p < 0.001. |
| Abdelhafiz AH, case series, UK, 2014 [21] | 8.0 patients with type 2 DM, mean (SD) age 86.5 (3.2) Y and tight glycaemic control. | Describe whether hypoglycaemic therapy can be withdrawn in patients with HbA1c ≤6.0% or having recurrent hypoglycaemia. | Mean (SD) HbA1c was: A. 6.2% (0.8) at the point of medication withdrawal. B. 6.5% (0.7) after one year of follow-up, p = 0.4. At the point of medication withdrawal compared to the initial point of starting medication: A. Increased mean number of comorbidities 6.8 vs. 4.1, p = 0.002. B. Decreased mean weight 75.4 vs. 88.0 Kg, p = 0.003. C. Increased mean number of medications 10.1 vs. 6.4, p = 0.01. D. 50.0% of patients had a new diagnosis of dementia. |
| Sjoblom P, et al., observational, Sweden, 2008 [22] | 32.0 nursing home patients with DM, mean (SD) age 84.4 (6.8) Y. | Explore the feasibility of hypoglycaemic medication withdrawal in patients with HbA1c ≤6.0%. | Intervention compared to the non-intervention group had: A. Lower mean (SD) BMI, 25.6 (4.5) vs. 26.5 (5.1). B. Lower eGFR 50 vs. 55 mL/min/1.73 m2. C. Longer mean (SD) duration of DM, 10.6 (8.9) vs. 9.0 (7.4) Y. D. Lower mean (SD) HbA1c 5.2 (0.4) vs. 7.1 (1.6). |
| Morita T, et al., prospective, Japan, 2017 [23] | 184 patients with diabetes aged 65–94 Y, F/U 5 Y. | Investigate if low HbA1c is associated with risk of support/care need certification. | 42 (22.8%) patients developed disability defined as a requirement of first support/care-need certification. Compared to patients with HbA1c ≥ 7.0%, patients with HbA1c < 0.6% were: A. Older, mean (SD) age 77.5 (6.5) vs. 75.1 (6.3) Y, and had more people aged >75 Y, 68.5% vs. 54.5%, Kruskal–Wallis p < 0.05/3 and 0.20/3 respectively. B. Had less insulin use 37.8% vs. 63.6%, p < 0.20/3. C. Had less dyslipidaemia 37.0% vs. 59.1%, p < 0.05/3. D. Had more hypoalbuminaemia 14.8% vs. 6.8%, p < 0.20/3. |
| Adame Perez SI, et al., cross-sectional, Canada, 2019 [24] | 41 subjects with DM, median (IQR) age 70.0 (65.0–74.0) Y. | Compare differences in body composition by frailty status. | Frail compared to non-frail patients had: A. Lower mean (SD) ASMI (kg/m2), 6.8 (1.0) vs. 7.7 (0.9), p = 0.02. B. More patients with low lean body mass, 57.1% vs. 14.7%, p = 0.01. C. Higher mean (SD) comorbidities, 6.0 (2.0) vs. 4.0 (2.0), p = 0.03. C. No difference in body weight, BMI, HbA1c, hypoglycaemic therapies, or hypoglycaemic episodes. |
| Kitamura A, et al., prospective, Japan, 2019 [25] | 1271 subjects, mean (SD) age 71.0 (5.6) Y, 176 had DM. | Clarify risks of death and disability. | A. Frail compared to non-frail were older 72.5 vs. 69.6 Y, p = 0.05, more patients had hypoalbuminaemia, 20.7 vs. 5.9%, p = 0.02 and lower BMI 20.8 vs. 2.0%, p = 0.06. B. There was no difference in HbA1c, lipid profile, average BMI, hypoglycaemic medications, and comorbidities. |
| Thorpe CT, retrospective, US, 2015 [26] | 15,880 patients ≥ 65.0 years old with DM and dementia. | Examine: A. Risk factors for tight glycaemic control B. Medications associated with the risk of hypoglycaemia. |
A. 52.0% of patients had tight glycaemic control (HbA1c < 7.0%). B. Factors associated with tight control were: 1. Older age (75.0–84.0 years, OR 1.16, 95.0% CI 1.07 to 1.126, p = 0.001, ≥75.0 years, (1.13, 1.02 to 1.125, p = 0.021. 2. Heart valve disease (OR 1.16, 95.0% CI 1.01 to 1.32, p = 0.033), chronic lung disease OR 1.10, 95.0% CI 1.01 to 1.21, p = 0.038), and deficiency anaemia (OR 1.12, 95.0% CI 1.02 to 1.22, p = 0.016). 3. Weight loss (OR 1.36, 95.0% CI 1.09 to 1.69, p = 0.006). C. Among tightly controlled patients, 75.0% used SU and/or insulin. |
| Sussman JB, et al., retrospective, US, 2015 [27] | 179,991 patients > 70.0 years old on active treatment for DM. | Examine the rate of medications deintensification. | Patients with very low HbA1c (<6.0%) compared to those with higher HbA1c (≥6.5%) were: A. Older (mean age 78.6 vs. 77.8 years). B. More comorbidities (mean CCIS 1.44 vs. 1.27). C. More low life expectancy <5.0 years (19.9% vs. 15.7%). D. More dementia (2.3 vs. 1.6%). E. More palliation in prior year (0.5 vs. 0.3%). F. More metastatic cancer (0.7 vs. 0.4%). |
| Yotsapont, et al., retrospective, Thailand, 2015 [28] | 143.0 patients > 85.0 years old with DM. | Describe clinical characteristics and outcomes of “oldest old” patients with DM. | Patients had: 1. Long duration of diabetes, mean (SD) 22.1 (11.1) Y. 2. Severe comorbidities, CCIS ≥ 5.0 in 35.3%. 3. Tight glycaemic control, HbA1c < 7.0% in 66.9%. 4. Frequent hypoglycaemia in 10.5%. 5. Multiple comorbidities: 23.4% diabetic retinopathy, 54.9% CKD, 15.8% CV disease, 18.0% stroke, 22.6% dementia. 6. Only 20.0% of those with HbA1c <6.0% received medication deintensification. |
SD = Standard deviation, Y = Year, CCIS = Charlson Comorbidity Index Score, aDCSI = Adjusted diabetic complication severity index, DM = Diabetes mellitus, CKD = Chronic kidney disease, CLD = Chronic liver disease, COPD = Chronic obstructive pulmonary disease, AF = Atrial fibrillation, CAD = Coronary artery disease, PVD = Peripheral vascular disease, OA = Osteoarthritis, CVD = Cerebrovascular disease, OR = Odds ratio, CI = Confidence interval, CFS = Clinical frailty scale, Hb = Haemoglobin, SBP = Systolic blood pressure, BMI = Body mass index, eGFR = Estimated glomerular filtration rate, CV = Cardiovascular, SMBG = Self-monitoring blood glucose, SU = Sulfonyl urea, CHF = Congestive heart failure, MMSE = Mini-mental state examination, GDS = Geriatric depression scale, F/U = Follow up, ASMI= Appendicular skeletal muscle mass index, OR = Odds ratio.