In the Journals

HbA1c level, variability associated with mortality risk in older adults

Angus Forbes
Angus Forbes

Among older adults with diabetes, low and high HbA1c levels, as well as the amount of variability, were each associated with increased mortality risk, according to findings reported in The Lancet Diabetes & Endocrinology.

Data from observational studies have shown J-shaped distributions for mortality and glycemic control, and varied outcomes from recent trials assessing intensive glucose lowering in patients with type 2 diabetes have led to an emphasis on less stringent glycemic targets, Angus Forbes, PhD, professor of midwifery and palliative care at King’s College London, and colleagues wrote in the study background.

“Low, declining and variable glycemic control in older people seem to convey a risk that is only marginally explained by hypoglycemic agents,” Forbes told Endocrine Today. “In the older populations, it is important to consider HbA1c values and trends as a health marker and not as a target. The current system of care is biased toward lowering glycemic control. This may lead to a lack of vigilance in relation to high-risk patients with lower HbA1c variability.”

In a retrospective study, Forbes and colleagues analyzed data from 54,803 adults aged at least 70 years with type 1 or type 2 diabetes, from 587 U.K. primary care practices participating in The Health Improvement Network (THIN) database. Primary outcome was time to all-cause mortality; primary exposure variables were mean HbA1c and HbA1c variability. The observation included a 4-year run-in period beginning in 2003 as a baseline, with a 5-year follow-up from 2007 to 2012. Researchers assessed mean HbA1c in three models: a baseline mean HbA1c for 2003-2006 (model 1), the mean across the 5-year follow-up period (model 2) and a time-varying, yearly updated mean HbA1c (model 3). Researchers calculated a glycemic variability score from 0 (low) to 100 (high) based on number of changes in HbA1c of 0.5% or more between 2003 and 2012 or to the point of mortality, based on changes in the annual mean HbA1c (minimum of six readings).

Within the cohort, 8,614 women (mean baseline age, 79 years) and 9,066 men (mean baseline age, 78 years) died during follow-up. Mean diabetes duration at baseline was 8.48 years for women and 9.09 years for men. Mean baseline HbA1c was 7.23% for women and 7.22% for men.

In assessing HbA1c values over time, researchers observed a J-shaped distribution for both sexes in which mortality risk increased with HbA1c values of at least 8% and with HbA1c values less than 6%. However, in model 1, researchers did not observe an association between excess mortality and HbA1c between 8% and 8.5% for men, and there was no risk observed for men and women with HbA1c less than 6% when stratified by sex in models 2 and 3.

Mortality increased substantially with increasing HbA1c variability in all models for the overall cohort and in analyses for men and women individually, according to researchers.

When compared with older adults with a glycemic variability score between 0 and 20, risk for all-cause mortality increased for women (adjusted HR = 1.51; 95% CI, 1.3-1.75) and men (aHR = 1.57; 95% CI, 1.35-1.82) with a score between 80 and 100 in model 1. This risk rose in model 2, when the risk for all-cause mortality more than doubled in women (aHR = 2.47; 95% CI, 2.08-2.93) and men (aHR = 2.21; 95% CI, 1.87-2.61) with a high glycemic variability score vs. those with a low score. For model 3, aHRs for all-cause mortality with high glycemic variability were 1.87 (95% CI, 1.59-2.19) for women and 1.54 for men (95% CI, 1.32-1.8) vs. adults with low glycemic variability.

“Fitting the mean HbA1c models with the glycemic variability score altered the risk distribution; this observation was most marked in the model 2 analysis, in which an increased risk was only apparent with HbA1c values of at least 9.5% in women and 9% in men,” the researchers wrote.

Forbes said the study suggests that health care professionals must reconsider how they evaluate the J-shaped curve observed in the relation between HbA1c thresholds and mortality, adding that mortality at the lower end of the distribution might be due to age-related factors rather than excess glycemic intensification.

“The data may also suggest that maintaining a more stable glucose environment with more gradual intensification is better in older adults,” Forbes said. “It is important to note that as an observational study, these are only associations, and further research is required to model the relative benefits of different approaches to glucose management in older people. This work would need to consider the high level of heterogeneity in the population, considering differences in sex, comorbidities and, most important, frailty.” – by Regina Schaffer

For more information:

Angus Forbes, PhD, can be reached at the Florence Nightingale Faculty of Nursing, Midwifery and Palliative Care, King’s College London, SE1 8WA, United Kingdom; email: angus.forbes@kcl.ac.uk.

Disclosures: King’s College London and Diabetes Frail funded this study. Forbes reports he has received honoraria or unrestricted educational grants from Abbott, Ascensia Diabetes Care, Eli Lilly and Roche. Please see the study for the other authors’ relevant financial disclosures.

 

 

Angus Forbes
Angus Forbes

Among older adults with diabetes, low and high HbA1c levels, as well as the amount of variability, were each associated with increased mortality risk, according to findings reported in The Lancet Diabetes & Endocrinology.

Data from observational studies have shown J-shaped distributions for mortality and glycemic control, and varied outcomes from recent trials assessing intensive glucose lowering in patients with type 2 diabetes have led to an emphasis on less stringent glycemic targets, Angus Forbes, PhD, professor of midwifery and palliative care at King’s College London, and colleagues wrote in the study background.

“Low, declining and variable glycemic control in older people seem to convey a risk that is only marginally explained by hypoglycemic agents,” Forbes told Endocrine Today. “In the older populations, it is important to consider HbA1c values and trends as a health marker and not as a target. The current system of care is biased toward lowering glycemic control. This may lead to a lack of vigilance in relation to high-risk patients with lower HbA1c variability.”

In a retrospective study, Forbes and colleagues analyzed data from 54,803 adults aged at least 70 years with type 1 or type 2 diabetes, from 587 U.K. primary care practices participating in The Health Improvement Network (THIN) database. Primary outcome was time to all-cause mortality; primary exposure variables were mean HbA1c and HbA1c variability. The observation included a 4-year run-in period beginning in 2003 as a baseline, with a 5-year follow-up from 2007 to 2012. Researchers assessed mean HbA1c in three models: a baseline mean HbA1c for 2003-2006 (model 1), the mean across the 5-year follow-up period (model 2) and a time-varying, yearly updated mean HbA1c (model 3). Researchers calculated a glycemic variability score from 0 (low) to 100 (high) based on number of changes in HbA1c of 0.5% or more between 2003 and 2012 or to the point of mortality, based on changes in the annual mean HbA1c (minimum of six readings).

Within the cohort, 8,614 women (mean baseline age, 79 years) and 9,066 men (mean baseline age, 78 years) died during follow-up. Mean diabetes duration at baseline was 8.48 years for women and 9.09 years for men. Mean baseline HbA1c was 7.23% for women and 7.22% for men.

In assessing HbA1c values over time, researchers observed a J-shaped distribution for both sexes in which mortality risk increased with HbA1c values of at least 8% and with HbA1c values less than 6%. However, in model 1, researchers did not observe an association between excess mortality and HbA1c between 8% and 8.5% for men, and there was no risk observed for men and women with HbA1c less than 6% when stratified by sex in models 2 and 3.

Mortality increased substantially with increasing HbA1c variability in all models for the overall cohort and in analyses for men and women individually, according to researchers.

When compared with older adults with a glycemic variability score between 0 and 20, risk for all-cause mortality increased for women (adjusted HR = 1.51; 95% CI, 1.3-1.75) and men (aHR = 1.57; 95% CI, 1.35-1.82) with a score between 80 and 100 in model 1. This risk rose in model 2, when the risk for all-cause mortality more than doubled in women (aHR = 2.47; 95% CI, 2.08-2.93) and men (aHR = 2.21; 95% CI, 1.87-2.61) with a high glycemic variability score vs. those with a low score. For model 3, aHRs for all-cause mortality with high glycemic variability were 1.87 (95% CI, 1.59-2.19) for women and 1.54 for men (95% CI, 1.32-1.8) vs. adults with low glycemic variability.

“Fitting the mean HbA1c models with the glycemic variability score altered the risk distribution; this observation was most marked in the model 2 analysis, in which an increased risk was only apparent with HbA1c values of at least 9.5% in women and 9% in men,” the researchers wrote.

Forbes said the study suggests that health care professionals must reconsider how they evaluate the J-shaped curve observed in the relation between HbA1c thresholds and mortality, adding that mortality at the lower end of the distribution might be due to age-related factors rather than excess glycemic intensification.

“The data may also suggest that maintaining a more stable glucose environment with more gradual intensification is better in older adults,” Forbes said. “It is important to note that as an observational study, these are only associations, and further research is required to model the relative benefits of different approaches to glucose management in older people. This work would need to consider the high level of heterogeneity in the population, considering differences in sex, comorbidities and, most important, frailty.” – by Regina Schaffer

For more information:

Angus Forbes, PhD, can be reached at the Florence Nightingale Faculty of Nursing, Midwifery and Palliative Care, King’s College London, SE1 8WA, United Kingdom; email: angus.forbes@kcl.ac.uk.

Disclosures: King’s College London and Diabetes Frail funded this study. Forbes reports he has received honoraria or unrestricted educational grants from Abbott, Ascensia Diabetes Care, Eli Lilly and Roche. Please see the study for the other authors’ relevant financial disclosures.