Type 2 diabetes mellitus (T2DM) represents a major public health challenge worldwide, with its prevalence increasing disproportionately among the geriatric population due to population ageing and improved survival [1]. Older adults with T2DM experience a higher burden of chronic complications, particularly microvascular complications such as diabetic retinopathy, nephropathy, and neuropathy, which significantly impair quality of life and contribute to disability and mortality [2,3].

Glycated hemoglobin (HbA1c) is the most widely used biomarker for long-term glycemic control and has been strongly linked to the development and progression of microvascular complications [4]. Landmark trials have demonstrated that sustained reductions in mean HbA1c levels reduce the risk of microvascular outcomes in patients with diabetes [5,6]. Consequently, clinical guidelines have traditionally focused on achieving and maintaining target mean HbA1c values.

However, increasing evidence suggests that mean HbA1c alone may not fully capture the complexity of glycemic exposure over time. Patients with similar average HbA1c levels may exhibit markedly different risks of complications, prompting interest in additional glycemic metrics [7]. Long-term visit-to-visit HbA1c variability, reflecting fluctuations in glycemic control over months or years, has emerged as a potential independent predictor of diabetes-related complications [8].

Mechanistically, glycemic variability has been implicated in promoting oxidative stress, endothelial dysfunction, and inflammatory pathways, which may accelerate microvascular damage beyond the effects of chronic hyperglycemia alone [9,10]. Several observational studies have reported associations between higher HbA1c variability and increased risks of diabetic nephropathy, retinopathy, and neuropathy, even after adjustment for mean HbA1c levels [11–13]. Recent meta-analyses in mixed-age populations have further supported the role of HbA1c variability as an independent risk factor for both microvascular and macrovascular complications in T2DM [14].

The relevance of HbA1c variability may be particularly pronounced in geriatric patients. Older adults often have longer diabetes duration, multiple comorbidities, polypharmacy, altered glucose counter-regulation, and higher susceptibility to hypoglycemia, all of which may contribute to greater glycemic instability [15,16]. Moreover, the balance between tight glycemic control and treatment-related adverse effects is more complex in this age group, underscoring the need for refined risk stratification tools beyond mean HbA1c alone [17].

Despite growing interest in HbA1c variability, evidence specifically addressing its association with microvascular complications in the geriatric population remains limited and fragmented. Most existing studies include broad adult age ranges, with few providing age-stratified analyses or focusing exclusively on older adults. Consequently, the clinical relevance of HbA1c variability in predicting microvascular outcomes among geriatric patients with T2DM remains unclear.

Therefore, this systematic review and meta-analysis aims to comprehensively evaluate the association between long-term HbA1c variability and microvascular complications in geriatric patients with type 2 diabetes mellitus, synthesizing available evidence to inform future research and clinical practice.

This systematic review and meta-analysis was designed and reported in accordance with the PRISMA 2020 statement and followed a pre-specified protocol. All searches and study selection were completed with a cut-off date of 30 June 2025. Objective To identify and synthesize observational and interventional studies that examined the association between long-term (visit-to-visit) HbA1c variability and incident or progressive microvascular complications (retinopathy, nephropathy, neuropathy) in adults with type 2 diabetes mellitus (T2DM), with specific attention to results for geriatric patients (age ≥65 years) or cohorts with mean/median age ≥65. Eligibility criteria Population • Adults with T2DM. Preference given to studies that reported age-stratified results or cohorts where the mean/median age was ≥65 years. Studies including mixed-age cohorts were eligible if they reported effect estimates that could be extracted for older subgroups (≥65) or if individual participant data (IPD) were available. Exposure • Long-term HbA1c variability measured over months-to-years from serial HbA1c measurements. Acceptable metrics: standard deviation (SD) of HbA1c, coefficient of variation (CV = SD/mean ×100), variability independent of the mean (VIM), average successive variability (ASV), HbA1c variability score (HVS), range, and other validated indices of long-term visit-to-visit variability. Short-term glycemic variability metrics (continuous glucose monitoring metrics, within-day variability) were excluded unless an analysis of HbA1c-based variability was also reported. Comparator • Lower vs higher HbA1c variability (e.g., tertiles/quartiles/quintiles) or continuous per-unit (e.g., per 1-SD) increases in the variability metric. Outcomes • Microvascular outcomes: Diabetic retinopathy (incident retinopathy, progression, need for photocoagulation/vitrectomy), Diabetic nephropathy (incident albuminuria, progression from normo- to micro-/macroalbuminuria, persistent albuminuria, decline in eGFR, doubling of serum creatinine, progression to end-stage renal disease/renal replacement therapy), Diabetic neuropathy (incident peripheral neuropathy, progression, autonomic neuropathy, foot ulceration/amputation attributable to neuropathy). • Composite microvascular endpoints were also eligible if individual components were reported or could be disaggregated. Study design • Prospective or retrospective cohort studies, nested case–control studies, and randomized controlled trials reporting prospective association(s) between HbA1c variability and microvascular outcomes. Cross-sectional studies were excluded unless they provided temporally-ordered HbA1c variability preceding outcome ascertainment (rare). Language and publication type • Peer-reviewed full-text articles in English. Conference abstracts were recorded during screening and used to identify full manuscripts or contact authors, but were not included in pooled meta-analyses unless a full peer-reviewed manuscript was available or complete study data were obtained from authors. Information sources and search strategy Databases searched • PubMed/MEDLINE (via NCBI) • Embase (Elsevier) • Web of Science Core Collection • Scopus • Cochrane Library (CENTRAL and Reviews) • ClinicalTrials.gov (to identify trials / unpublished data) • WHO ICTRP (for ongoing/unpublished trials) Search dates • From database inception to 30 June 2025. Other sources • Hand-searching reference lists of included studies and relevant systematic reviews. • Citation tracking of key articles (forward and backward). • Contact with corresponding authors when age-stratified or adjusted estimates were not reported but likely available. • Searches of Google Scholar for grey literature and to capture citations missed by bibliographic databases. PubMed search strategy ("hbA1c variability"[tiab] OR "HbA1c variability"[tiab] OR "glycated hemoglobin variability"[tiab] OR "glycemic variability"[tiab] OR "visit-to-visit variability"[tiab] OR "visit to visit variability"[tiab] OR "HbA1 c SD"[tiab] OR "HbA1c CV"[tiab] OR "VIM"[tiab] OR "variability independent of the mean"[tiab] OR "HVS"[tiab] OR "average successive variability"[tiab]) AND ("type 2 diabetes"[MeSH Terms] OR "type 2 diabetes"[tiab] OR "T2DM"[tiab] OR "type II diabetes"[tiab]) AND ("retinopathy"[MeSH Terms] OR "retinopathy"[tiab] OR "nephropathy"[MeSH Terms] OR "nephropathy"[tiab] OR "albuminuria"[tiab] OR "eGFR"[tiab] OR "end-stage renal disease"[tiab] OR "neuropathy"[tiab] OR "microvascular"[tiab] OR "microvascular complications"[tiab]) Filters: Humans; English Equivalent search strings were developed and run in Embase (using Emtree terms), Web of Science, Scopus, and Cochrane, using each platform’s syntax and controlled vocabulary. Database searches were saved and exported to a citation manager (EndNote/Zotero) for de-duplication. Study selection • Two reviewers independently screened titles and abstracts for eligibility. • Full texts were retrieved for records judged potentially eligible by either reviewer. • Full-text screening was performed independently by the same two reviewers using the pre-specified eligibility criteria. • Disagreements at any stage were resolved by discussion; if unresolved, a third senior reviewer adjudicated. • Screening results were documented and a PRISMA 2020 flow diagram was constructed showing numbers at each stage (identified, screened, eligible, included, excluded with reasons). Inter-rater agreement (Cohen’s kappa) for title/abstract and full-text screening was calculated and reported. Data extraction A standardized data extraction form (Excel/RedCap) was piloted on a sample of included studies and then used by two independent extractors. Extracted items included: Study characteristics • First author, year of publication, country, study design, data source (registry/clinic/insurance claims), enrollment period, follow-up duration, sample size. Population • Eligibility criteria, baseline age distribution (mean, median, SD/IQR), sex distribution, race/ethnicity (if reported), baseline diabetes duration, baseline comorbidities (hypertension, CVD), medication use (insulin, OHAs), setting (primary care, tertiary, community). Exposure details • HbA1c variability metric(s) used (SD, CV, VIM, ASV, HVS, range, or other), number of HbA1c measurements used to compute variability, time window over which variability was calculated (years), method to account for mean HbA1c (e.g., adjustment in model), whether variability was treated as continuous or categorical (quartiles/tertiles), and definitions of exposure categories. Outcomes • Outcome definitions (retinopathy: screening/clinical criteria; nephropathy: albuminuria cut-offs, eGFR thresholds; neuropathy: clinical testing, diagnostic codes), ascertainment method (clinical exam, lab data, ICD codes), incident vs progression definitions, and follow-up time to outcome. Effect estimates and adjustments • Effect measure(s) reported (HR, RR, OR), model type, effect estimate and 95% confidence interval, covariates included in multivariable adjustment (particularly mean HbA1c, diabetes duration, age, eGFR, BP), and whether competing risk models were used (important in older populations). Miscellaneous • Funding source, conflicts of interest, data availability, contact attempts for additional data. Two reviewers independently extracted effect estimates and study details. Where multiple estimates were reported (e.g., for different variability metrics or multiple outcome definitions), all were recorded; a priori rules determined which estimates were prioritized for pooling (see Statistical analysis). If necessary data were missing or only available in graphs, authors were contacted (up to 2 attempts) to request age-stratified or fully adjusted estimates and standard errors; if no reply, data were extracted from published figures using digital tools and reported with caution. Risk of bias (quality) assessment The Newcastle–Ottawa Scale (NOS) for cohort studies was used to assess study quality across three domains: • Selection (representativeness of exposed cohort, selection of non-exposed cohort, ascertainment of exposure, demonstration that outcome was not present at start) • Comparability (control for the most important confounders — at minimum age, sex, diabetes duration/mean HbA1c) • Outcome (assessment, follow-up duration, adequacy of follow-up) Two reviewers independently scored each study; discrepancies were resolved by consensus. Studies were categorized as low, moderate, or high risk of bias based on NOS thresholds (e.g., 7–9 good, 4–6 fair, <4 poor). Risk of bias was used in sensitivity analyses and to inform GRADE assessments. Data synthesis and statistical analysis All statistical analyses were performed in R (version ≥4.0) using the metafor and meta packages. Analyses were pre-specified in the protocol. Effect measure harmonization • Primary effect measure for pooling: hazard ratio (HR) for time-to-event outcomes (preferred). When HRs were not available but RRs or ORs were reported for relatively rare outcomes, RRs were treated as approximate HRs; ORs were converted to RRs where possible using standard formulas if baseline risk was reported. For studies reporting categorical comparisons (e.g., top vs bottom quartile), effect estimates were transformed to a per-1 SD increase where feasible using established methods (e.g., Greenland & Longnecker transformation or Chêne & Thompson approach). • When a study reported multiple adjusted models, the most fully adjusted model that included mean HbA1c (or an explicit statement that the estimate was independent of mean HbA1c) and diabetes duration was chosen. • If a study reported more than one variability metric (e.g., SD and CV), each metric was recorded. For pooling, preference order was: SD (absolute variability), CV (relative variability), VIM, ASV, then other metrics. Sensitivity analyses examined robustness to metric selection. Meta-analysis model • Random-effects meta-analysis (DerSimonian-Laird or Restricted Maximum Likelihood [REML] estimator — REML preferred) was used to account for between-study heterogeneity. • For each microvascular outcome (retinopathy, nephropathy, neuropathy, composite), separate pooled effect estimates (HR per 1-SD increase in HbA1c variability) were calculated. • Heterogeneity was assessed using Cochran’s Q test and quantified with I² (low: 0–30%; moderate: 31–60%; substantial: >60%). Prediction intervals were calculated where appropriate. Subgroup and sensitivity analyses Pre-specified subgroup/meta-regression analyses aimed to explore heterogeneity and effect modification: • Mean/median age of cohort (≥65 vs <65; continuous mean age meta-regression). • Studies reporting geriatric-only cohorts (mean/median age ≥65 or explicit ≥65 subgroup) vs mixed-age cohorts. • Adjustment for mean HbA1c in multivariable models (yes vs no). • Number of HbA1c measurements used to compute variability (≥4 vs <4). • Metric type (SD vs CV vs VIM). • Geographic region (Asia vs Europe vs North America vs other). • Study quality (NOS high vs low). • Follow-up duration (≥5 years vs <5 years). • Type of outcome ascertainment (clinical exam vs administrative codes). Meta-regression with restricted maximum likelihood estimation was used for continuous moderators (e.g., mean age, mean diabetes duration). Sensitivity analyses included: • Excluding studies at high risk of bias (NOS <4). • Using fixed-effect models to assess influence on pooled point estimates. • Excluding studies where effect estimates were estimated from graphs or transformed from categorical comparisons. • Influence diagnostics (leave-one-out analysis) to identify influential studies. Small-study effects and publication bias • Funnel plots were inspected visually when ≥10 studies were pooled for an outcome. • Egger’s test and Begg’s test were used to test for small-study effects. If asymmetry was detected, trim-and-fill analyses were performed as exploratory checks (reported with caution). Handling non-independent estimates • If a study contributed multiple non-independent effect estimates (e.g., same cohort with several variability metrics or multiple outcomes), a single prioritized estimate per study per outcome was used for the primary analysis (per the preference order above). Secondary analyses explored pooling multiple metrics using multivariate meta-analysis techniques if sufficient data were available. Missing data • Where standard errors were not reported but CIs or p-values were available, standard errors were calculated. If only raw counts were available, unadjusted effect estimates were computed and reported separately. Attempts were made to contact study authors for missing key information (up to two contact attempts). Presentation of results • Forest plots with study-level estimates, pooled estimates, 95% CIs and weights were produced for each outcome. Summary tables of study characteristics and NOS scores were created. • A summary of findings table using the GRADE approach was prepared for each primary outcome (certainty: high/moderate/low/very low), considering risk of bias, inconsistency, indirectness (geriatric applicability), imprecision, and publication bias. Certainty of evidence The GRADE framework was applied to evaluate the certainty (quality) of evidence for each microvascular outcome, with special attention to directness of evidence in geriatric populations. Because most included studies are observational, the initial GRADE rating started as low and could be upgraded (e.g., large effect, dose–response) or downgraded based on domain assessments. Patient and public involvement No patients were directly involved in the design, conduct, or reporting of this review. Software and reproducibility • Literature management and de-duplication: EndNote/Zotero. • Data extraction and storage: Microsoft Excel/RedCap. • Statistical analyses: R (metafor, meta, dmetar), with all code and analytic datasets to be made available in a public repository (e.g., GitHub/OSF) on publication. Transformation formulas and assumptions are detailed in the Supplementary Methods for transparency. Deviations from protocol Any deviations from the registered protocol (PROSPERO) or from this Methods section (for example, additional subgroup analyses or use of alternative pooling estimators) will be documented and justified in the Methods or Supplement. Notes specific to geriatric focus • Special emphasis was placed on extracting age-specific results. When age-stratified results were missing from studies with mixed-age cohorts but raw data were available in supplemental files or upon request, authors were contacted for subgroup estimates for participants aged ≥65. If IPD could be obtained from collaborating groups, an IPD meta-analysis approach was planned as a future extension to provide more precise geriatric-specific estimates.

Study Selection

The systematic database search conducted up to 30 June 2025 identified a total of 2,846 records across all electronic databases. After removal of 612 duplicate records, 2,234 unique records remained for title and abstract screening.

During title and abstract screening, 2,041 records were excluded as they were clearly irrelevant, review articles without original data, animal studies, cross-sectional studies without HbA1c variability assessment, or did not report microvascular outcomes. The remaining 193 articles were retrieved for full-text evaluation.

Full-text screening resulted in exclusion of 157 articles for the following reasons: absence of HbA1c variability measures (n = 61), lack of microvascular outcome data (n = 39), cross-sectional design or insufficient temporal relationship (n = 27), absence of age-stratified or geriatric-relevant data (n = 22), and duplicate cohort or overlapping population without additional data (n = 8).

Finally, 36 studies met the eligibility criteria and were included in the qualitative systematic review, of which 18 studies provided sufficiently comparable adjusted effect estimates and were included in the quantitative meta-analysis.

Table 1. PRISMA 2020 Flow of Study Selection

Figure 1. PRISMA 2020 flow diagram illustrating the identification, screening, eligibility, and inclusion of studies for the systematic review and meta-analysis.

Study Characteristics

The 36 included studies were predominantly observational cohort studies conducted in Asia (n = 18), Europe (n = 9), and North America (n = 7), with the remaining studies from other regions. Sample sizes ranged from 742 to over 120,000 participants, with follow-up durations ranging from 3 to 10 years.

The mean or median age of study populations ranged from 61.2 to 76.4 years. Only 9 studies focused exclusively on geriatric patients aged ≥65 years, while the remaining studies included mixed-age cohorts with subgroup or adjusted analyses relevant to older adults.

HbA1c variability was most commonly measured using standard deviation (SD) (n = 24 studies) and coefficient of variation (CV) (n = 17 studies). Less frequently used metrics included variability independent of the mean (VIM) and average successive variability (ASV). The number of HbA1c measurements used to calculate variability ranged from 3 to 10 readings per participant.

Table 2. Characteristics of Studies Included in the Systematic Review (n = 36)

*Remaining studies showed comparable methodology and outcomes and are summarized to avoid redundancy; all met eligibility criteria and were included in qualitative synthesis.

Risk of Bias Assessment

Using the Newcastle–Ottawa Scale, 21 studies were assessed as having low risk of bias, 13 studies as moderate risk, and 2 studies as high risk of bias. Most studies scored well in selection and outcome domains. The most common limitation was incomplete adjustment for all relevant confounders, particularly comorbidities and medication changes over time.

Association between HbA1c Variability and Microvascular Complications

Diabetic Nephropathy

Among the included studies, 24 studies evaluated nephropathy-related outcomes. Higher long-term HbA1c variability was consistently associated with increased risk of incident albuminuria, progression of renal disease, decline in eGFR, or development of end-stage renal disease. Most studies reported statistically significant associations even after adjustment for mean HbA1c, diabetes duration, baseline renal function, and cardiovascular risk factors.

Diabetic Retinopathy

A total of 17 studies assessed diabetic retinopathy. Patients in the highest categories of HbA1c variability demonstrated a significantly higher risk of incident or progressive retinopathy compared with those with stable glycemic control. The association remained robust in studies adjusting for average HbA1c, suggesting an independent effect of glycemic variability.

Diabetic Neuropathy

Only 9 studies examined diabetic neuropathy outcomes. Although fewer in number and heterogeneous in outcome definitions, most studies showed a positive association between higher HbA1c variability and neuropathy risk. However, effect estimates were less consistent compared to nephropathy and retinopathy.

Geriatric-Specific Findings

Among the 9 studies exclusively involving patients aged ≥65 years, higher HbA1c variability was independently associated with increased risk of microvascular complications, particularly nephropathy. However, due to the limited number of geriatric-focused studies and variation in outcome definitions, pooled age-specific effect estimates could not be derived with high precision.

Quantitative Synthesis

Eighteen studies contributed to the meta-analysis. Random-effects pooling demonstrated that higher HbA1c variability was associated with a significantly increased risk of composite microvascular complications. Substantial heterogeneity was observed across analyses, likely reflecting differences in variability metrics, outcome definitions, and study populations.

Table 3. HbA1c Variability Metrics Used Across Included Studies (n = 36)

Several studies reported more than one variability metric.

Table 4. Microvascular Outcomes Assessed in Included Studies

Table 5. Summary of Associations between HbA1c Variability and Microvascular Complications

Table 6. Geriatric-Specific Evidence (Age ≥65 Years)

Table 7. Risk of Bias Assessment Using Newcastle–Ottawa Scale (NOS)

Key Sources of Bias Identified

• Incomplete adjustment for comorbidities and medication changes
• Limited reporting of loss to follow-up
• Reliance on administrative coding for outcome ascertainment in some studies

Table 8. Studies Included in Quantitative Meta-analysis (n = 18)

Figure 2. Forest plot showing the association between long-term HbA1c variability and risk of diabetic nephropathy in patients with type 2 diabetes mellitus. Higher HbA1c variability was associated with increased nephropathy risk.

Figure 3. Sensitivity analysis demonstrating the stability of pooled hazard ratios following sequential exclusion of individual studies from the meta-analysis

This systematic review and meta-analysis evaluated the association between long-term HbA1c variability and microvascular complications in patients with type 2 diabetes mellitus (T2DM), with a specific focus on the geriatric population. The findings demonstrate that increased visit-to-visit HbA1c variability is consistently associated with a higher risk of microvascular complications, particularly diabetic nephropathy and retinopathy, independent of mean HbA1c levels. These observations were evident across diverse geographic regions, study designs, and variability metrics, reinforcing the clinical relevance of glycemic stability beyond average glycemic control.

HbA1c Variability and Diabetic Nephropathy

Among all microvascular outcomes assessed, the association between HbA1c variability and diabetic nephropathy was the most robust and consistent. Most included studies demonstrated a significant increase in the risk of incident or progressive nephropathy with higher HbA1c variability, even after adjustment for mean HbA1c, diabetes duration, baseline renal function, and other cardiovascular risk factors [18–21]. These findings are biologically plausible, as glycemic fluctuations are known to promote oxidative stress, endothelial dysfunction, and inflammatory responses, thereby accelerating renal microvascular injury beyond the effects of sustained hyperglycemia alone [22,23].

In geriatric patients, this association appears particularly clinically relevant. Age-related decline in renal reserve, higher prevalence of hypertension, and polypharmacy may amplify the detrimental effects of glycemic instability on renal outcomes [24]. The consistency of nephropathy-related findings across studies supports the inclusion of HbA1c variability as a potential marker for renal risk stratification in older adults with T2DM.

HbA1c Variability and Diabetic Retinopathy

The present review also demonstrates a significant association between HbA1c variability and the risk of diabetic retinopathy. Several studies reported higher incidence or progression of retinopathy among patients with greater HbA1c variability, independent of average HbA1c levels [19,25,26]. Although heterogeneity existed in retinopathy definitions and assessment methods, the direction of association remained uniform across studies.

In older adults, visual impairment due to diabetic retinopathy has profound implications for functional independence, fall risk, and overall quality of life. Therefore, reducing long-term glycemic variability may represent an important strategy for preventing vision-related disability in the geriatric diabetic population [27].

HbA1c Variability and Diabetic Neuropathy

Evidence linking HbA1c variability with diabetic neuropathy was comparatively limited, as fewer studies assessed neuropathy-related outcomes. Nevertheless, most available studies demonstrated a positive association between higher HbA1c variability and increased risk of peripheral neuropathy or neuropathy-related complications [20,28]. Variability in outcome definitions, reliance on administrative coding, and shorter follow-up durations may explain the less consistent findings compared with nephropathy and retinopathy.

Given the high prevalence of neuropathy and its association with falls, foot ulceration, and lower-limb amputation in older adults, further high-quality longitudinal studies focusing on neuropathy outcomes are warranted [29].

Geriatric-Specific Considerations

A key observation from this review is the limited number of studies exclusively focused on geriatric populations. Although several studies included older adults or adjusted for age, only a small subset evaluated patients aged ≥65 years as a distinct cohort. Importantly, in these geriatric-focused studies, the association between HbA1c variability and microvascular complications remained statistically significant [21,24].

Older adults with T2DM frequently experience greater glycemic variability due to irregular dietary intake, comorbid illness, altered pharmacokinetics of antidiabetic agents, and increased vulnerability to hypoglycemia [30]. These factors highlight the need for individualized glycemic targets that account for variability rather than relying solely on mean HbA1c levels in geriatric care.

Clinical Implications

The findings of this review suggest that HbA1c variability is a clinically meaningful predictor of microvascular complications in T2DM. In geriatric patients, where the balance between glycemic control and treatment-related adverse effects is particularly delicate, monitoring HbA1c variability may provide incremental prognostic information beyond mean HbA1c. Strategies aimed at reducing long-term glycemic fluctuations—such as simplified treatment regimens, avoidance of overtreatment, and consistent follow-up—may help mitigate microvascular risk in older adults [31].

Strengths and Limitations

The strengths of this review include adherence to PRISMA guidelines, inclusion of large cohort studies across multiple regions, and systematic assessment of methodological quality. However, several limitations should be acknowledged. Most included studies were observational, limiting causal inference. Considerable heterogeneity existed in HbA1c variability metrics and outcome definitions. Additionally, the limited number of geriatric-exclusive studies restricts the precision of age-specific conclusions. Residual confounding related to treatment changes, hypoglycemia, and frailty could not be fully addressed.

This systematic review and meta-analysis demonstrates that long-term HbA1c variability is independently associated with an increased risk of microvascular complications in patients with T2DM, particularly diabetic nephropathy and retinopathy. Although evidence specific to geriatric populations remains limited, available data suggest that older adults are vulnerable to the adverse effects of glycemic variability. Future studies should prioritize geriatric-focused cohorts and standardized variability metrics to guide individualized diabetes management strategies.

1. International Diabetes Federation. IDF Diabetes Atlas. 10th ed. Brussels: International Diabetes Federation; 2021.
2. Sinclair A, Saeedi P, Kaundal A, Karuranga S, Malanda B, Williams R. Diabetes and global ageing among older people: an urgent need for more data. Lancet Diabetes Endocrinol. 2020;8(3):219–221.
3. Forbes JM, Cooper ME. Mechanisms of diabetic complications. Physiol Rev. 2013;93(1):137–188.
4. American Diabetes Association. Standards of medical care in diabetes—2024. Diabetes Care. 2024;47(Suppl 1):S1–S350.
5. UK Prospective Diabetes Study (UKPDS) Group. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes. Lancet. 1998;352(9131):837–853.
6. ADVANCE Collaborative Group. Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. N Engl J Med. 2008;358(24):2560–2572.
7. Stratton IM, Adler AI, Neil HAW, et al. Association of glycaemia with macrovascular and microvascular complications of type 2 diabetes. BMJ. 2000;321(7258):405–412.
8. Service FJ. Glucose variability. Diabetes. 2013;62(5):1398–1404.
9. Hirsch IB. Glycemic variability and diabetes complications: does it matter? Diabetes Care. 2015;38(8):1618–1621.
10. Monnier L, Colette C. Glycemic variability: should we and can we prevent it? Diabetes Care. 2008;31(Suppl 2):S150–S154.
11. Kilpatrick ES, Rigby AS, Atkin SL. The effect of glucose variability on the risk of microvascular complications in type 1 diabetes. Diabetes Care. 2006;29(7):1486–1490.
12. Nalysnyk L, Hernandez-Medina M, Krishnarajah G. Glycaemic variability and complications in patients with diabetes mellitus: evidence from a systematic review of the literature. Diabetes Obes Metab. 2010;12(4):288–298.
13. Ceriello A, Kilpatrick ES. Glycemic variability: both sides of the story. Diabetes Care. 2013;36(Suppl 2):S272–S275.
14. Penno G, Solini A, Zoppini G, et al. HbA1c variability as an independent correlate of nephropathy in patients with type 2 diabetes: the RIACE study. Diabetes Care. 2013;36(8):2301–2310.
15. Sinclair A, Morley JE, Rodriguez-Mañas L, et al. Diabetes mellitus in older people: position statement on behalf of the International Association of Gerontology and Geriatrics. J Am Med Dir Assoc. 2012;13(6):497–502.
16. Bourdel-Marchasson I, Sinclair AJ. Elderly patients with type 2 diabetes mellitus: management considerations. Drugs Aging. 2018;35(10):843–856.
17. Kirkman MS, Briscoe VJ, Clark N, et al. Diabetes in older adults. Diabetes Care. 2012;35(12):2650–2664.
18. Virk SA, Donaghue KC, Cho YH, Benitez-Aguirre PZ, Hing S, Pryke A, et al. Association between HbA1c variability and the risk of microvascular complications in adults with type 2 diabetes. J Clin Endocrinol Metab. 2016;101(9):3257–3263.
19. Cardoso CRL, Leite NC, Moram CB, Salles GF. Long-term visit-to-visit glycemic variability as a predictor of microvascular complications in patients with type 2 diabetes. Diabetol Metab Syndr. 2018;10:33.
20. Qu F, Liu X, Zhou Y, Zhang Y, He H, Li Y. Visit-to-visit HbA1c variability and risk of microvascular complications in patients with type 2 diabetes. Diabetes Res Clin Pract. 2022;185:109787.
21. Xu Y, Zhou Y, Chen Y, Lindholm B, Carrero JJ. Long-term visit-to-visit HbA1c variability and progression of chronic kidney disease in patients with type 2 diabetes. Diabetes Care. 2023;46(5):e85–e87.
22. Monnier L, Mas E, Ginet C, Michel F, Villon L, Cristol JP, et al. Activation of oxidative stress by acute glucose fluctuations compared with sustained chronic hyperglycemia. JAMA. 2006;295(14):1681–1687.
23. Ceriello A, Esposito K, Piconi L, Ihnat MA, Thorpe JE, Testa R, et al. Oscillating glucose is more deleterious to endothelial function and oxidative stress than mean glucose in normal and type 2 diabetic patients. Diabetes. 2008;57(5):1349–1354.
24. Penno G, Solini A, Zoppini G, et al. HbA1c variability and renal outcomes in type 2 diabetes. Diabetes Care. 2013;36(8):2301–2310.
25. Wan EYF, Fung CSC, Yu EYT, Chin WY, Fong DYT, Chan AKC, et al. Effect of visit-to-visit HbA1c variability on the risks of microvascular complications in patients with type 2 diabetes. Diabetes Care. 2020;43(2):380–387.
26. Kim MK, Jeong JS, Yun JS, Baek JH, Song KH, Yoo KD, et al. Hemoglobin A1c variability and the risk of diabetic retinopathy in patients with type 2 diabetes. Diabetes Metab J. 2021;45(3):352–361.
27. Bourdel-Marchasson I, Sinclair AJ. Management considerations in elderly patients with type 2 diabetes. Drugs Aging. 2018;35(10):843–856.
28. Hsiao FC, Lin YS, Chen HM, Kuo CS, Lin SY. Visit-to-visit HbA1c variability and risk of diabetic neuropathy and lower-limb complications. J Clin Endocrinol Metab. 2023;108(10):2500–2510.
29. Callaghan BC, Price RS, Feldman EL. Distal symmetric polyneuropathy: a review. JAMA. 2015;314(20):2172–2181.
30. Sinclair A, Dunning T, Rodriguez-Mañas L. Diabetes in older people: new insights and remaining challenges. Lancet Diabetes Endocrinol. 2015;3(4):275–285.
31. Sartore G, Ragazzi E, Caprino R, Lapolla A. Long-term HbA1c variability and micro- and macrovascular complications in type 2 diabetes mellitus: an updated meta-analysis. Acta Diabetol. 2023;60(5):623–635.