Dementia Risk Increase In Diabetic Patients With Poor Glycemic Control
Overview
This study investigated the association between glycemic variability and dementia risk among 171,964 individuals with type 2 diabetes. Despite optimal glycosylated hemoglobin A1c (HbA1c) levels, variability in HbA1c over the long term was found to increase the hazard of dementia. The analysis, which adjusted for demographics and comorbidities, revealed that greater variability in HbA1c (measured by standard deviation) was linked to a higher dementia hazard, with an adjusted hazard ratio of 1.15 (95% confidence interval: 1.12, 1.17).
Stratified analyses showed that individuals with higher variability in HbA1c, particularly in the highest quintiles, faced a significantly greater risk of dementia, especially among those maintaining mean HbA1c levels below 6% or between 6% and 8%. However, this association was not observed among individuals with mean HbA1c levels equal to or greater than 8%.
These findings underscore the clinical significance of minimizing glycemic variability to reduce dementia risk, even in individuals maintaining HbA1c levels within optimal clinical targets. The study contributes to the growing understanding of the impact of glycemic control on cognitive health and emphasizes the importance of consistent management strategies in diabetes care.
Introduction
Type 2 diabetes is linked to a nearly twofold increased risk of dementia, yet the precise mechanisms behind this association remain unclear. While glycemic control, typically assessed through mean levels of glycated hemoglobin A1c (HbA1c), is often proposed as a mediator, large-scale trials focusing on stringent mean glycemic control have not conclusively shown reduced dementia risk. A less explored aspect of glycemic control is variability, which is frequently studied in relation to dysglycemic events. However, few studies have separately investigated visit-to-visit HbA1c variability, despite its association with various adverse health outcomes, such as cardiovascular disease and mortality, independent of mean glycemic levels.
Recent meta-analytic findings have suggested that greater visit-to-visit HbA1c variability correlates with an elevated risk of dementia. This study aims to delve deeper into this relationship using diverse metrics of HbA1c variability and to explore potential variations in this association based on mean HbA1c levels, as well as factors such as sex, race, and ethnicity. Conducted within a large, integrated healthcare system in Northern California, this research seeks to provide insights into how HbA1c variability may influence dementia risk across different demographic and clinical profiles.
Method
Kaiser Permanente Northern California (KPNC) is a comprehensive healthcare system serving over 4.5 million members, encompassing approximately 30% of its surrounding geographic area. Compared to insured adults in the same region, KPNC members show similar demographic profiles in terms of race, ethnicity, educational attainment, and health indicators, but are less likely to experience low socioeconomic status. The KPNC Diabetes Registry, utilizing a validated algorithm with 99% sensitivity, identifies individuals with diabetes using pharmacy and laboratory data, hospitalization records, and outpatient diagnoses.
For this study, researchers focused on individuals aged 50 years and older with type 2 diabetes from January 1, 1996, to December 31, 2018, totaling 494,828 individuals. Exclusions were made for prevalent dementia, less than 5 years of follow-up, or a history of hospital admissions for severe glycemic events at baseline. Additionally, individuals without at least one HbA1c measurement per year during the initial 3 years of follow-up were excluded. Cohort entry was defined as the first date an individual met the age and type 2 diabetes criteria, and follow-up continued until dementia diagnosis, membership lapse, death, or the end of the study period.
Dementia diagnoses were identified through electronic medical records using specific ICD-9/10 codes for Alzheimer’s disease, non-specific dementia, and vascular dementia. HbA1c variability within the first 3 years of follow-up was assessed using multiple metrics including standard deviation (SD) and coefficient of variation. Various operationalizations were utilized to account for different aspects of variability measurement limitations, such as mean HbA1c, number of measurements, and outlier influence.
Demographic characteristics and baseline health conditions (e.g., peripheral artery disease, nephropathy, retinopathy) were obtained from electronic medical records at cohort entry. Mean HbA1c levels were calculated using all measurements from the same 3-year period used to determine HbA1c variability. This study, approved by the KPNC Internal Review Board, highlights the complex relationship between HbA1c variability and dementia risk in a large, diverse cohort of older adults with type 2 diabetes within a managed care setting.
Statistical Analysis
The study investigated the association between glycemic variability measured by glycosylated hemoglobin A1c (HbA1c) and the risk of dementia among over 170,000 individuals aged 50 years and older with type 2 diabetes. Cox proportional hazards models were utilized, with age as the timescale, to assess this relationship. The main exposure of interest was the standard deviation (SD) of HbA1c levels, both as a continuous measure and divided into quintiles, with the lowest quintile serving as the reference.
Several models were employed for adjustment: Model 1 included age, sex, race, and ethnicity; Model 2 further adjusted for baseline health conditions such as myocardial infarction, peripheral artery disease, neuropathy, nephropathy, stroke, retinopathy, and prior hyperosmolarity and ketoacidosis; and Model 3 additionally adjusted for the number of HbA1c measurements available.
Additionally, analyses were stratified by mean HbA1c levels to explore variations in the association between HbA1c variability and dementia risk across different clinical thresholds. Stratified analyses by sex and race/ethnicity were conducted to examine potential differences in these associations. Sensitivity analyses were performed varying the duration of glycemic control measurements from 5 to 10 years, while maintaining a 2-year lag between exposure and outcome periods to mitigate reverse causation.
The findings indicated that greater variability in HbA1c levels was significantly associated with an elevated risk of dementia, particularly among individuals with lower mean HbA1c levels. These results suggest that HbA1c variability may serve as a potential marker for dementia risk in individuals with type 2 diabetes, warranting further investigation into its modifiability and clinical implications. The study adhered to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for comprehensive reporting of observational research.
Result
This study analyzed data from 171,964 participants over an average follow-up period of 5.9 years, during which 9% developed dementia. The average age at study entry was 61.1 years, with 48% of participants being women. The sample was predominantly White (49%), followed by Asian (19%), Hispanic (15%), and Black (9%) individuals. Participants had an average of 6.2 HbA1c measurements in the first 3 years, with a mean HbA1c of 7.4%. Various measures of HbA1c variability, such as standard deviation (SD), coefficient of variation, and average real variability, were assessed for their association with dementia risk.
The findings indicated that greater HbA1c variability was consistently associated with an increased hazard of dementia across multiple metrics. Adjusted hazard ratios (aHR) showed significant associations for SD (aHR = 1.15), SD adjusted for number of measurements (aHR = 1.16), coefficient of variation (aHR = 2.56), z-scored coefficient of variation (aHR = 1.08), and average real variability (aHR = 1.10). These associations remained robust even after adjusting for baseline health conditions.
Further analysis by quintiles of SD and coefficient of variation revealed a graded increase in dementia hazard with higher variability. Participants in higher quintiles of SD or coefficient of variation showed progressively elevated risks of dementia compared to those in the lowest quintiles.
Stratified analyses by mean HbA1c levels indicated that the association between HbA1c variability and dementia risk was particularly pronounced in participants with mean HbA1c levels below 8%. However, this association was not significant for those with mean HbA1c levels of 8% or higher.
Overall, the study underscores the detrimental impact of glycemic variability on dementia risk, highlighting the importance of stabilizing glycemic control, especially in individuals with lower to moderate mean HbA1c levels, to potentially mitigate dementia risk. Further research is warranted to elucidate the underlying mechanisms and implications for clinical management in diverse populations.
Conclusion
In a comprehensive study involving over 170,000 individuals diagnosed with type 2 diabetes, researchers observed a significant correlation between higher variability in HbA1c levels and increased risk of dementia across various demographic groups, including sex, race, and ethnicity. This association persisted regardless of whether variability was measured by standard deviation or coefficient of variation. Individuals in the second to fifth quintiles of HbA1c variability exhibited a 20% to 50% higher risk of developing dementia compared to those in the first quintile (indicating lower variability). Importantly, these findings remained robust even after adjusting for age, race, ethnicity, sex, baseline health conditions, and frequency of HbA1c measurements.
Interestingly, the study also revealed that the relationship between HbA1c variability and dementia risk was influenced by long-term glycemic control, as reflected by mean HbA1c levels. The adverse effects of variability were most pronounced in individuals with mean HbA1c concentrations below 6%, a group traditionally associated with lower average glucose levels. This counterintuitive finding suggests that while maintaining lower average HbA1c levels is beneficial, high variability within this range may exacerbate dementia risk, possibly due to frequent excursions into hypoglycemic ranges which are known to be detrimental to brain health.
The mechanisms underlying these associations are complex and likely involve pathways related to inflammation, oxidative stress, vascular dysfunction, and potentially epigenetic changes. Previous studies have linked glycemic variability not only to cognitive decline but also to vascular complications and neuropathic changes, all of which are implicated in the pathogenesis of dementia.
Strengths of the study include its large and diverse sample size, comprehensive electronic health record data capturing longitudinal HbA1c measures and incident dementia, and inclusion of a substantial proportion of non-White participants. However, limitations include the retrospective nature of the data analysis, lack of specific details on certain dementia risk factors like frailty or obesity, and the use of HbA1c as a proxy for glycemic variability, which may not fully capture glucose fluctuations in all populations.
In conclusion, this study underscores the importance of minimizing glycemic variability in individuals with type 2 diabetes to potentially reduce the risk of dementia, particularly in those with tightly controlled but fluctuating HbA1c levels. Future research incorporating advanced glucose monitoring technologies and assessing the impact of specific antidiabetic therapies will be crucial in elucidating these relationships further and guiding clinical management strategies aimed at preserving brain health in diabetic populations.
Oncology Related Tools
Other
Latest Research
