[IP] In Diabetes, Get Glucose Control 'Just Right'
In Diabetes, Get Glucose Control 'Just Right'
By Crystal Phend, Senior Staff Writer, MedPage Today
Published: January 27, 2010
Reviewed by Dori F. Zaleznik, MD; Associate Clinical Professor of
Medicine, Harvard Medical School, Boston and
Dorothy Caputo, MA, RN, BC-ADM, CDE, Nurse Planner
Earn CME/CE credit
for reading medical news
Intensive treatment to achieve normoglycemia in type 2 diabetes may be
almost as bad for survival and heart health as leaving glucose levels
elevated, according to a study of real-world practice.
The study turned up a U-shaped link between all-cause mortality and
glycosylated hemoglobin levels in which those at a median of 7.5%
carried the lowest risk, found Craig J. Currie, PhD, of Cardiff
University in Cardiff, Wales, and colleagues.
By comparison, patients who reached normal glycosylated hemoglobin
levels with a median of 6.4% were at 52% greater risk of dying from
any cause during the study period (95% confidence interval 32% to 76%).
Those in the highest decile with a median of 10.5% hemoglobin A1c were
at 79% elevated risk (95% CI 55% to 106%), the researchers reported
online in The Lancet.Action Points
Explain to interested patients that the American Diabetes Association
recommends a hemoglobin A1c treatment target of less than 7% for
nonpregnant adults in general with tight glucose control reserved for
patients in whom it can be achieved without significant hypoglycemia
or other adverse effects of treatment.
"If confirmed, diabetes guidelines might need revision to include a
minimum hemoglobin A1c value," they wrote in the study.
Richard Bergenstal, MD, president for medicine and science of the
American Diabetes Association, agreed that confirmation would be key
and cautioned against overinterpreting the results.
Practice doesn't change based on epidemiological evidence alone,
particularly when it fits with only part of the clinical trial
literature, he noted.
Among the several recent large clinical trials to examine tight
glucose control with hemoglobin A1c goals below the standard 7%, only
one showed a negative impact on survival -- 22% excess mortality in
the ACCORD trialwith an ultra-tight target under 6.0%.
A meta-analysis of the trials suggested no reduction in all-cause or
cardiovascular mortality, stroke, amputations, or even microvascular
complications with aggressive glycemic control strategies aiming to
bring A1c under 7%.
Although the reason behind the elevated mortality in ACCORD remains a
mystery, "the most plausible explanation for these results is
hypoglycemia: the treatment target was probably too low, or glucose
lowering was too rapid, or the combinations of treatments led to
hypoglycemia," according to an accompanying editorial in The Lancet.
The new observational data may shed some light on this issue, wrote
Beverley Balkau, PhD, and Dominique Simon, MD, PhD, both of the INSERM
Center for Research in Epidemiology and Population Health in
Currie's group retrospectively reviewed data from medical records of
British primary care physicians in the U.K. General Practice Research
Database from November 1986 to November 2008.
They identified two cohorts of patients 50 years and older with
primary type 2 diabetes: 27,965 whose treatment had been intensified
from oral monotherapy to combinations of oral blood-glucose lowering
agents, and 20,005 who switched to regimens that included insulin.
Initial hemoglobin A1c levels were 9.0% and 10.0% in the two groups
before intensification of treatment.
"This two-cohort approach was intended to establish whether any
emergent patterns were independent of diabetes treatment regimen," the
researchers explained in The Lancet.
After an average of 4.5 to 5.2 years of follow-up, patients in the
lowest and highest deciles of achieved hemoglobin A1c were at elevated
risk of death from any cause in both unadjusted and adjusted analyses.
In the cohort treated with oral drugs alone, only the highest and
lowest deciles, for whom the median hemoglobin levels were 10.5% and
6.4% respectively, were at elevated risk.
In the group treated with insulin, significantly elevated risk was
seen in the three lowest and two highest groups compared with the
lowest risk decile (median 7.5% hemoglobin A1c).
The researchers also found the same U-shaped relationship between
glucose control and new-onset, large-vessel cardiac disease events.
The lowest hemoglobin A1c group was at 54% increased risk for cardiac
events (28% to 84%) while the highest decile was at 36% elevated risk
(95% CI 14% to 61%) compared with the group who had a median 7.5%
One of the risk factors for these hazards appeared to be insulin-based
A sensitivity analysis excluding patients with high cardiovascular
risk revealed 46% higher mortality risk (95% CI 34% to 59%) with
insulin-based therapy than with oral combination therapy.
Insulin treatment was also associated with 31% elevated likelihood of
progression to a first large-vessel disease event (95% CI 22% to 42%).
Since previous studies have shown greater hypoglycemia risk with
insulin treatment than with oral agents, such as sulfonylurea therapy,
this result further implicates hypoglycemia in the premature death
associated with tight glucose control, Balkau and Simon said.
"Priority should be given to insulin sensitisers for as long as
possible in patients with type 2 diabetes, because these drugs allow a
low hemoglobin A1c to be targeted without any risk of hypoglycemia,"
they wrote in the editorial.
However, the researchers noted that another possible explanation is
that the insulin-treated patients were older with more comorbidities
and longer disease duration.
They also cautioned that residual sources of confounding were
possible, such as different prescribing patterns for cardiovascular
prophylaxis across the hemoglobin A1c groups.
Other limitations of the study included missing data, different timing
and methodologies for measuring hemoglobin A1c, and inclusion of
patients who could be assigned to both cohorts.
Overall, these results should be a reminder that the "lower is better"
paradigm has given way to individualized treatment targets, Bergenstal
"We need to keep trying to understand the data so we get the right
patients into the right A1c targets," he toldMedPage Today.
The study was funded by Eli Lilly and Company.
Currie reported conflicts of interest with Astellas, Diabetes UK, the
European Association for the Study of Diabetes, the Engineering and
Physical Sciences Research Council, Ferring, GSK, Lilly, Medtronic,
the Medical Research Council, Pfizer, sanofi-aventis, the National
Health Service, Wyeth, Amylin, Aryx, Boeringher Ingelheim, Bristol-
Myers Squibb, Eisel, Ferring, Ipsen, Merck, and Takeda.
Several co-authors reported being employed by Eli Lilly and Company.
Other co-authors reported financial conflicts of interest with Abbott,
Allergan, BMS, GSK, Lilly, Novartis, Novo Nordisk, MSD, Roche, sanofi-
aventis, Takeda, Astellas, Ferring, Medtronic, and Wyeth (Pfizer).
Balkau reported having served as a speaker for sanofi-aventis and on
advisory panels for AstraZeneca, Bristol Myers Squibb, Lilly, and
sanofi-aventis. Simon reported having served as a speaker for
GlaxoSmith Kline, sanofi-aventis, Servier, and on advisory panels for
AstraZeneca, Bristol Myers Squibb, GlaxoSmith Kline, and Novartis.
Bergenstal reported receiving research funding and serving on advisory
boards for various pharmaceutical companies related to novel diabetes
treatments but without any related personal compensation.
for HELP or to subscribe/unsubscribe/change list versions,