Papakitsou I, Papazachariou A, Filippatos TD. Association of glycated hemoglobin levels with 3-year mortality in hospitalized older adults with diabetes: The role of frailty. World J Clin Cases 2026; 14(14): 120574 [DOI: 10.12998/wjcc.v14.i14.120574]
Corresponding Author of This Article
Theodosios D Filippatos, MD, PhD, Associate Professor, School of Medicine, University of Crete, Panepistimiou Ave, Crete, Heraklion 71003, Greece. filtheo@uoc.gr
Research Domain of This Article
Geriatrics & Gerontology
Article-Type of This Article
Observational Study
Open-Access Policy of This Article
This article is an open-access article which was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/
May 16, 2026 (publication date) through Apr 27, 2026
Times Cited of This Article
Times Cited (0)
Journal Information of This Article
Publication Name
World Journal of Clinical Cases
ISSN
2307-8960
Publisher of This Article
Baishideng Publishing Group Inc, 7041 Koll Center Parkway, Suite 160, Pleasanton, CA 94566, USA
Share the Article
Papakitsou I, Papazachariou A, Filippatos TD. Association of glycated hemoglobin levels with 3-year mortality in hospitalized older adults with diabetes: The role of frailty. World J Clin Cases 2026; 14(14): 120574 [DOI: 10.12998/wjcc.v14.i14.120574]
Ioanna Papakitsou, Andria Papazachariou, Theodosios D Filippatos, School of Medicine, University of Crete, Crete, Heraklion 71003, Greece
Ioanna Papakitsou, Andria Papazachariou, Theodosios D Filippatos, Department of Internal Medicine, University Hospital of Heraklion, Heraklion 71500, Greece
Author contributions: Papakitsou I contributed to study design, data collection, statistical analysis, and manuscript drafting; Papazachariou A contributed to data collection, data interpretation, statistical analysis and manuscript revision; Filippatos TD conceived and supervised the study, contributed to study design, data interpretation, and critically revised the manuscript for important intellectual content; and all authors approved the final version of the manuscript.
Institutional review board statement: The original study protocol, including the present sub-analysis, was approved by the Ethics Committee of the University Hospital of Heraklion (approval code 716/16-01-2019).
Informed consent statement: All study participants, or either legal guardians, provided written consent prior to study enrollment.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
STROBE statement: The authors have read the STROBE Statement-checklist of items, and the manuscript was prepared and revised according to the STROBE Statement-checklist of items.
Data sharing statement: There is no additional data available.
Corresponding author: Theodosios D Filippatos, MD, PhD, Associate Professor, School of Medicine, University of Crete, Panepistimiou Ave, Crete, Heraklion 71003, Greece. filtheo@uoc.gr
Received: March 3, 2026 Revised: March 26, 2026 Accepted: April 15, 2026 Published online: May 16, 2026 Processing time: 56 Days and 4.7 Hours
Abstract
BACKGROUND
Tight glycemic control in older adults with type 2 diabetes mellitus (T2DM) has been associated with increased risk of hypoglycemia, functional decline, and mortality, particularly in frail individuals. However, real-world glycemic patterns and their prognostic implications across different frailty and multimorbidity states remain unclear.
AIM
To evaluate glycemic control at admission in very old adults with T2DM according to frailty and comorbidity burden, and to investigate the association between glycated hemoglobin (HbA1c) levels and 3-year post discharge-mortality.
METHODS
This study is a sub-analysis of a prospective cohort study including hospitalized patients aged ≥ 65 years admitted to a medical ward. Baseline demographic, clinical, and laboratory data were collected during hospitalization. Frailty was assessed using the Clinical Frailty Scale (CFS), functional status using the Katz Index, and comorbidity burden using the Charlson Comorbidity Index. Patients were followed for three years after discharge. HbA1c was analyzed as both a categorical (< 6.5%, 6.5%-8.0%, > 8.0%) and continuous variable. Multivariable logistic regression models were used to examine the association between HbA1c and 3-year post discharge all-cause mortality.
RESULTS
A total of 430 hospitalized older adults with type 2 diabetes (mean age 81.5 ± 7.3 years; 55.8% women) were included. Severe frailty (CFS ≥ 6) was present in 46.7% of participants. The median HbA1c was 6.1% (interquartile range: 5.5-7.1), with no significant differences across frailty, functional status, or comorbidity categories. Among them, 62 patients died during hospitalization; the 368 remaining patients exhibited a 54.9% 3-year post-discharge mortality. Lower HbA1c levels were independently associated with increased 3-year mortality (odds ratio per 1% HbA1c decrease 1.24, 95%CI: 1.05-1.46; P = 0.010), with significant effect modification by frailty (interaction OR: 0.97 per CFS unit; 95%CI: 0.93-0.99; P = 0.034), indicating an approximate 3% relative reduction in the strength of the HbA1c-mortality association per CFS increment.
CONCLUSION
In very old medical inpatients with T2DM, glycemic control was similar irrespective of frailty, functional status, or comorbidity burden. Lower HbA1c levels were independently associated with increased 3-year-post-discharge-mortality, with significant effect modification by frailty, supporting a frailty-informed approach to glycemic targets.
Core Tip: In very old hospitalized adults with type 2 diabetes, glycemic control did not differ according to frailty or comorbidity burden, suggesting a lack of individualized treatment in routine practice. Lower glycated hemoglobin levels were independently associated with increased 3-year mortality, particularly in less frail individuals. These findings support a frailty-informed approach to glycemic targets in very old patients.
Citation: Papakitsou I, Papazachariou A, Filippatos TD. Association of glycated hemoglobin levels with 3-year mortality in hospitalized older adults with diabetes: The role of frailty. World J Clin Cases 2026; 14(14): 120574
Type 2 diabetes mellitus (T2DM) is one of the most common health challenges of the twenty first century. More than 10% of adults aged over 65 years old are diagnosed with T2DM, with the incidence constantly rising especially in elderly[1]. According to recent data from the International Diabetes Federation, the incidence of T2DM is expected to increase further in the coming decades[2], mainly due to the aging population and sedentary lifestyle[3].
Elderly patients with T2DM constitute a heterogeneous population, with significant differences in functional status, comorbidities, and vulnerability. Chronological age alone does not adequately reflect overall and actual health status, as many older people exhibit frailty and functional dependence, factors that influence prognosis and long-term clinical outcomes[4,5]. It is widely acknowledged that frailty and disability are associated with an increased risk of hypoglycemia, hospitalization, and mortality, especially when strict antidiabetic treatment regimens are applied to elderly with limited physiological reserve[6]. At the same time, previous studies have reported a J-shaped or U-shaped relationship between glycated hemoglobin (HbA1c) levels and mortality, with increased mortality risk observed at both very low and very high HbA1c levels in older patients, supporting the need for individualized glycemic targets[7-11].
According to recent guidelines, antidiabetic regimens for older people should be adjusted based on frailty, performance status and additional comorbidities[12-16]. However, data regarding the application of these guidelines in everyday clinical practice for very old frail individuals are still lacking. Furthermore, the prognostic significance of HbA1c in this particularly vulnerable group and the potential modifying role of frailty in its relationship with long-term mortality have not been adequately investigated. The aim of this study was to evaluate glycemic control at admission in very old medical patients with T2DM and to investigate the association of HbA1c levels with three-year post-discharge mortality, taking into account frailty and comorbidities.
MATERIALS AND METHODS
Study design
This study is a sub-analysis of a prospective cohort study conducted in the Department of Internal Medicine at the University Hospital of Heraklion, Crete, Greece, that evaluated clinical characteristics, frailty, and outcomes in 949 hospitalized older adults admitted to an Internal Medicine Ward investigating predictors of 3-year post-discharge mortality[17]. The original cohort enrolled only patients who admitted for acute medical conditions, aged ≥ 65 years in accordance with the World Health Organization definition[18]. Overall inclusion criteria were age ≥ 65 years, hospitalization in the internal medicine ward during the study period, and a documented diagnosis of T2DM with available baseline HbA1c measurements. Exclusion criteria included scheduled (elective) admissions, inability to provide informed consent, and missing baseline HbA1c measurements; therefore, analyses were performed using complete-case data. The present sub-analysis represents the full eligible T2DM subgroup of the parent cohort, minimizing additional selection bias related to subgroup selection.
Baseline demographic data (age, sex, weight and height), comorbidities [hypertension, dyslipidemia, heart failure (HF), coronary heart disease, cerebrovascular disease, peripheral artery disease, chronic kidney disease (CKD), and dementia], chronic medication use (antidiabetic, antihypertensive, and lipid-lowering agents), as well as functional and laboratory data (complete blood count, serum creatinine, HbA1c, and serum albumin levels) were collected at admission[17]. All laboratory measurements were performed at the Central Laboratory of the University Hospital of Heraklion. HbA1c levels were assessed by high-performance liquid chromatography using the Bio-Rad D-10 system. Routine biochemical and hematological parameters, including complete blood count, serum creatinine, and albumin, were measured using standardized automated enzymatic and colorimetric assays[17]. The reference ranges in our laboratory were as follows: HbA1c < 5.7% (normal), serum creatinine 0.6-1.3 mg/dL, and serum albumin 3.5-5.0 g/dL.
Functional status was evaluated using the Katz Index of Independence in activities of daily living, categorized as follows: Severe limitation: Score 0, independent (0-3 points), score 1-moderate limitation (4-5 points), and score 2-no limitation (6 points). Frailty status was assessed using the Clinical Frailty Scale (CFS). CFS categories were defined as 0 (mild or no frailty; scores 1-3), 1 (moderate frailty; scores 4-5), and 2 (severe frailty; scores ≥ 6). Comorbidity burden was assessed using the Charlson Comorbidity Index (CCI), categorized into three groups: Low/0 (scores 1-3), moderate/1 (scores 4-5), and high/2 (scores ≥ 6).
Patients who survived the index hospitalization were followed for three years after hospital discharge to assess all-cause 3-year mortality. Mortality status was obtained from hospital records and/or national registry data.
The original study protocol, including the present sub-analysis, was approved by the Ethics Committee of the University Hospital of Heraklion (approval code 716/16-01-2019). Written informed consent was obtained from all participants or from a legal representative when patients were unable to provide consent. Consecutive patient enrollment was used to minimize selection bias. Standardized laboratory methods and validated assessment tools were applied to reduce measurement bias. Multivariable models were constructed to adjust for potential confounding.
Statistical analysis
Continuous variables were tested for normality using the Kolmogorov-Smirnov test and visual inspection of histograms. Normally distributed data are presented as mean ± SD, while non-normally distributed variables are presented as median with interquartile range (IQR). Categorical variables are presented as absolute numbers and percentages. All statistical analyses were two-sided, and statistical significance was defined as a P value ≤ 0.05.
Baseline characteristics and outcomes were compared across HbA1c categories (< 6.5%, 6.5%-8.0%, and > 8.0%). These cut-offs were selected a priori based on current clinical guidelines and previous literature in older adults with T2DM, aiming to reflect distinct glycemic control patterns. Comparisons of continuous variables were performed using ANOVA analysis for normally distributed data and the Kruskal-Wallis test for non-normally distributed data. Categorical variables were compared using the χ2 test or Fisher’s exact test, as appropriate.
Mortality was assessed as a binary outcome at 3 years post-discharge. As detailed time-to-event data were not consistently available, time-to-event analyses were not performed. Therefore, the association between HbA1c levels and 3-year post-discharge all-cause mortality was evaluated using multivariable logistic regression analyses. Age, sex, frailty status, and major comorbidities were included as potential confounders in the multivariable models. HbA1c was examined both as a categorical variable [< 6.5%, 6.5%-8.0% (reference), > 8.0%] to explore potential non-linear associations, and as a continuous variable to assess the linear relationship between glycemic control and mortality.
To formally assess potential U-shaped associations, HbA1c was additionally modeled as a centred continuous variable with a quadratic term (HbA1c and HbA1c2). To evaluate whether the association between HbA1c and mortality differed according to frailty status, a multiplicative interaction term was constructed as the product of continuous HbA1c and ordinal CFS and entered into the multivariable logistic regression model. Results are presented as odds ratios (ORs) with 95%CI.
Statistical analyses were performed using SPSS software, version 25.0 (IBM Corp., Armonk, NY, United States). Mortality status was available for all discharged participants during the 3-year follow-up period, and no loss to follow-up occurred.
RESULTS
During the study period, a total of 1134 individuals were admitted to the medical ward. All participants were recruited from a tertiary hospital in Crete, Greece, and the study population consisted predominantly of individuals of Caucasian European origin. Of these, 1100 patients were aged ≥ 65 years, 151 were excluded due to scheduled admissions or inability to provide informed consent. The final eligible cohort consisted of 949 hospitalized older adults admitted for acute medical conditions. Among them, 434 patients had a diagnosis of T2DM. Four patients were excluded due to missing baseline HbA1c measurements, resulting in a cohort of 430 hospitalized very old adults with T2DM (mean age 81.5 ± 7.3 years) included in the baseline analysis. Among these, 62 patients died during the index hospitalization. Consequently, 368 patients with T2DM who were discharged alive constituted the study sample for the assessment of 3-year post-discharge mortality, as illustrated in Figure 1.
Figure 1 Study flowchart.
T2DM: Type 2 diabetes mellitus; HbA1c: Glycated hemoglobin.
From the 430 medical inpatients, two-hundred forty (55.8%) were females. The median duration of T2DM was 13 years (IQR: 11-16), and the median HbA1c was 6.1% (IQR: 5.5-7.1). Among the sample, 201 (46.7%) patients were severely frail (CFS ≥ 6), 196 (45.5%) had severe limitation (Katz index = 0-3) and 291 participants (67.7%) had ≥ 6 chronic conditions (CCI ≥ 6). Based on HbA1c levels, participants were categorized into three groups: HbA1c < 6.5% (n = 246), HbA1c 6.5%-8.0% (n = 137), and HbA1c > 8.0% (n = 47). The main findings of the study were that lower HbA1c levels were associated with increased 3-year mortality, while frailty emerged as the strongest predictor of adverse outcomes.
Baseline demographic characteristics, comorbidities, functional status, and clinical variables according to HbA1c categories are presented in Table 1. No statistically significant differences were observed among the three groups in respect to age, sex, body mass index, diabetes duration and comorbidity burden as calculated with CCI categories (all P > 0.05). The distribution of CFS categories differed significantly across HbA1c groups (P = 0.003), with a higher proportion of severely frail patients (CFS ≥ 6) observed in the highest HbA1c category. Similarly, bedridden status was more frequent in patients with HbA1c > 8.0% (P = 0.020).
Table 1 Baseline demographics, comorbidities, and laboratory findings stratified by glycated hemoglobin levels in older patients with diabetes mellitus, n (%).
During the 3-year follow-up period after hospital discharge, 202 (54.9%) patients died. Crude mortality rates differed across HbA1c categories, being highest among patients with HbA1c < 6.5% (63.4%, n = 128), 55.6% (n = 20) in patients with HbA1c > 8.0% and 26.7% (n = 54) with those with HbA1c 6.5-8.0% (P = 0.026). Notably, the highest mortality was observed in patients with HbA1c < 6.5%, despite similar glycemic levels across frailty and comorbidity categories.
No statistically significant differences in median HbA1c values were observed across categories of comorbidity burden or frailty (Figure 2). Specifically, median HbA1c was 6.0% (IQR: 5.9-7.1) in patients with CCI 1-3 (n = 7), 6.1% (IQR: 5.3-7.0) in those with CCI 4-5 (n = 132), and 6.1% (IQR: 5.5-7.3) among patients with CCI ≥ 6 (n = 291) (P = 0.73). Similarly, HbA1c levels did not differ substantially across CFS categories; 6.4% (IQR: 5.4-7.1) in patients with CFS 1-3 (n = 145), 6.0% (IQR: 5.3-7.0) in those with CFS 4-5 (n = 84), and 6.1% (IQR: 5.5-7.5) in patients with CFS ≥ 6 (n = 201) (P = 0.08).
Figure 2 Distribution of glycated hemoglobin levels according to Clinical Frailty Scale categories and Charlson Comorbidity Index categories.
A: Clinical Frailty Scale; B: Charlson Comorbidity Index. Boxes represent the interquartile range, horizontal lines indicate median values, whiskers represent 1.5 × interquartile range, and dots indicate outliers. HbA1c: Glycated hemoglobin; CFS: Clinical Frailty Scale; CCI: Charlson Comorbidity Index.
In addition, treatment patterns were examined to explore whether glycemic control differences were reflected in antidiabetic therapy. As shown in Table 2, the distribution of antidiabetic medications was largely similar across HbA1c categories. Metformin use did not differ significantly between groups (58.5% in HbA1c < 6.5%, 62.8% in HbA1c 6.5%-8.0%, and 59.6% in HbA1c > 8.0%; P = 0.718), nor did the use of sulfonylureas (18.3%, 18.2%, and 23.4%, respectively; P = 0.698). Similarly, no significant differences were observed in the use of SGLT2 inhibitors (5.3%, 8.0%, and 2.1%; P = 0.284), GLP-1 receptor agonists (2.0%, 3.6%, and 0.0%; P = 0.322), basal insulin (32.9%, 27.7%, and 21.3%; P = 0.219), or rapid-acting insulin (13.4%, 10.9%, and 12.8%; P = 0.783). In contrast, DPP-4 inhibitors were more frequently prescribed in patients with higher HbA1c levels (61.7% in HbA1c > 8.0% vs 42.7% and 39.4% in the lower categories, P = 0.026). The distribution of antidiabetic treatment across HbA1c categories is summarized in Table 2.
Table 2 Distribution of antidiabetic treatment according to glycated hemoglobin categories in the total study sample, n (%).
To evaluate the association between glycemic control and 3-year post-discharge all-cause mortality, and to investigate whether this association differed according to frailty status, multivariable logistic regression analyses were performed (Table 3). Sex was not significantly associated with mortality in the adjusted model (OR: 0.88, 95%CI: 0.60-1.62, P = 0.959). Three-year post-discharge mortality rates did not differ significantly between men and women (P = 0.249). Lower HbA1c levels were independently associated with increased 3-year mortality. Specifically, each 1% decrease in HbA1c was associated with 24% higher odds of mortality (OR: 1.24, 95%CI: 1.05-1.46, P = 0.010), after adjustment for age, sex, frailty status, and major comorbidities. When HbA1c was modeled as a centred continuous variable with an additional quadratic term, the squared HbA1c term was not statistically significant, indicating no evidence of a U-shaped association between HbA1c and 3-year mortality. Frailty was a strong independent predictor of mortality, with a graded increase in risk across higher CFS categories. Compared with patients with CFS scores 1-3, those with moderate frailty (CFS: 4-5) had more than twofold higher odds of mortality (OR: 2.10, 95%CI: 1.12-3.91, P = 0.021), while severely frail patients (CFS ≥ 6) had markedly increased odds of death (OR: 7.57, 95%CI: 3.65-15.70, P < 0.001).
Table 3 Multivariate logistic regression analysis for 3-year post discharge mortality.
In a secondary analysis (Table 4), when HbA1c was analyzed as a categorical variable, patients with HbA1c levels < 6.5% had significantly higher odds of 3-year mortality compared with those with HbA1c levels between 6.5% and 8.0% (OR: 1.73, 95%CI: 1.03-2.91, P = 0.039), after multivariable adjustment. In contrast, HbA1c levels > 8.0% were not significantly associated with mortality (OR: 0.81, 95%CI: 0.34-1.94, P = 0.639). Frailty remained a strong independent predictor of mortality, with markedly increased odds observed in patients with higher CFS categories. Compared with patients with CFS 1-3, those with CFS 4-5 had nearly twofold higher odds of 3-year mortality (OR: 1.98, 95%CI: 1.06-3.73, P = 0.034), while patients with CFS ≥ 6 had more than sevenfold higher odds (OR: 7.02, 95%CI: 3.38-14.57, P < 0.001). In addition, the presence of malignancy was associated with a substantially increased risk of death (OR: 8.08, 95%CI: 2.24-29.18, P = 0.001).
Table 4 Multivariate logistic regression model including the association between glycated hemoglobin categories and 3-year post discharge mortality.
To further examine whether the association between HbA1c and 3-year mortality varied according to frailty status, an interaction term between continuous HbA1c and CFS was introduced into the multivariable logistic regression model (Supplementary Table 1). The interaction term reached statistical significance (OR: 0.97 per unit increase in CFS; 95%CI: 0.93-0.99; P = 0.034), indicating that the strength of the association between lower HbA1c and 3-year mortality differed across frailty levels. Although lower HbA1c levels were independently associated with increased mortality overall (OR: 1.17 per 1% decrease; 95%CI: 1.02-1.36; P = 0.029), the strength of this association progressively appeared to diminish with increasing frailty. Specifically, the interaction term corresponds to an approximate 3% reduction in the effect size of HbA1c for each increment in CFS category.
Taken together, these findings suggest that lower HbA1c levels are associated with increased mortality independently of frailty, although this association appears to be attenuated with increasing frailty.
DISCUSSION
In this cohort study of medical hospitalized very old adults with T2DM, low HbA1c levels at admission were significantly associated with higher 3-year post-discharge mortality after adjustment for age, sex, frailty status and multiple comorbidities. In this sample, HbA1c values were broadly similar across frailty and comorbidity categories, suggesting that glycemic control was not systematically aligned with patients’ overall vulnerability. Frailty emerged as the strongest determinant of mortality, with a stepwise and marked increase in the risk of death among patients with moderate and severe frailty, independent of glycemic control. Importantly, a statistically significant interaction between HbA1c and frailty was observed, indicating that the association between lower HbA1c and mortality varied according to frailty level, being more pronounced among less frail individuals and progressively attenuated as frailty increased.
Optimal glycemic control in very old patients with T2DM remains a challenge, due to population heterogeneity, frequent coexistence of multiple comorbidities, and variable functional status. Τhe results of randomized control trials like Action to Control Cardiovascular Risk in Diabetes (ACCORD) and Veteran Affairs Diabetes Trial (VADT) showed no additional benefit in overall mortality from intensive glycemic control and, in particularly the case of ACCORD, showed an increased risk of death[8,19]. These trials focused mainly on young or early old community dwelling patients and revealed the need of personalized glycemic goals. Recent guidelines focus on older patients suggesting that the choice of glycemic targets should be based on overall health status, functional capacity, and life expectancy, and not solely on achieving low HbA1c levels[12,14,20-22]. In this context, the present study supports a cautious interpretation of low HbA1c values and offers real-world data from a particularly vulnerable population of very old patients.
Given the existing gap in the literature regarding very old and frail patients with T2DM, this study is among the few to investigate long-term post-discharge mortality in hospitalized very old patients after simultaneous assessment of HbA1c levels and frailty, with an extended follow-up period. Our results showed that the 3-year post-discharge mortality rate was 54.9%, indicating that more than half of the discharged patients died within the subsequent three years. This finding was consistent with our cohort population’s characteristics (very old individuals with high prevalence of comorbidity and frailty) and with previous reports emphasizing the prognostic impact of hospitalization in older adults with diabetes[5,6,23]. In this context, the identification of prognostic factors related to long-term outcomes is of particular clinical relevance. Within our cohort, lower HbA1c levels were significantly associated with an increased risk of 3-year post-discharge mortality, even after adjustment for age, sex, and frailty status, indicating that low HbA1c constitutes an independent prognostic marker of adverse outcomes in this vulnerable population. Previous observational studies in older adults have described J- or U-shaped associations between HbA1c and mortality, with increased risk observed at both low and high HbA1c levels[24-28]. However, in our analysis, we did not find statistical evidence of a U-shaped relationship, as the quadratic term was not significant. Instead, the excess risk was mainly driven by lower HbA1c levels. This may be partly explained by the relatively small number of patients in the highest HbA1c category, as well as by the very advanced age and high burden of frailty and comorbidity in our population, where overall vulnerability may outweigh the possible independent effect of hyperglycaemia[29,30]. Consistent with our findings, Hamada and Gulliford[31] demonstrated that low HbA1c was associated with increased mortality in individuals aged 80 years and older with T2DM supporting the notion that very low glycemic levels may be particularly detrimental in advanced age.
The association of lower HbA1c levels and increased mortality in very old adults is likely multifactorial and should be interpreted with caution. Firstly, hypoglycemia represents a plausible mechanism, particularly in the context of intensive glucose-lowering treatment. Data from the VADT and the Action in Diabetes and Vascular Disease study demonstrated that hypoglycemic episodes were linked to increased cardiovascular and non-cardiovascular mortality[7,32]. The ACCORD trial suggested that hypoglycemia acted more as a marker of vulnerability and provided a basis for adverse outcomes and complications in the long term[33,34]. In elderly patients, there is a greater risk of complications from hypoglycemia such as falls, cardiac arrhythmias, increased risk of hospitalization and functional decline[35]. Yet, low HbA1c levels in older persons may indicate underlying health decline in addition to glycemic management. Chronic inflammatory states such as frailty-associated inflammation, chronic diseases (HF, CKD, atherosclerosis), malignancy, and chronic or recurrent infections (including urinary or respiratory tract infections and decubitus ulcers), may contribute to lower HbA1c levels independently of diabetes management[36]. Furthermore, conditions such as protein-energy malnutrition, unintentional weight loss, sarcopenia, declining metabolic demand, and age-related catabolic processes may also lead to progressively lower HbA1c levels over time[36]. These observations suggest that low HbA1c may represent a marker of biological vulnerability rather than a direct causal factor.
The prevalence of frailty in this study was very high; nearly half of the studied population had severe frailty with CFS ≥ 6 showing that very old hospitalized patients with T2DM usually are more vulnerable with low physical reserve. The high prevalence of frailty in people with T2DM has been extensively documented in the literature and is associated with increased mortality, functional decline, and greater use of health services[37-41]. Several studies have explored the link between frailty and mortality in different patient populations. Frailty, assessed using the CFS, was a significant predictor of mortality across multiple follow-up periods, extending up to 7 years[40]. Our findings are in line with previous studies that have shown that hospitalized older people with T2DM constitute a particularly vulnerable population, in which biological age and functional status play a crucial role in prognosis[17,42]. In addition, according to the multivariate analysis, severely frail individuals had more than sevenfold higher odds of 3-year mortality compared with those with low frailty burden, underscoring the dominant prognostic role of biological age and functional reserve over traditional metabolic parameters.
An important finding of the present study was that HbA1c levels were broadly similar across frailty and comorbidity categories. This observation suggests that glycemic control was not systematically tailored according to patients’ overall vulnerability. In clinical practice, current guidelines recommend individualized glycemic targets in older adults, taking into account frailty, functional status, and life expectancy[43]. However, our findings indicate that such an individualized approach may not be consistently implemented in very old hospitalized patients. This may partly reflect therapeutic inertia, defined as the failure to appropriately adjust treatment intensity despite changes in clinical status, as well as the complexity of managing diabetes in frail individuals with multiple comorbidities[23,44]. Consequently, these results highlight a gap between guideline recommendations and real-world practice and support the need for more personalized, frailty-informed treatment strategies in this vulnerable population. Nevertheless, a significant interaction between HbA1c and frailty was observed, suggesting that the prognostic impact of HbA1c varies across levels of vulnerability. Specifically, the association between lower HbA1c and mortality appeared more pronounced among individuals with lower frailty burden and progressively attenuated as frailty increased. One possible explanation is that, in less frail patients, lower HbA1c may more directly reflect intensive glycemic management and potential overtreatment, thereby contributing to adverse outcomes. In contrast, among highly frail individuals, overall vulnerability and multimorbidity may dominate the risk profile, diminishing the relative contribution of glycemic control to mortality risk. This finding contrasts with previous reports, such as the study by Yanagita et al[45], who observed an association between lower HbA1c levels and greater frailty severity in elderly patients with diabetes while Muszalik et al[46] reported that poor metabolic control was linked to frailty syndrome. Similarly, Paterni et al[47], in older patients with hip fracture, showed that frailty modified the prognostic impact of HbA1c, particularly at higher levels. The differences between these studies and our findings are likely related to the different clinical setting of the present study, and our patients’ characteristics. They were hospitalized, very elderly, and with a high burden of comorbidities and functional impairment and it are possible that the impact of acute illness and overall biological burden may override the relationship between HbA1c and frailty, making HbA1c less representative of overall frailty. Overall, our findings suggest that, in very old hospitalized patients, frailty and HbA1c may be distinct but coexisting dimensions of the clinical picture, with low HbA1c maintaining its independent prognostic significance for mortality.
A particularly important result of our study was the absence of substantial differences in antidiabetic treatment between HbA1c categories, despite clear differences in mortality, frailty and functional status of patients. Except from the more frequent use of DPP-4 inhibitors in the group with higher HbA1c, the therapeutic strategy was uniform across HbA1c groups irrespectively of glycemic control, physical reserve or comorbidity. This finding shows that in the everyday clinical practice the management of T2DM in very old patients remains consistent, without systematic adjustment to the patient’s general health status and vulnerability[48]. This observation contrasts with current guidelines and expert consensus statements, which emphasize that the choice and intensity of antidiabetic treatment in older persons should be individualized based on the risk of hypoglycemia, functional capacity, frailty, and life expectancy[23,35]. Therefore, our findings highlight a potential gap between theoretical recommendations and real clinical practice in very old patients with T2DΜ. The lack of differences in antidiabetic therapy across HbA1c groups in our cohort may partly explain why low HbA1c was not associated with lower frailty burden, supporting the need for more individualized and patient-centered treatment strategies in this vulnerable population. The limited use of newer glucose-lowering agents in this cohort likely reflects real-world prescribing patterns in very old patients with multiple comorbidities.
The findings of the present study should be interpreted along with the guidelines for glycemic control in older patients. The 2025 ADA Standards of Care suggest a personalized approach, recommending less strict glycemic goals for very old and vulnerable patients[49,50]. In this context, our results, which show increased mortality in patients with low HbA1c levels regardless of the degree of vulnerability, supports a cautious interpretation of low HbA1c values in very old hospitalized individuals and the current direction towards more realistic and individualized therapeutic goals in very old patients with T2DΜ.
Despite its prospective design and inclusion of very old persons, this study has certain limitations. Its observational nature does not allow for causal conclusions to be drawn, while episodes of hypoglycemia were not systematically recorded. Moreover, HbA1c was measured at admission and there was no information on serial changes in glycemic control or antidiabetic treatment. In addition, limited information on nutritional status, weight loss, or potential overtreatment may have influenced HbA1c levels and outcomes in this population. Furthermore, the absence of detailed time-to-event data precluded survival analysis, limiting the ability to assess the temporal dynamics of mortality. Finally, the single-center design may limit generalizability; however, the hospital serves both urban and rural populations. Residual confounding cannot be excluded and may have influenced the observed associations. In addition, the absence of detailed data on hypoglycemic episodes limits the interpretation of the relationship between low HbA1c and mortality.
CONCLUSION
In conclusion, in our study, HbA1c levels were similar regardless of the degree of frailty, functional status, or comorbidity burden of the patients. Furthermore, low HbA1c levels emerged as an independent risk factor for increased three-year mortality after discharge. The prognostic impact of HbA1c appeared to vary according to frailty status, with a stronger association observed among less frail individuals. These findings suggest that in very old patients with type 2 diabetes, strict glycemic control does not necessarily imply a better prognosis and highlight the need for individualized glycemic goals and personalized treatment based on the overall clinical condition.
ADVANCE Collaborative Group; Patel A, MacMahon S, Chalmers J, Neal B, Billot L, Woodward M, Marre M, Cooper M, Glasziou P, Grobbee D, Hamet P, Harrap S, Heller S, Liu L, Mancia G, Mogensen CE, Pan C, Poulter N, Rodgers A, Williams B, Bompoint S, de Galan BE, Joshi R, Travert F. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes.N Engl J Med. 2008;358:2560-2572.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 5517][Cited by in RCA: 4773][Article Influence: 265.2][Reference Citation Analysis (0)]
Duckworth W, Abraira C, Moritz T, Reda D, Emanuele N, Reaven PD, Zieve FJ, Marks J, Davis SN, Hayward R, Warren SR, Goldman S, McCarren M, Vitek ME, Henderson WG, Huang GD; VADT Investigators. Glucose control and vascular complications in veterans with type 2 diabetes.N Engl J Med. 2009;360:129-139.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 3873][Cited by in RCA: 3231][Article Influence: 190.1][Reference Citation Analysis (0)]
Qaseem A, Obley AJ, Shamliyan T, Hicks LA, Harrod CS, Crandall CJ; Clinical Guidelines Committee of the American College of Physicians, Balk EM, Cooney TG, Cross JT Jr, Fitterman N, Lin JS, Maroto M, Miller MC, Shekelle P, Tice JA, Tufte JE, Etxeandia-Ikobaltzeta I, Yost J. Newer Pharmacologic Treatments in Adults With Type 2 Diabetes: A Clinical Guideline From the American College of Physicians.Ann Intern Med. 2024;177:658-666.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 70][Cited by in RCA: 60][Article Influence: 30.0][Reference Citation Analysis (0)]
Davies MJ, Aroda VR, Collins BS, Gabbay RA, Green J, Maruthur NM, Rosas SE, Del Prato S, Mathieu C, Mingrone G, Rossing P, Tankova T, Tsapas A, Buse JB. Management of Hyperglycemia in Type 2 Diabetes, 2022. A Consensus Report by the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD).Diabetes Care. 2022;45:2753-2786.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 1463][Cited by in RCA: 1142][Article Influence: 285.5][Reference Citation Analysis (0)]
Cefalu WT, Andersen DK, Arreaza-Rubín G, Pin CL, Sato S, Verchere CB, Woo M, Rosenblum ND; Symposium planning committee, moderators, and speakers. Heterogeneity of Diabetes: β-Cells, Phenotypes, and Precision Medicine: Proceedings of an International Symposium of the Canadian Institutes of Health Research's Institute of Nutrition, Metabolism and Diabetes and the U.S. National Institutes of Health's National Institute of Diabetes and Digestive and Kidney Diseases.Diabetes Care. 2022;45:3-22.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 26][Cited by in RCA: 21][Article Influence: 5.3][Reference Citation Analysis (0)]
Cybulski M, Krajewska-Kułak E, Sowa P, Orzechowska M, Van Damme-Ostapowicz K, Rozwadowska E, Guzowski A. Elderly Peoples' Perception of Young People - A Preliminary Study.Iran J Public Health. 2013;42:1099-1105.
[PubMed] [DOI]
Abraira C, Colwell J, Nuttall F, Sawin CT, Henderson W, Comstock JP, Emanuele NV, Levin SR, Pacold I, Lee HS. Cardiovascular events and correlates in the Veterans Affairs Diabetes Feasibility Trial. Veterans Affairs Cooperative Study on Glycemic Control and Complications in Type II Diabetes.Arch Intern Med. 1997;157:181-188.
[PubMed] [DOI]
Sugimoto T, Sakurai T, Uchida K, Kuroda Y, Tokuda H, Omura T, Noguchi T, Komatsu A, Nakagawa T, Fujita K, Matsumoto N, Ono R, Crane PK, Saito T. Impact of Type 2 Diabetes and Glycated Hemoglobin Levels Within the Recommended Target Range on Mortality in Older Adults With Cognitive Impairment Receiving Care at a Memory Clinic: NCGG-STORIES.Diabetes Care. 2024;47:864-872.
[RCA] [PubMed] [DOI] [Full Text][Cited by in RCA: 5][Reference Citation Analysis (0)]
Ray KK, Seshasai SR, Wijesuriya S, Sivakumaran R, Nethercott S, Preiss D, Erqou S, Sattar N. Effect of intensive control of glucose on cardiovascular outcomes and death in patients with diabetes mellitus: a meta-analysis of randomised controlled trials.Lancet. 2009;373:1765-1772.
[RCA] [PubMed] [DOI] [Full Text][Cited by in Crossref: 1113][Cited by in RCA: 927][Article Influence: 54.5][Reference Citation Analysis (0)]
Bonds DE, Miller ME, Bergenstal RM, Buse JB, Byington RP, Cutler JA, Dudl RJ, Ismail-Beigi F, Kimel AR, Hoogwerf B, Horowitz KR, Savage PJ, Seaquist ER, Simmons DL, Sivitz WI, Speril-Hillen JM, Sweeney ME. The association between symptomatic, severe hypoglycaemia and mortality in type 2 diabetes: retrospective epidemiological analysis of the ACCORD study.BMJ. 2010;340:b4909.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 779][Cited by in RCA: 719][Article Influence: 44.9][Reference Citation Analysis (0)]
Riddle MC, Ambrosius WT, Brillon DJ, Buse JB, Byington RP, Cohen RM, Goff DC Jr, Malozowski S, Margolis KL, Probstfield JL, Schnall A, Seaquist ER; Action to Control Cardiovascular Risk in Diabetes Investigators. Epidemiologic relationships between A1C and all-cause mortality during a median 3.4-year follow-up of glycemic treatment in the ACCORD trial.Diabetes Care. 2010;33:983-990.
[RCA] [PubMed] [DOI] [Full Text] [Full Text (PDF)][Cited by in Crossref: 359][Cited by in RCA: 328][Article Influence: 20.5][Reference Citation Analysis (0)]
Scientific quality: Grade A, Grade A, Grade B, Grade C
Novelty: Grade A, Grade B, Grade B, Grade C
Creativity or innovation: Grade A, Grade B, Grade B, Grade C
Scientific significance: Grade A, Grade B, Grade B, Grade C
P-Reviewer: Li M, Associate Chief Physician, China; Malik S, PhD, Professor, Researcher, Pakistan; Racz A, MD, PhD, Professor, Croatia S-Editor: Liu JH L-Editor: A P-Editor: Xu J
