Jump to content

HbA1c - A Poor Measure of Glucose Control in Healthy Folks


Recommended Posts

[Admin note: I've (obviously) shifted this conversation about HbA1c from its old home in the thread about best biomarkers to a new thread dedicated to the topic, since I think it is of general and lasting enough interest to deserve it's own thread.]





Mike Lustgarten, a regular on the CR Society Facebook Group prides himself on trying to optimize every biomarker based on any available literature from healthy agers and long-lived populations. I don't always agree with his assessment of optimal levels for certain biomarkers, but his blog is a good place to look for population-based evidence for optimal levels of various biomarkers, that is if you want to avoid doing the literature search on PubMed yourself...



Has Lustgarten mentioned his fasting glucose / post-prandial glucose / A1C% targets?  I've been curious what others are setting for a goal in these areas.  Last I checked, my A1C measured 4.9% on a low-carb diet (down from 5.3% on a higher carb diet), although I'm not sure if lower is continually better or if there's some optimal U-Curve there (beneficial hormesis?)

Link to comment
Share on other sites

<6% seems like a fairly large target -- With some rough math, I could spike my glucose to an average of 180 mg/dL 3 times a day for 2hrs each, then run an average of 100 mg/dL for the remainder of the day and stay within "normal" range.  Outside of a diabetic disease state, I can't think of anything I could eat to rise beyond that range.


Perhaps 4.2 - 4.6% is the optimal target?  For minimizing glycation and reduced its age-promoting effects at 4.6%, you would have to maintain a fairly tight ~80mg/dL fasting range with ~4hrs post-prandially averaging around 100 mg/dL -- I'm not too far off from this range, actually.  At one meal a day or very tight control, ~4.4% doesn't seem like an impossible target.

Link to comment
Share on other sites

[...]Perhaps 4.2 - 4.6% is the optimal target?  For minimizing glycation and reduced its age-promoting effects at 4.6%, you would have to maintain a fairly tight ~80mg/dL fasting range with ~4hrs post-prandially averaging around 100 mg/dL -- I'm not too far off from this range, actually.  At one meal a day or very tight control, ~4.4% doesn't seem like an impossible target.



Who are you?


Why hemoglobin A1c is not a reliable marker

Link to comment
Share on other sites


[...]Perhaps 4.2 - 4.6% is the optimal target?  For minimizing glycation and reduced its age-promoting effects at 4.6%, you would have to maintain a fairly tight ~80mg/dL fasting range with ~4hrs post-prandially averaging around 100 mg/dL -- I'm not too far off from this range, actually.  At one meal a day or very tight control, ~4.4% doesn't seem like an impossible target.



Who are you?


Why hemoglobin A1c is not a reliable marker


I was more active on here in the past... apologies for barging back in.


I'm looking at A1C as a surrogate for fasting and post-prandial glucose -- I often measure these both and my A1C tends to match up fairly closely if not spot on (I've worn CGM where my A1C measured 2% higher than predicted.)  Surely, Lustgarten with all his obsessive biometric tracking has some rough target and approach for reaching it?

Link to comment
Share on other sites

Well I didn't mean to question your identity, rather, meant to say, that the meaning of your A1c number is dependent on your particular profile (i.e. who you are physiologically)- so a CR'd person is going to have a different Ac1 number than an AL person, quite apart from any dietary practices, simply because the life cycle of your blood components is going to be different. I supplied the link that explains it. Sorry for any misunderstanding - your postings here speak for themselves, no need to question who you are :) 

Link to comment
Share on other sites

Sirtuin and TomB,


Thanks for prompting about this.


I've always thought HbA1c (or glycated hemoglobin, also sometimes simply A1c, HBA1c or HgA1c) is a pretty flaky measure of glucose control due to individual variations in red blood cell (RBC) turnover and other factors, but I've never really looked into it before. That study [1] that Chris Kresser points to appears to be part of the story. It found that red blood cell turnover was reduced in folks with healthy glucose control. As a result, red blood cells remain in circulation longer, and therefore have more time to get sugar molecules (glucose) attached to their hemoglobin molecules, which is (in simplest terms) what glycated hemoglobin (HbA1c) represents - the percentage of red blood cells that appear to have hemoglobin that is "gummed up" with glucose. So in healthy people with good glucose control, HbA1c may appear higher, making it look like their glucose control isn't as good as it actually is.


Interestingly and contrary to [1], study [2] did not find any difference between the average age of RBCs in diabetics vs normal people. Nonetheless, they found large differences in average RBC age within each population. In particularly, the average age of RBCs across their study population ranged from 38 to 60 days, independent of diabetes status. That is an average of 49 days ± 20%! Since the authors of [2] found the rate of glycation of RBCs was linear with exposure time, they were able to fit a linear model to it. Using there model, they determined that HbA1c could vary by ± 15% simply as a result of variation in how long RBCs persist on average in a person's bloodstream. So a person whose HB1Ac should be 5.2 (if they had an average RBC lifespan), might have a measured HbA1c as low as 4.4 (5.2 - (0.15 * 5.2)) if they had really fast RBC turnover or as high as 6.0 (5.2 + (0.15 * 5.2)) if their RBCs persisted in their bloodstream for much longer than average.


Such a large variability in measured HbA1c as a function of RBC turnover rate (e.g. 4.4 - 6.0) seems to seriously undermine the diagnostic utility of HbA1c as a measure of glucose control and (pre-)diabetes status, at least among healthy people, since the reference range for HbA1c is reported to only be 4.8 - 5.6!.


So what factors affect RBC turnover rate? I couldn't find any studies showing CR per se influences RBC lifespan. But low iron status / anemia is definitely associated with elevated HbA1c [3], and correcting the anemia through iron supplements (100 mg/day for 3 months) dramatically reversed their elevated HbA1c. Study [3] found that the fasting as post-meal glucose levels of anemic folks were no different from non-anemic controls and no different before vs. after their anemia was corrected. But being anemic nevertheless resulted in higher levels of glycated hemoglobin. Low iron status seems reasonably common about CR folks, particularly those who don't supplement iron and who don't eat a lot of heme-iron-rich meats and animal products.


Other factors influencing HbA1c include genetic variations in the hemoglobin molecule making it more prone to glycation [4] (see also PMID 20144281), and certain other pathologies / illnesses, including "haemoglobinopathies, splenomegaly, rheumatoid arthritis or as a side effect from drugs such as antiretrovirals and dapsone" [5]. Surgery or donating blood can also (obviously) impact red blood cell turnover. High dose of vitamin C can cause elevated blood glucose and HBa1C Even ethnicity and other demographics can influence HbA1c levels "Older non-diabetic subjects appear to have higher HbA1c values than younger individuals and differences in the HbA1c have also been found between individuals from different races, with Afro-Caribbeans and individuals of South Asian descent having values approximately 3 mmol/mol higher than Caucasians." [5].


In short, it looks like lots of factors other than glucose control can influence HbA1c level and influence it pretty dramatically. It therefore doesn't seem like a metric that is worth putting too much stock in.


Given all that cautionary background about HbA1c as a pretty worthless measure, there are a couple studies of the association between HbA1c and all-cause mortality [6][7] in people without (diagnosed) diabetes. First we have [6], which was a study of the NHANES folks once again. Here is the upshot of what they found in graphical form:



As you can see, these results are pretty consistent with HbA1c being a pretty worthless predictor of health or longevity, at least within a reasonable range. With the "normal" reference range (4.8 - 5.6) there was almost no association between HbA1c and mortality risk. HbA1c levels between 6 and 6.4 were associated with about a 25% increased risk of all-cause mortality. On the low end, below HbA1c of 4.5, you can see mortality risk rose pretty quickly. People with HbA1c down around 4 had about a 75% increased risk of mortality.


Another study [7] of 5000+ German adults, found a similar U-shaped association between HbA1c and mortality among people who weren't known diabetics:



Here is the data from [7] in tabular form:




As you can see, the all-cause mortality nadir in [7] was an HbA1c just below 5.5. Compared with a value from 5.0 - 5.7, people an HbA1c up to 6.4 weren't at a statistically significant increased risk of mortality. But they found more of a penalty for being below 5. Compared with an HbA1c between 5.0 and 5.7, folks who were less than 5.0 has a 70% increase in mortality risk. Eyeballing the graph, it looks like 5.0 was associated with a 20-25% increase, and 4.5 was associated with a 75% increase in mortality risk.


This additional apparently penalty on the low end of HbA1c in [7] relative to [6] remains unexplained, and seems to be a further demonstration of just how unreliable HbA1c is as a metric of health.


In short, HbA1c appears to be a pretty poor, highly variable measure of both glucose control and mortality risk, at least when it is within a reasonable range (~4.5 - 6.0). Direct glucose measurements, when fasting but particularly after meals, are much better metrics for assessing glucose control and diabetes risk.





[1] Diabetes Care. 2004 Apr;27(4):931-5.


Relationship between GHb concentration and erythrocyte survival determined from

breath carbon monoxide concentration.


Virtue MA(1), Furne JK, Nuttall FQ, Levitt MD.


Author information:

(1)Minneapolis Veterans Administration Medical Center, Minneapolis, Minnesota

55417, USA.


Comment in

Diabetes Care. 2004 Apr;27(4):1013-4.


OBJECTIVE: Subjects with decreased erythrocyte survival have an unusually low GHb

percentage. The goal of this study was to determine whether hyperglycemia, as

reflected by GHb percentage, is associated with decreased erythrocyte survival.

RESEARCH DESIGN AND METHODS: Erythrocyte survival was quantitated in 23 subjects

with type 2 diabetes, and these values were correlated with the subjects' GHb

percentage. Erythrocyte survival was determined from the difference between the

subjects' alveolar carbon monoxide (CO) concentration and atmospheric CO

concentration. Reticulocyte counts were obtained in 16 subjects.

RESULTS: Although the vast majority of the subjects had erythrocyte life spans

that fell within the normal range (123 +/- 23 days), there was a highly

significant inverse correlation (r = -0.66, P < 0.01) between life span and GHb

percentage, with an average decline in life span of 6.9 days for each 1% rise in

GHb. The reticulocyte count inversely correlated with erythrocyte life span (r =

-0.77, P < 0.01).

CONCLUSIONS: Hyperglycemia, as evidenced by high GHb percentage, is associated

with an appreciable decrease in erythrocyte life span. Because GHb appears to be

formed over the lifetime of the erythrocyte, this decreased erythrocyte survival

suggests that high GHb percentages may systematically underestimate the true

degree of hyperglycemia.


PMID: 15047651



[2] Blood. 2008 Nov 15;112(10):4284-91. doi: 10.1182/blood-2008-04-154112. Epub 2008

Aug 11.


Red cell life span heterogeneity in hematologically normal people is sufficient

to alter HbA1c.


Cohen RM(1), Franco RS, Khera PK, Smith EP, Lindsell CJ, Ciraolo PJ, Palascak MB,

Joiner CH.


Free full text: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2581997/


Although red blood cell (RBC) life span is a known determinant of percentage

hemoglobin A1c (HbA1c), its variation has been considered insufficient to affect

clinical decisions in hematologically normal persons. However, an unexplained

discordance between HbA1c and other measures of glycemic control can be observed

that could be, in part, the result of differences in RBC life span. To explore

the hypothesis that variation in RBC life span could alter measured HbA1c

sufficiently to explain some of this discordance, we determined RBC life span

using a biotin label in 6 people with diabetes and 6 nondiabetic controls. Mean

RBC age was calculated from the RBC survival curve for all circulating RBCs and

for labeled RBCs at multiple time points as they aged. In addition, HbA1c in

magnetically isolated labeled RBCs and in isolated transferrin

receptor-positivereticulocytes was used to determine the in vivo synthetic rate

of HbA1c. The mean age of circulating RBCs ranged from 39 to 56 days in diabetic

subjects and 38 to 60 days in nondiabetic controls. HbA1c synthesis was linear

and correlated with mean whole blood HbA1c (R(2) = 0.91). The observed variation

in RBC survival was large enough to cause clinically important differences in

HbA1c for a given mean blood glucose.


DOI: 10.1182/blood-2008-04-154112

PMCID: PMC2581997

PMID: 18694998



[3] Acta Haematol. 2004;112(3):126-8.


Effect of iron deficiency anemia on the levels of hemoglobin A1c in nondiabetic



Coban E(1), Ozdogan M, Timuragaoglu A.


Full text: http://sci-hub.cc/10.1159/000079722


The major form of glycohemoglobin is hemoglobin A1c (HbA1c). The HbA1c fraction

is abnormally elevated in chronic hyperglycemic diabetic patients and correlates

positively with glycemic control. Previous studies suggest that iron deficiency

anemia (IDA) affects the levels of HbA1c. The aim of this study was to determine

the effect of IDA on HbA1c levels in nondiabetic patients. The population studied

consisted of 50 patients (30 women, 20 men, mean age 35.7 +/- 11.9 years) with

IDA and 50 healthy subjects that were matched for age and sex. Patients who had

glucose tolerance abnormalities (impaired glucose tolerance or diabetes

mellitus), hemoglobinopathies, hemolytic anemia, chronic alcohol ingestion and

chronic renal failure were excluded from the study. Hematologic investigations,

fasting and postprandial glucose and HbA1c levels were measured in all subjects

before iron therapy. All patients with IDA were treated with iron 100 mg/day for

3 months. We repeated the laboratory investigation after iron therapy. Before

iron treatment, the mean HbA1c (7.4 +/- 0.8%) level in patients with IDA was

higher than in a healthy group (5.9% +/- 0.5) (p < 0.001). In patients with IDA,

HbA1c decreased significantly after iron treatment from a mean of 7.4% +/- 0.8 to

6.2% +/- 0.6 (p < 0.001). Iron deficiency must be corrected before any diagnostic

or therapeutic decision is made based on HbA1c.


DOI: 10.1159/000079722

PMID: 15345893



[4] J Investig Med High Impact Case Rep. 2016 Jan 28;4(1):2324709616628549. doi:



Led Astray by Hemoglobin A1c: A Case of Misdiagnosis of Diabetes by Falsely

Elevated Hemoglobin A1c.


Chen J(1), Diesburg-Stanwood A(2), Bodor G(2), Rasouli N(1).


Author information:

(1)University of Colorado School of Medicine, Aurora, CO, USA; VA Eastern

Colorado Health Care Systems, Denver, CO, USA. (2)VA Eastern Colorado Health Care

Systems, Denver, CO, USA.


Hemoglobin A1c (A1c) is used frequently to diagnose and treat diabetes mellitus.

Therefore, it is important be aware of factors that may interfere with the

accuracy of A1c measurements. This is a case of a rare hemoglobin variant that

falsely elevated a nondiabetic patient's A1c level and led to a misdiagnosis of

diabetes. A 67-year-old male presented to endocrine clinic for further management

after he was diagnosed with diabetes based on an elevated A1c of 10.7%, which is

approximately equivalent to an average blood glucose of 260 mg/dL. Multiple

repeat A1c levels remained >10%, but his home fasting and random glucose

monitoring ranged from 92 to 130 mg/dL. Hemoglobin electrophoresis and subsequent

genetic analysis diagnosed the patient with hemoglobin Wayne, a rare hemoglobin

variant. This variant falsely elevates A1c levels when A1c is measured using

cation-exchange high-performance liquid chromatography. When the boronate

affinity method was applied instead, the patient's A1c level was actually 4.7%.

Though hemoglobin Wayne is clinically silent, this patient was erroneously

diagnosed with diabetes and started on an antiglycemic medication. Due to this

misdiagnosis, the patient was at risk of escalation in his "diabetes management"

and hypoglycemia. Therefore, it is important that providers are aware of factors

that may result in hemoglobin A1c inaccuracy including hemoglobin variants.


DOI: 10.1177/2324709616628549

PMCID: PMC4735504

PMID: 26848480



[5] JRSM Open. 2016 Jan 18;7(2):2054270415619321. doi: 10.1177/2054270415619321.

eCollection 2016.


Rising HbA1c in the presence of optimal glycaemic control as assessed by

self-monitoring - iron deficiency anaemia.


Mudenha ET(1), Aarella VG(1), Chandrasekaram S(1), Fernando DJ(1).


Free full text: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4740944/


HbA1c can be affected by determinants other than glucose and an awareness of this

is important to avoid unnecessary hypoglycaemia.


DOI: 10.1177/2054270415619321

PMCID: PMC4740944

PMID: 26877879



[6] Circ Cardiovasc Qual Outcomes. 2010 Nov;3(6):661-7. doi:

10.1161/CIRCOUTCOMES.110.957936. Epub 2010 Oct 5.


Low hemoglobin A1c and risk of all-cause mortality among US adults without



Carson AP(1), Fox CS, McGuire DK, Levitan EB, Laclaustra M, Mann DM, Muntner P.


Free full text: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4734630/


BACKGROUND: Among individuals without diabetes, elevated hemoglobin A1c (HbA1c)

has been associated with increased morbidity and mortality, but the literature is

sparse regarding the prognostic importance of low HbA1c.

METHODS AND RESULTS: National Health and Nutrition Examination Survey III (NHANES

III) participants, 20 years and older, were followed up to 12 years (median

follow-up, 8.8 years) for all-cause mortality. Cox proportional hazards

regression was used to calculate hazard ratios (HR) and 95% confidence intervals

(CI) for the association between HbA1c levels and all-cause mortality for 14 099

participants without diabetes. There were 1825 deaths during the follow-up

period. Participants with a low HbA1c (<4.0%) had the highest levels of mean red

blood cell volume, ferritin, and liver enzymes and the lowest levels of mean

total cholesterol and diastolic blood pressure compared with their counterparts

with HbA1c levels between 4.0% and 6.4%. An HbA1c <4.0% versus 5.0% to 5.4% was

associated with an increased risk of all-cause mortality (HR, 3.73; 95% CI, 1.45

to 9.63) after adjustment for age, race-ethnicity, and sex. This association was

attenuated but remained statistically significant after further multivariable

adjustment for lifestyle, cardiovascular factors, metabolic factors, red blood

cell indices, iron storage indices, and liver function indices (HR, 2.90; 95% CI,

1.25 to 6.76).

CONCLUSIONS: In this nationally representative cohort, low HbA1c was associated

with increased all-cause mortality among US adults without diabetes. Additional

research is needed to confirm these results and identify potential mechanisms

that may be underlying this association.


DOI: 10.1161/CIRCOUTCOMES.110.957936

PMCID: PMC4734630

PMID: 20923991



[7] Diabetes Care. 2015 Feb;38(2):249-56. doi: 10.2337/dc14-1787. Epub 2014 Nov 20.


Association between hemoglobin A1c and all-cause mortality: results of the

mortality follow-up of the German National Health Interview and Examination

Survey 1998.


Paprott R(1), Schaffrath Rosario A(1), Busch MA(1), Du Y(1), Thiele S(2),

Scheidt-Nave C(1), Heidemann C(3).


Author information:

(1)Department of Epidemiology and Health Monitoring, Robert Koch Institute,

Berlin, Germany. (2)Department of Food Economics and Consumption Studies,

Christian-Albrechts-University Kiel, Kiel, Germany. (3)Department of Epidemiology

and Health Monitoring, Robert Koch Institute, Berlin, Germany heidemannc@rki.de.


OBJECTIVE: This study examined the association of HbA1c-defined glycemic status

and continuous HbA1c with all-cause mortality.

RESEARCH DESIGN AND METHODS: The study population comprised 6,299 participants

(aged 18-79 years) of the German National Health Interview and Examination Survey

1998, who were followed up for mortality for an average of 11.6 years. Glycemic

status was defined as known diabetes (self-reported diagnosis or intake of

antidiabetic medication) and based on HbA1c levels according to American Diabetes

Association diagnostic criteria as undiagnosed diabetes (≥6.5% [≥48 mmol/mol]),

prediabetes with very high (6.0-6.4% [42-46 mmol/mol]) or high diabetes risk

(5.7-5.9% [39-41 mmol/mol]), and normoglycemia (<5.7% [<39 mmol/mol]).

Associations between glycemic status and mortality were examined by Cox

regression adjusting for age, sex, education, lifestyle factors, anthropometric

measures, and history of chronic diseases (reference: normoglycemia). Spline

models were fitted to investigate associations between continuous HbA1c and

mortality among participants without known diabetes.

RESULTS: Excess mortality risk was observed for participants with known diabetes

(hazard ratio 1.41 [95% CI 1.08-1.84]) and undiagnosed diabetes (1.63

[1.23-2.17]) but not for those with high (1.02 [0.80-1.30]) or very high diabetes

risk (0.87 [0.67-1.13]). Spline models revealed a U-shaped association, with

lowest risk at HbA1c levels 5.4-5.6% (36-38 mmol/mol) and a significantly

increased risk at ≤5.0% (≤31 mmol/mol) and ≥6.4% (≥46 mmol/mol).

CONCLUSIONS: Unlike known and undiagnosed diabetes, HbA1c levels in the

prediabetic range were not associated with an increased mortality risk. The

observed U-shaped relationship adds to existing evidence that not only high but

also low HbA1c levels might be associated with all-cause mortality.


© 2015 by the American Diabetes Association. Readers may use this article as long

as the work is properly cited, the use is educational and not for profit, and the

work is not altered.


DOI: 10.2337/dc14-1787

PMID: 25414153

Link to comment
Share on other sites

Well, "worthless" depends on what is measured. Precision when it comes to mortality, is one thing. But "useful" can be had in other contexts. For example:




Risk factors for progression of brain atrophy in aging: six-year follow-up of normal subjects. (link to full text)



Neurologically asymptomatic elderly experience continuing brain volume loss, which appears to accelerate with age. Glycated hemoglobin A (HbA1c) was identified as a risk factor for a greater rate of brain atrophy. [...]


What is particularly striking, is how there is steady and pretty dramatic increase in brain volume loss in asymptotic individuals with each increasing fraction of A1c starting with a huge increase in loss from 5.2 to 5.3 (there are several interesting charts in that full text link). I think the upshot is that you want to keep your A1c lower just to slow the loss of brain volume, which happens to everyone with aging, no matter how healthy - of course, having additional risk factors (like metabolic syndrome associated ones) makes things that much worse.


Then there is this:


PMID: 24675382


Does cancer risk increase with HbA1c, independent of diabetes? (full text link)



The data reveal that chronic hyperglycaemia correlates with increased cancer risk for a number of cancers, except prostate cancer. Evidence is also provided that risk is already increased in the pre-diabetic and normal ranges for several cancers.



These results merit urgent investigation into the risks and advantages of updating recommendations for stricter glycaemic control in diabetic and non-diabetic subjects, as this could help reduce the risk of cancer incidence and mortality.


There are several interesting charts in the full text link, but the upshot is that A1c is something to watch.





Link to comment
Share on other sites

  • 1 month later...

I guess that Dean is going to come back at me with pooh pooh arguments, and the subjects were diabetics, but it did seem to me that a lot of covariates were included in the analyses for the two below papers suggesting Hemoglobin A1c is an important metric.

Hemoglobin A1C testing and amputation rates in black, Hispanic, and white Medicare patients.
Suckow BD, Newhall KA, Bekelis K, Faerber AE, Gottlieb DJ, Skinner JS, Stone DH, Goodney PP.
Ann Vasc Surg. 2016 Jul 26. pii: S0890-5096(16)30509-X. doi: 10.1016/j.avsg.2016.03.035. [Epub ahead of print]
PMID:  27474195



Major (above or below-knee) amputation is a complication of diabetes, and is seen more common among black and Hispanic patients. While amputation rates have declined for patients with diabetes in the last decade, it remains unknown if these improvements have equitably extended across racial groups, and if measures of diabetic care, such as hemoglobin A1C testing, are associated with these improvements. We set out to characterize secular changes in amputation rates among black, Hispanic, and white patients, and to determine associations between hemoglobin A1c testing and amputation risk.


We identified 11,942,840 Medicare patients (55% female) with diabetes over the age of 65 between 2002-2012, and followed them for a mean of 6.6 years. Of these, 86% were white, 11.5% were black and 2.5% were Hispanic. We recorded the occurrence of major amputation and hemoglobin A1C testing during this time period, and studied secular changes in amputation rate by race (black, Hispanic, and white). Finally, we examined associations between amputation risk and hemoglobin A1C testing. We measured both the presence of any testing, as well as testing consistency using three categories: poor consistency (hemoglobin A1C testing in 0-50% of years), medium consistency (testing in 50-90% of years), and high consistency (testing in >90% of the years in the cohort).


Between 2002 and 2012, the average major lower extremity amputation rate in diabetic Medicare patients was 1.78 per 1,000 per year for black patients, 1.15 per 1,000 per year for Hispanic patients and 0.56 per 1,000 per year for white patients (p<0.001). Over the study period, the incidence of major amputation in Medicare patients with diabetes declined by 54%, from 1.15 per 1,000 in 2002 to 0.53 per 1,000 in 2012 (rate ratio = 0.53, 95% CI =0.51-0.54). The reduction in amputation rate was similar across racial groups: 52% for black patients, 61% for Hispanic patients, and 55% for white patients. In multivariable analysis adjusting for patient characteristics, including race, any use of hemoglobin A1C testing was associated with a 15% decline in amputation risk (HR 0.85, 95% CI 0.83-0.87, p<0.001). High consistency hemoglobin A1C testing was associated with a 39% decline in amputation (HR 0.61, 95% CI 0.59-0.62, p<0.0001).


Although more frequent among racial minorities, major lower extremity amputation rates have declined similarly across black, Hispanic, and white patients over the last decade. Hemoglobin A1C testing, particularly the consistency of testing over time, may be an effective component metric of longitudinal quality measures towards limiting amputation in all races.

Consistency of Hemoglobin A1c Testing and Cardiovascular Outcomes in Medicare Patients With Diabetes
  Philip P. Goodney, Karina A. Newhall, Kimon Bekelis, Daniel Gottlieb, Richard Comi, Sushela Chaudrain, Adrienne E. Faerber, Todd A. Mackenzie, and Jonathan S. Skinner
  J Am Heart Assoc. 2016;5:e003566. Originally published August 10, 2016. doi: 10.1161/JAHA.116.003566


Background Annual hemoglobin A1c testing is recommended for patients with diabetes mellitus. However, it is unknown how consistently patients with diabetes mellitus receive hemoglobin A1c testing over time, or whether testing consistency is associated with adverse cardiovascular outcomes.

Methods and Results We identified 1 574 415 Medicare patients (2002–2012) with diabetes mellitus over the age of 65. We followed each patient for a minimum of 3 years to determine their consistency in hemoglobin A1C testing, using 3 categories: low (testing in 0 or 1 of 3 years), medium (testing in 2 of 3 years), and high (testing in all 3 years). In unweighted and inverse propensity‐weighted cohorts, we examined associations between testing consistency and major adverse cardiovascular events, defined as death, myocardial infarction, stroke, amputation, or the need for leg revascularization. Overall, 70.2% of patients received high‐consistency testing, 17.6% of patients received medium‐consistency testing, and 12.2% of patients received low‐consistency testing. When compared to high‐consistency testing, low‐consistency testing was associated with a higher risk of adverse cardiovascular events or death in unweighted analyses (hazard ratio

=1.21; 95% CI, 1.20–1.23; P<0.001), inverse propensity‐weighted analyses (HR=1.16; 95% CI, 1.15–1.17; P<0.001), and weighted analyses limited to patients who had at least 4 physician visits annually (HR=1.15; 95% CI, 1.15–1.16; P<0.001). Less‐consistent testing was associated with worse results for each cardiovascular outcome and in analyses using all years as the exposure.

Conclusions Consistent annual hemoglobin A1c testing is associated with fewer adverse cardiovascular outcomes in this observational cohort of Medicare patients of diabetes mellitus.

Key Words:

    cardiovascular outcomes
    diabetes mellitus
    health disparities
    health outcomes
    hemoglobin A1c
Link to comment
Share on other sites



I guess that Dean is going to come back at me with pooh pooh arguments...


Au contraire Al. I love Winnie the Pooh. Two of favorite Pooh quotes (by author AA Milne):


I am a Bear of Very Little Brain, and long words Bother me.


When you are a Bear of Very Little Brain, and you Think of Things, you find sometimes that a Thing which seemed very Thingish inside you is quite different when it gets out into the open and has other people looking at it.

It's very reminiscent of one of my other favorite sentiments:


I don't know what I think until I hear what I say and see what I do.


Speaking of long words, did you notice the long word your two papers (PMID: 27474195 and doi:10.1161/JAHA.116.003566) have in common? "Consistency" - that is, consistency in HBA1c testing. Nowhere in either of them do they talk about results of the HBA1c testing, to say nothing of whether or not the diabetic patients so tested received any different treatment, to say nothing of whether or not such treatment was effective at bringing down the patients HBA1c, or fasting/postprandial glucose.


So it is quite plausible (in fact, seems to me likely) that consistent (i.e. yearly) HBA1c testing is a proxy for both access to quality medical care, and in particular a caring, knowledgeable doctor and most of all, conscientiousness on the part of the diabetic patient, all three of which would expected to be inversely correlated with the kind of neglect that eventually results to the amputation of a limb, independent of HBA1c level. The authors of your first paper seem to agree in the full text:


While our study adjusts for many medical and surgical aspects of care, we readily recognize that vascular, podiatric and wound care, lipid testing and statin therapy, blood pressure control, and the management of obesity are all other preventive measures that need optimal approaches to ensure the best results. Hemoglobin A1C testing may simply represent a proxy for access to these services, and future efforts to better understand the role of preventive measures in limiting amputation certainly are necessary. 


The authors of your second paper, which found reduced CVD mortality and heart attacks in diabetics who received consistent HBA1c test, support the same hypothesis - namely that consistent HBA1c testing is a proxy for other things, particularly patient engagement with their healthcare provider, which is really the cause of the observed benefits. From the full text:


A theoretical framework proposing an explanatory mechanism has been described by Presseau et al.,29 wherein they hypothesized that highly consistent hemoglobin A1c testing is a derivative of the consistent patient and physician interactions that occur in the setting of clinical trials. Our results are consistent with this view. Whereas Presseau et al. strike a cautionary note—that raising consistency alone will not necessarily improve the quality of patient‐physician interactions—testing consistency allows a determination of which patients were most commonly engaged with their healthcare providers and are thereby likely to achieve stronger relationships with their healthcare team over time. These stronger relationships could potentially manifest in better outcomes, an increasingly common theme in cardiovascular care.


Note - neither I nor the authors, are denying that seriously elevated HBA1c in diabetics is a marker for poor glucose control. And neither am I denying that seriously elevated HBA1c in diabetics is probably correlated with adverse consequences. But the two studies you post are terribly weak and indirect evidence to support these points, if that is what you meant them to be.


What these two studies basically say, is "be the kind of person who is fortunate enough to have health insurance and who is conscientious enough to visit your doctor regularly and gets tests done, especially if you've got a serious illness like diabetes".



Link to comment
Share on other sites

  • 8 months later...

Intimately related to this thread: one way to get a guess as to one's RBC turnover rate is one's bilirubin level. Bilirubin (to quote Wikipedia) "is a yellow compound that occurs in the normal catabolic pathway that breaks down heme in vertebrates. This catabolism is a necessary process in the body's clearance of waste products that arise from the destruction of aged red blood cells." Since slower turnover of RBC — which, as Dean has just laid out, is both mechanistically expected and empirically observed to result in an "artificially" high HbA1c — will lead to lower bilirubin, you would expect that low bilirubin would also be linked to artificially high HbA1c.
Dean has apparently forgotten this, or didn't make the connection, but years ago, on the original CR Listserv (and in a thread that, was originally about and titled "Dean Pomerleau's Blood Test Results", including a surprising-seeming high HbA1c), poster "M P" alerted us to the fact "that high levels of bilirubin ("hyperbilirunemia") can also interfere with some A1C testing methods (assays) resulting in inaccurately high reported A1C." Dean himself then dug up (1), which found that in a group of apparently healthy men and women with no history of CVD or taking antidiabetic, antihypertensive, or anticholesterol medication, "The standardized regression coefficient of HbA1c for TB [total bilirubin] and that of TB for HbA1c was -0.12 (P=0.007) and -0.06 (P=0.02), respectively, in the [male] smokers, -0.20 (P<0.0001) and -0.07 (P<0.0001), respectively, in the [male] nonsmokers, and -0.21 (P<0.0001) and -0.14 (P<0.0001), respectively, in the women. The odds ratio of 1 SD increment in TB for HbA1c ≥5.4% was not significant in the smokers, 0.67 (P=0.002) in the nonsmokers, and 0.55 (P<0.0001) in the women."
G— and JWP also contributed to this discussion, but I unfortunatley seem to have deleted their posts. See also here, here, here, here, here, and here (at least in the diabetic groups — their reporting on the normals is sufficiently convoluted that I don't want to say without further reading that such can actually be parsed out of their report).


Since many people (myself included) report very low bilirubin, it's likely that many of us may have somewhat inflated HbA1c.
[1] Circ J. 2011;75(1):190-5. Epub 2010 Dec 2.
Bilirubin is negatively associated with hemoglobin a(1c) independently of other cardiovascular risk factors in apparently healthy Japanese men and women.
Oda E, Kawai R.
PMID: 21139250

Link to comment
Share on other sites

  • 2 months later...
  • 9 months later...

Interesting analysis guys.

I never realized that people on CR generally have low bilirubin. I remember when I first started CR mine was quite high, and since then it has lowered a little. Last time it was measured at 15 (ref 1-22). My HbA1c was measured at 4.8%.

Link to comment
Share on other sites


This topic is now archived and is closed to further replies.

  • Create New...