India carries one of the largest diabetes burdens in the world. The national ICMR-INDIAB survey of 113 043 adults estimated a weighted prevalence of 11.4% for diabetes and 15.3% for prediabetes, with the epidemic still rising in most states¹. Because a substantial share of this burden remains undiagnosed, the accuracy of the screening test matters at population scale: a small systematic bias in a widely used test translates into large absolute numbers of people wrongly labelled or wrongly reassured.

Glycated haemoglobin (HbA1c) has become central to that task. In 2009 an International Expert Committee recommended an HbA1c of 6.5% or higher for diagnosing diabetes², and the American Diabetes Association continues to endorse HbA1c of at least 6.5% for diabetes and 5.7–6.4% for prediabetes³. The test needs no fasting, has low within-person variability and relates closely to microvascular risk. Its biological premise, however, is narrow. HbA1c reflects the glucose exposure of haemoglobin integrated over the lifespan of the circulating erythrocyte, so any condition that alters erythrocyte age, haemoglobin structure or assay behaviour can shift HbA1c without any change in glycaemia⁴. Even among haematologically normal people, variation in mean erythrocyte age is large enough to change clinical decisions at a given mean glucose⁵. The ADA therefore advises that diabetes be diagnosed with glucose criteria when red-cell turnover is abnormal³.

The analytical basis of the test explains why erythrocyte biology matters. HbA1c is formed by the slow, non-enzymatic attachment of glucose to the N-terminal valine of the haemoglobin β-chain, which then rearranges into a stable ketoamine. The fraction of haemoglobin glycated at any moment is therefore a function of two quantities: the prevailing glucose concentration and the time for which each haemoglobin molecule has been exposed to it. Clinical laboratories measure this fraction by ion-exchange high-performance liquid chromatography, boronate affinity chromatography, capillary electrophoresis or immunoassay. Immunoassays, which predominate in many Indian hospital laboratories because of their low cost and ease of automation, detect the glycated N-terminal epitope with an antibody and express the result as a ratio to total haemoglobin measured photometrically in the same sample. Each method responds differently to haemoglobin variants and to changes in the composition of the red-cell population, and the systematic literature on anaemia and HbA1c shows wide method-dependent variation^4,6.

Anaemia is the most prevalent of these conditions. The Global Burden of Disease Study 2021 estimated that 24.3% of the world population, or 1.92 billion people, were anaemic, with dietary iron deficiency the leading cause and women and South Asian populations disproportionately affected⁷. In India, the fifth National Family Health Survey documented a rising prevalence of anaemia among women of reproductive age, clustered geographically with undernutrition⁸. The co-existence of a high diabetes burden and a high anaemia burden means that any non-glycaemic distortion of HbA1c by iron-deficient erythropoiesis could cause misclassification on a large scale.

The direction of that distortion is contested. Brooks and colleagues first reported raised glycated haemoglobin in iron deficiency that fell with iron repletion⁹. Coban and colleagues found a mean HbA1c of 7.4% in non-diabetic patients with iron deficiency anaemia compared with 5.9% in controls, falling to 6.2% after three months of iron¹⁰, and a study from Chennai reported 7.6% versus 5.5%¹¹. Iron treatment lowered HbA1c in children with type 1 diabetes despite unchanged glycaemia¹². Other studies point the other way. Gram-Hansen and colleagues found no baseline difference¹³. Sinha and colleagues in Delhi found a lower HbA1c in iron deficiency anaemia (4.6% vs 5.5%) that rose after treatment¹⁴. In the United States National Health and Nutrition Examination Survey (NHANES), Ford and colleagues observed a positive gradient of HbA1c with haemoglobin across the population¹⁵, whereas Kim and colleagues found that iron deficiency in women shifted HbA1c from below to above 5.5%¹⁶. In young adults from Pune, HbA1c identified prediabetes three times as often as the oral glucose tolerance test, and lower haemoglobin independently predicted HbA1c¹⁷.

Evidence syntheses have not settled the question. A systematic review concluded that iron deficiency tends to raise HbA1c while other anaemias may lower it, and identified the degree of anaemia as a key unanswered question⁶. A meta-analysis restricted to standardised assays found no significant difference¹⁸. A 2026 meta-analysis of 30 studies estimated HbA1c to be 0.66% higher in iron deficiency anaemia but attributed most heterogeneity to assay method and anaemia definitions¹⁹, and a contemporaneous review in non-diabetic adults found no conclusive overall difference and inconsistent relationships with anaemia severity²⁰. Three gaps follow. First, many studies did not show that glycaemia was comparable between groups, so a difference in HbA1c could reflect a difference in glucose. Second, the role of anaemia severity has rarely been examined within a single sample. Third, Indian data obtained with immunoassay methods, which are widely used in district and teaching hospitals, remain sparse.

We hypothesised that non-diabetic individuals with microcytic hypochromic anaemia would have a higher HbA1c than non-anaemic individuals with comparable fasting and postprandial glucose.

OBJECTIVES

The primary objective was to compare HbA1c between non-diabetic individuals with microcytic hypochromic anaemia and non-anaemic controls. The secondary objectives were (i) to compare absolute HbA1c, the mass of glycated haemoglobin per unit volume of blood, between the groups; (ii) to examine the association of HbA1c with haemoglobin and red-cell indices; and (iii) to explore whether the difference in HbA1c varied by anaemia severity and by sex.

Study design and setting We conducted a hospital-based case–control study in the Department of Internal Medicine, , Raichur institute of medical sciences (Rims), Raichur, Karnataka, India. over a period of six months .The study is reported according to the STROBE statement for case–control studies21. Participants Participants were drawn from the medical outpatient department and inpatient wards of the hospital. Cases were non-diabetic patients with microcytic hypochromic anaemia, defined as haemoglobin below 13 g/dL in men or below 12 g/dL in women, mean corpuscular volume (MCV) below 76 fL, mean corpuscular haemoglobin below 27 pg and mean corpuscular haemoglobin concentration below 33 g/dL, with a microcytic hypochromic picture on peripheral smear. Controls were non-anaemic individuals with normal red-cell indices and a normal peripheral smear, recruited from the same departments and selected to be comparable to cases in age and sex .Fifty cases and 50 controls were enrolled Individuals were excluded if they had known diabetes or were taking glucose-lowering medication, or had symptoms of diabetes with fasting plasma glucose above 126 mg/dL or postprandial plasma glucose above 200 mg/dL. Further exclusions were acute blood loss, haemolytic anaemia, known haemoglobinopathy or recent surgery; kidney disease, liver disease or another major systemic illness; pregnancy (a urine pregnancy test was performed in women with amenorrhoea); an endocrinopathy affecting glucose tolerance; blood transfusion in the preceding two months; and blood urea above 20 mg/dL or serum creatinine above 1.2 mg/dL in men or 0.9 mg/dL in women. Operational definitions and outcomes Non-diabetic status was defined by the absence of a history of diabetes or glucose-lowering treatment together with plasma glucose below the exclusion thresholds above. The exposure was microcytic hypochromic anaemia as defined for cases. The primary outcome was HbA1c expressed as a percentage of total haemoglobin. Secondary outcomes were absolute HbA1c and the proportion of participants with HbA1c in the ADA prediabetes range of 5.7% or above3. Fasting and postprandial plasma glucose were treated as measures of prevailing glycaemia, so that a difference in HbA1c between groups with comparable glucose could be attributed to non-glycaemic factors. Age, sex, BMI and fasting glucose were treated as covariates; occupation, diet and menorrhagia were recorded to describe the groups and to examine the comparability of cases and controls. Sample size No formal sample size calculation was performed; the sample of 50 cases and 50 controls was determined by the number of eligible patients who could be enrolled during the study period. We therefore report effect estimates with 95% confidence intervals to convey precision. Data collection and variables A structured proforma recorded age, sex, occupation (student, homemaker or employed), diet (vegetarian or non-vegetarian) and, in women, a history of menorrhagia. Height and weight were measured and body mass index (BMI) was calculated as weight in kilograms divided by the square of height in metres. Venous blood was collected after informed consent. A complete blood count, including haemoglobin, red blood cell (RBC) count, MCV, red cell distribution width (RDW), platelet count and white blood cell count, was performed on a Sysmex automated haematology analyser, and a peripheral smear was examined. Fasting and postprandial plasma glucose and serum creatinine were measured. HbA1c was measured by turbidimetric inhibition immunoassay .Absolute HbA1c, expressed in g/dL, was calculated as HbA1c (%) multiplied by haemoglobin (g/dL) and divided by 100, as described by Sinha and colleagues14. Anaemia severity among cases was graded with the World Health Organization haemoglobin thresholds as severe (below 8.0 g/dL), moderate (8.0–10.9 g/dL) or mild (11.0 g/dL up to the sex-specific anaemia threshold)22. Because only one case had mild anaemia, mild and moderate categories were combined for analysis. The Mentzer index (MCV divided by RBC count) was calculated for each case; a value above 13 favours iron deficiency over thalassaemia trait23. Data quality The study master chart was re-entered from the source record and verified against the original summary tables; every group mean and standard deviation for the haematological and glycaemic variables was reproduced exactly. Mean corpuscular haemoglobin and mean corpuscular haemoglobin concentration in the master chart were calculated values rather than analyser output, and several calculated concentrations were physiologically implausible for microcytic anaemia. These two indices were therefore not analysed. No variable had missing values. Statistical analysis Continuous variables are presented as mean ± standard deviation (SD) and, where distributions were skewed, as median and interquartile range (IQR); categorical variables are presented as number and percentage. Normality was assessed with the Shapiro–Wilk test and equality of variances with Levene's test. Groups were compared using Welch's unequal-variance t-test, with the mean difference (MD) and its 95% confidence interval (CI); the Mann–Whitney U test was used as a distribution-free sensitivity analysis for the primary outcome. Categorical variables were compared with Pearson's χ² test or Fisher's exact test where expected cell counts were small. The association of HbA1c with haematological and glycaemic variables was examined with Pearson's correlation coefficient (r) and Spearman's rank correlation (ρ), in the whole sample and separately within each group. Multivariable linear regression was used to estimate the difference in HbA1c between cases and controls adjusted for age, sex, BMI and fasting plasma glucose, which were chosen in advance as established determinants of HbA1c; a second model added postprandial glucose. Effect modification by sex was tested with an anaemia × sex interaction term. Within cases, HbA1c was regressed on haemoglobin with and without adjustment for the same covariates. Heteroskedasticity-consistent (HC3) standard errors were used throughout, and collinearity was assessed with variance inflation factors. Analyses by severity and by sex were exploratory. Two sensitivity analyses were performed: one excluding five control women whose serum creatinine of 1.0 mg/dL exceeded the prespecified female threshold, and one excluding nine records whose demographic and glycaemic values were identical to another record and could represent duplicate entry. All tests were two-sided with α = 0.05. Analyses were performed in Python 3.13.17 using SciPy 1.18.1 and statsmodels 0.15.0 Ethical considerations The study was approved by the Institutional Ethics Committee of , Raichur institute of medical sciences (Rims), Raichur, Karnataka, India and conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants. Participant data were anonymised before analysis.

Participant characteristics

All 100 enrolled participants (50 cases and 50 controls) were included in the analysis, and no data were missing. Participants were aged 12–65 years (median 17) and 76 were women. Cases and controls did not differ significantly in age (26.0 ± 15.8 vs 21.4 ± 13.0 years; p=0.117), sex (women 80.0% vs 72.0%; p=0.349), age-group distribution (p=0.306), occupation (p=0.362) or diet (vegetarian 16.0% vs 10.0%; p=0.372) (Table 1). Among women, a history of menorrhagia was equally common in cases and controls (62.5% vs 61.1%; p=0.901). Cases had a lower BMI than controls (19.92 ± 3.72 vs 22.51 ± 3.06 kg/m²; MD −2.60, 95% CI −3.95 to −1.24; p<0.001).

Haematological and biochemical parameters

As required by the case definition, cases had lower haemoglobin (8.15 ± 1.38 vs 13.71 ± 0.83 g/dL), RBC count, and MCV (69.62 ± 3.17 vs 91.20 ± 3.67 fL), and a higher RDW (19.65 ± 2.88 vs 13.21 ± 0.68%), all with p<0.001 (Table 2). Mean haemoglobin was 7.72 ± 1.77 g/dL in anaemic men and 8.26 ± 1.27 g/dL in anaemic women. Haemoglobin among cases ranged from 5.4 to 11.0 g/dL: 22 cases had severe, 27 moderate and one mild anaemia. The Mentzer index among cases averaged 22.9 ± 2.5 (range 18.6–28.3) and exceeded 13 in all 50. Platelet and white cell counts did not differ between groups.

Glycaemia was comparable. Fasting plasma glucose was 78.72 ± 10.14 mg/dL in cases and 81.14 ± 7.81 mg/dL in controls (MD −2.42, 95% CI −6.02 to 1.18; p=0.185), and postprandial glucose was 124.68 ± 8.96 and 126.32 ± 8.43 mg/dL respectively (MD −1.64, 95% CI −5.09 to 1.81; p=0.348). No participant had a fasting glucose of 100 mg/dL or above. Serum creatinine was lower in cases (0.66 ± 0.13 vs 0.83 ± 0.12 mg/dL; p<0.001).

Primary outcome: HbA1c

Mean HbA1c was 5.32 ± 0.33% in cases and 5.14 ± 0.53% in controls (MD 0.18%, 95% CI 0.005 to 0.357; Welch t=2.05, df 81.6; p=0.044), a standardised difference (Cohen's d) of 0.41 (Table 3; Figure 1). HbA1c was more dispersed among controls (Levene p<0.001): 13 controls but only three cases had an HbA1c of 4.6% or lower. Median HbA1c was 5.35% (IQR 5.20–5.57) in cases and 5.10% (IQR 4.65–5.50) in controls; the rank-based comparison did not reach significance (Mann–Whitney p=0.158). No participant had HbA1c of 6.5% or above, and the proportion with HbA1c in the prediabetes range of 5.7% or above did not differ (14.0% vs 22.0%; p=0.436).

Absolute HbA1c was lower in cases than in controls (0.437 ± 0.094 vs 0.704 ± 0.073 g/dL; MD −0.267, 95% CI −0.300 to −0.233; p<0.001).

Subgroups by sex and anaemia severity

The difference in HbA1c was confined to women. Anaemic women had a mean HbA1c of 5.37 ± 0.30% compared with 5.11 ± 0.55% in control women (MD 0.26%; p=0.014), whereas anaemic men had 5.15 ± 0.39% compared with 5.24 ± 0.49% in control men (MD −0.09%; p=0.611). The anaemia × sex interaction was significant (B 0.500, 95% CI 0.105 to 0.894; p=0.013).

HbA1c differed sharply by severity (Table 3; Figure 2). Cases with severe anaemia (n=22; mean haemoglobin 6.99 g/dL) had a mean HbA1c of 5.12 ± 0.29%, compared with 5.48 ± 0.27% in those with moderate or mild anaemia (n=28; mean haemoglobin 9.07 g/dL) (MD −0.36%; p<0.001). Relative to controls, HbA1c was 0.34% higher in moderate or mild anaemia (95% CI 0.16 to 0.52; p<0.001) but no different in severe anaemia (MD −0.02%, 95% CI −0.21 to 0.17; p=0.827).

Cases with severe anaemia differed from those with moderate or mild anaemia in other respects. They were younger (mean age 21.1 vs 29.9 years), had a lower BMI (18.78 vs 20.81 kg/m²) and a lower fasting glucose (75.36 vs 81.36 mg/dL), and had a higher RDW (20.77 ± 2.80 vs 18.77 ± 2.67%; p=0.014). The proportion of women was similar (17 of 22 vs 23 of 28). None of the 22 cases with severe anaemia had an HbA1c of 5.7% or above (range 4.6–5.5%), compared with 7 of 28 (25.0%) cases with moderate or mild anaemia; the latter proportion did not differ significantly from that among controls (p=0.785). The six cases with haemoglobin below 6 g/dL had a mean HbA1c of 4.70%.

Correlates of HbA1c

Within cases, HbA1c correlated strongly and positively with haemoglobin (r=0.75; ρ=0.66; both p<0.001) and negatively with RDW (r=−0.51; p<0.001), but not with MCV (r=0.12; p=0.396), RBC count (r=−0.05; p=0.738), fasting glucose (r=0.03; p=0.840) or postprandial glucose (r=−0.17; p=0.239) (Table 4). Within controls, HbA1c correlated negatively with RBC count (r=−0.39; p=0.005) and positively with BMI (r=0.38; p=0.006) and RDW (r=0.29; p=0.041). In the whole sample, HbA1c was not correlated with haemoglobin (r=−0.11; p=0.284) or fasting glucose (r=0.04; p=0.661).

Multivariable analysis

After adjustment for age, sex, BMI and fasting glucose, the difference in HbA1c between cases and controls increased from 0.181% (95% CI 0.006 to 0.356; p=0.043) to 0.314% (95% CI 0.115 to 0.513; p=0.002) (Table 5). BMI was independently associated with HbA1c (B 0.043% per kg/m², 95% CI 0.022 to 0.064; p<0.001); age, sex and fasting glucose were not. Additional adjustment for postprandial glucose left the estimate unchanged (B 0.310, 95% CI 0.102 to 0.518; p=0.004). Variance inflation factors were all 1.35 or lower. Within cases, HbA1c increased by 0.177% per 1 g/dL increase in haemoglobin (95% CI 0.137 to 0.217; R²=0.557), and by 0.181% (95% CI 0.132 to 0.230) after covariate adjustment. RDW was strongly correlated with haemoglobin among cases (r=−0.70) and was not associated with HbA1c once haemoglobin was in the model (B 0.002, 95% CI −0.034 to 0.038; p=0.903).

Sensitivity analyses

Excluding the five control women with serum creatinine above the prespecified threshold did not change the findings (unadjusted MD 0.20%, p=0.040; adjusted B 0.308, 95% CI 0.101 to 0.515; p=0.004). Excluding nine possible duplicate records (n=91) also yielded similar estimates (unadjusted MD 0.19%, p=0.041; adjusted B 0.321, 95% CI 0.126 to 0.517; p=0.001).

Table 1. Demographic and clinical characteristics of cases and controls

ᵃ Welch's t-test. ᵇ Pearson's χ² test. BMI, body mass index; SD, standard deviation.

Table 2. Haematological and biochemical parameters

Values are mean ± SD. Mean difference = anaemia minus controls; p from Welch's t-test. Confidence intervals for sex-specific haemoglobin differences are omitted for brevity. MCV, mean corpuscular volume; RBC, red blood cell; RDW, red cell distribution width.

Table 3. HbA1c and absolute HbA1c by group, sex and anaemia severity

ᵃ Welch's t-test. ᵇ Mann–Whitney U test. ᶜ Fisher's exact test. Severe versus moderate/mild anaemia: difference −0.36%, p<0.001. Absolute HbA1c = HbA1c (%) × haemoglobin (g/dL) / 100. Hb, haemoglobin; IQR, interquartile range.

Table 4. Correlation of HbA1c with haematological, glycaemic and anthropometric variables

Pearson's correlation coefficient. Spearman's ρ for haemoglobin among cases was 0.66 (p<0.001). Abbreviations as in Table 2.

Table 5. Linear regression of HbA1c (%) on anaemia status and covariates

Ordinary least squares with heteroskedasticity-consistent (HC3) standard errors. B is the difference in HbA1c (percentage points). Variance inflation factors ≤1.35. BMI, body mass index.

Figure 1. Distribution of HbA1c in non-anaemic controls and participants with microcytic hypochromic anaemia.

Boxes show median and interquartile range; whiskers show range; dots are individual participants; diamonds with bars are means with 95% confidence intervals. The dashed line marks the 5.7% prediabetes threshold. Difference in means from Welch's t-test.

Figure 2. HbA1c according to haemoglobin concentration and anaemia severity.

(A) HbA1c against haemoglobin in the 50 anaemic participants, with the least-squares regression line (B = 0.177% per g/dL; r = 0.75); the dotted line shows the mean HbA1c of controls. (B) Mean HbA1c (95% CI) in controls, moderate/mild anaemia (haemoglobin 8.0–11.0 g/dL) and

In this hospital-based case–control study from South India, non-diabetic individuals with microcytic hypochromic anaemia had a modestly higher HbA1c than non-anaemic controls despite comparable fasting and postprandial glucose. The difference roughly doubled after adjustment for BMI, was confined to women, and was not uniform across the range of anaemia: HbA1c was raised by about a third of a percentage point in moderate anaemia but was indistinguishable from controls in severe anaemia, producing a strong positive relationship between haemoglobin and HbA1c within the anaemic group. Comparison with previous studies The direction of the main finding agrees with most of the literature on iron deficiency. Elevated HbA1c in iron deficiency has been reported since Brooks and colleagues9, and has been confirmed in non-diabetic adults in Turkey10 and Chennai11, in women in NHANES16, in young Indians from Pune17 and in diabetic patients with controlled glycaemia from Mangalore24. The most recent meta-analysis pooled a difference of 0.66%19. Our crude difference of 0.18% is considerably smaller than the 1.5 percentage points (7.4 − 5.9) reported by Coban and colleagues10 and the 2.1 percentage points (7.6 − 5.5) reported from Chennai11. It is closer to the modest shifts reported by Son and colleagues, who found no difference when glucose was normal but a higher HbA1c in anaemic subjects once glucose exceeded the prediabetes threshold25, and by Rafat and colleagues in non-diabetic pregnant women with iron deficiency26. Three features of our sample probably account for the smaller magnitude. Glycaemia was uniformly normal, with no fasting glucose of 100 mg/dL or above, and Christy and colleagues found that the effect of iron deficiency was greatest in people with fasting glucose of 100–126 mg/dL24; glycation depends on glucose concentration, so a fixed increase in erythrocyte age generates less additional HbA1c at lower glucose. HbA1c was measured by immunoassay, and the 2026 meta-analysis identified assay method as a major source of heterogeneity19. Finally, almost half of our cases had severe anaemia, which, as discussed below, did not show the effect. Our findings also help to reconcile the discordant studies. Sinha and colleagues found a lower baseline HbA1c in iron deficiency anaemia that rose with treatment14; in NHANES, HbA1c rose steadily across the population distribution of haemoglobin15; and in Saudi women, HbA1c was inversely related to haemoglobin even though group means did not differ27. A single direction of effect cannot explain all of these results. A non-linear relationship can: if iron deficiency raises HbA1c at moderate degrees of anaemia but the effect disappears or reverses as anaemia becomes severe, the observed group difference will depend on the severity mix of the sample. Systematic reviews have flagged exactly this uncertainty, noting that the degree of anaemia that materially affects HbA1c is unknown6 and that relationships with severity are inconsistent across studies20. To our knowledge, few studies have shown within a single sample that the elevation present in moderate anaemia is absent in severe anaemia. The finding is exploratory and needs replication, but it offers a testable explanation for two decades of conflicting reports. Possible mechanisms The leading explanation for elevated HbA1c in iron deficiency concerns erythrocyte age. When iron limits erythropoiesis, fewer new erythrocytes enter the circulation and the mean age of circulating cells increases; because glycation accumulates over each cell's lifespan, an older red-cell population carries more HbA1c at the same glucose concentration4,5. The rapid fall in HbA1c with the reticulocytosis that follows iron or vitamin B12 replacement supports this model13, and a meta-analysis of iron replacement trials confirmed that HbA1c falls with treatment28. A second, complementary mechanism is oxidative: plasma malondialdehyde and fructosamine are raised in iron deficiency anaemia and fall after iron therapy, and lipid peroxidation products may promote glycation independently of glucose29. Why the effect should disappear in severe anaemia is less certain, and the following explanations are hypotheses. Severe iron deficiency produces fragile, poorly haemoglobinised microcytes whose survival may be shortened, offsetting the ageing effect of reduced production. Severe anaemia also imposes a strong erythropoietic drive, so intermittent bursts of reticulocyte release, for example after dietary iron intake or blood loss, could lower mean erythrocyte age. The inverse association between RDW and HbA1c among cases is consistent with a more heterogeneous, younger red-cell population in severe anaemia, although RDW carried no information beyond haemoglobin itself. An analytical contribution cannot be excluded: immunoassays express HbA1c as a ratio to total haemoglobin measured in the same sample, and at very low haemoglobin concentrations the denominator approaches the lower end of the assay's validated range. Studies that measure erythrocyte lifespan, reticulocyte counts and HbA1c by more than one method across the severity spectrum would distinguish these possibilities. Confounding should also be considered. Cases with severe anaemia had a slightly lower fasting glucose than those with moderate anaemia, and lower glycaemia would lower HbA1c. However, the difference in fasting glucose was small, glucose was not correlated with HbA1c among cases, and the slope of HbA1c on haemoglobin was essentially unchanged after adjustment for fasting glucose, BMI, age and sex (0.177% unadjusted vs 0.181% adjusted per g/dL). Differences in glycaemia are therefore unlikely to explain the severity gradient, although residual confounding by unmeasured nutritional factors cannot be excluded. A further observation supports a role for red-cell turnover even within the normal range. Among controls, HbA1c fell as the RBC count rose (r=−0.39), while glucose was unrelated to HbA1c. This pattern is consistent with the concept of a glycation gap, in which individuals with similar glucose exposure differ in HbA1c because of differences in erythrocyte survival5. It cautions against interpreting the wider spread of HbA1c among controls as measurement noise; inter-individual variation in red-cell kinetics probably contributes to HbA1c in non-anaemic people as well4. Sex, body mass index and absolute HbA1c The confinement of the effect to women mirrors NHANES, where the shift in HbA1c distribution associated with iron deficiency was observed in women16. Menstrual iron loss produces a chronic, slowly evolving iron deficiency that may favour the prolonged erythrocyte survival underlying HbA1c elevation. However, only ten men were anaemic, and the subgroup estimate in men is too imprecise for any conclusion; the interaction should be regarded as hypothesis-generating. Adjustment for BMI strengthened the association. Anaemic participants were thinner, and BMI was positively associated with HbA1c, so the lower BMI of cases partly masked the anaemia-related increase. This negative confounding is plausible in India, where anaemia and undernutrition co-occur geographically and socially8, and it suggests that unadjusted comparisons in undernourished populations may underestimate the effect of iron deficiency on HbA1c. Absolute HbA1c was substantially lower in anaemic participants, as reported by Sinha and colleagues14 and in Saudi women27. Because absolute HbA1c is the product of HbA1c and haemoglobin, this difference follows largely from the lower haemoglobin of cases and should not be read as evidence against increased glycation per unit of haemoglobin. We found no evidence that absolute HbA1c offers a diagnostic advantage over conventional HbA1c in anaemia. Clinical relevance The analytical method deserves attention in interpreting these results. Because the immunoassay used here reports glycated haemoglobin as a fraction of total haemoglobin, the result is sensitive to anything that changes the glycated fraction within the red-cell population rather than the absolute amount of glycated protein. This is the appropriate property for a marker of glycaemia, but it also means that the test faithfully records changes in erythrocyte age that have nothing to do with glucose. Laboratories that serve populations with a high prevalence of anaemia could usefully report haemoglobin alongside HbA1c and flag results from anaemic samples, as already recommended for haemoglobin variants4. The absolute shift we observed is small relative to the 6.5% diagnostic threshold, no participant reached that value, and the proportion with HbA1c in the prediabetes range did not differ between groups. Bias near the 5.7% threshold is nonetheless relevant in India, where HbA1c-based screening of young, frequently iron-deficient populations has been shown to inflate the apparent prevalence of prediabetes17. An upward bias of 0.3% in moderate iron deficiency, combined with an absent or opposite bias in severe anaemia, means that HbA1c cannot be corrected by a single adjustment factor. Clinicians should check haemoglobin and red-cell indices when interpreting HbA1c and, in line with ADA guidance3, use glucose-based criteria when anaemia is present and HbA1c is near a decision threshold. In people with diabetes, where pooled evidence on iron deficiency and HbA1c is conflicting30, iron status should be considered before treatment is intensified on the basis of an unexpectedly high HbA1c. Strengths Fasting and postprandial glucose were measured in both groups and were comparable, allowing the difference in HbA1c to be attributed to non-glycaemic factors. The dataset was complete and was verified against the original records. We estimated adjusted effects with robust standard errors, tested effect modification by sex, examined severity within the anaemic group and confirmed the findings in two sensitivity analyses. The Mentzer index supported iron deficiency as the predominant cause of microcytosis. Limitations Several limitations should be considered. Iron status was not measured: without serum ferritin or transferrin saturation, iron deficiency anaemia could not be confirmed, and although every case had a Mentzer index above 13, this index cannot exclude thalassaemia trait or anaemia of inflammation, and haemoglobin electrophoresis was not performed. Postprandial glucose was not obtained under a standardised oral glucose tolerance test, and HbA1c was measured once, with a single immunoassay that cannot detect haemoglobin variants. The sample was small, no a priori sample size calculation was made, and the analyses by sex and severity were exploratory; detecting the observed crude difference of 0.18% with 80% power at α = 0.05 would require about 95 participants per group (pooled SD 0.44%; d = 0.18/0.44 = 0.41). Only ten anaemic men were included. Controls were comparable in age and sex at group level but were not individually matched, and the sample was hospital-based and young (median age 17 years), which limits generalisability to older adults in whom diabetes screening is most common. Five control women had serum creatinine slightly above the prespecified threshold and nine records may have been duplicate entries, although excluding either set did not change the results. Mean corpuscular haemoglobin and concentration could not be analysed. Vitamin B12 and folate status, which are frequently deficient in young Indians and themselves influence HbA1c13,17, were not measured, nor were reticulocyte counts or markers of inflammation; the severity analysis may also be confounded by the younger age and lower BMI of severely anaemic participants. Finally, the design was cross-sectional in its measurement, so we could not observe the change in HbA1c after correction of anaemia. Clinical implications In settings with a high prevalence of iron deficiency, a borderline HbA1c should be interpreted together with the haemoglobin level and red-cell indices. When anaemia is present, glucose-based tests should be used to diagnose or exclude diabetes and prediabetes, and the direction of any bias should not be assumed to be the same in moderate and severe anaemia. Future research Larger studies with full iron studies, haemoglobin electrophoresis, a standardised oral glucose tolerance test, reticulocyte counts and HbA1c measured by both chromatographic and immunoassay methods are needed to define how the effect of iron deficiency on HbA1c varies across the full range of severity and between sexes. Longitudinal studies measuring HbA1c before and after iron repletion would establish whether the severity-dependent pattern reflects changes in erythrocyte lifespan.

Among non-diabetic individuals with comparable glycaemia, microcytic hypochromic anaemia was associated with a modest increase in HbA1c, which was larger after adjustment for BMI and was observed in women. The elevation was present in moderate anaemia but absent in severe anaemia, indicating a non-linear relationship that may explain the conflicting results of earlier studies. HbA1c near diagnostic thresholds should be interpreted with caution in anaemic individuals and confirmed with glucose-based testing.

1. Anjana RM, Unnikrishnan R, Deepa M, Pradeepa R, Tandon N, Das AK, et al. Metabolic non-communicable disease health report of India: the ICMR-INDIAB national cross-sectional study (ICMR-INDIAB-17). Lancet Diabetes Endocrinol. 2023;11(7):474-89. doi:10.1016/S2213-8587(23)00119-5 2. International Expert Committee. International Expert Committee report on the role of the A1C assay in the diagnosis of diabetes. Diabetes Care. 2009;32(7):1327-34. doi:10.2337/dc09-9033 3. American Diabetes Association Professional Practice Committee. 2. Diagnosis and classification of diabetes: Standards of Care in Diabetes—2025. Diabetes Care. 2025;48(Suppl 1):S27-49. doi:10.2337/dc25-S002 4. Radin MS. Pitfalls in hemoglobin A1c measurement: when results may be misleading. J Gen Intern Med. 2014;29(2):388-94. doi:10.1007/s11606-013-2595-x 5. Cohen RM, Franco RS, Khera PK, Smith EP, Lindsell CJ, Ciraolo PJ, et al. Red cell life span heterogeneity in hematologically normal people is sufficient to alter HbA1c. Blood. 2008;112(10):4284-91. doi:10.1182/blood-2008-04-154112 6. English E, Idris I, Smith G, Dhatariya K, Kilpatrick ES, John WG. The effect of anaemia and abnormalities of erythrocyte indices on HbA1c analysis: a systematic review. Diabetologia. 2015;58(7):1409-21. doi:10.1007/s00125-015-3599-3 7. GBD 2021 Anaemia Collaborators. Prevalence, years lived with disability, and trends in anaemia burden by severity and cause, 1990-2021: findings from the Global Burden of Disease Study 2021. Lancet Haematol. 2023;10(9):e713-34. doi:10.1016/S2352-3026(23)00160-6 8. Ghosh K, Chakraborty AS, Haloi B, Zakir S. Spatial clustering of malnutrition and anemia among reproductive women and its associated risk factors in India: evidence from National Family Health Survey-5. Food Nutr Bull. 2024;45(1):3-11. doi:10.1177/03795721241234086 9. Brooks AP, Metcalfe J, Day JL, Edwards MS. Iron deficiency and glycosylated haemoglobin A. Lancet. 1980;2(8186):141. doi:10.1016/s0140-6736(80)90019-7 10. Coban E, Ozdogan M, Timuragaoglu A. Effect of iron deficiency anemia on the levels of hemoglobin A1c in nondiabetic patients. Acta Haematol. 2004;112(3):126-8. doi:10.1159/000079722 11. Shanthi B, Revathy C, Manjula Devi AJ, Subhashree. Effect of iron deficiency on glycation of haemoglobin in nondiabetics. J Clin Diagn Res. 2013;7(1):15-7. doi:10.7860/JCDR/2012/4881.2659 12. Tarim O, Küçükerdoğan A, Günay U, Eralp O, Ercan I. Effects of iron deficiency anemia on hemoglobin A1c in type 1 diabetes mellitus. Pediatr Int. 1999;41(4):357-62. doi:10.1046/j.1442-200x.1999.01083.x 13. Gram-Hansen P, Eriksen J, Mourits-Andersen T, Olesen L. Glycosylated haemoglobin (HbA1c) in iron- and vitamin B12 deficiency. J Intern Med. 1990;227(2):133-6. doi:10.1111/j.1365-2796.1990.tb00131.x 14. Sinha N, Mishra TK, Singh T, Gupta N. Effect of iron deficiency anemia on hemoglobin A1c levels. Ann Lab Med. 2012;32(1):17-22. doi:10.3343/alm.2012.32.1.17 15. Ford ES, Cowie CC, Li C, Handelsman Y, Bloomgarden ZT. Iron-deficiency anemia, non-iron-deficiency anemia and HbA1c among adults in the US. J Diabetes. 2011;3(1):67-73. doi:10.1111/j.1753-0407.2010.00100.x 16. Kim C, Bullard KM, Herman WH, Beckles GL. Association between iron deficiency and A1C levels among adults without diabetes in the National Health and Nutrition Examination Survey, 1999-2006. Diabetes Care. 2010;33(4):780-5. doi:10.2337/dc09-0836 17. Hardikar PS, Joshi SM, Bhat DS, Raut DA, Katre PA, Lubree HG, et al. Spuriously high prevalence of prediabetes diagnosed by HbA(1c) in young Indians partly explained by hematological factors and iron deficiency anemia. Diabetes Care. 2012;35(4):797-802. doi:10.2337/dc11-1321 18. Cavagnolli G, Pimentel AL, Freitas PA, Gross JL, Camargo JL. Factors affecting A1C in non-diabetic individuals: review and meta-analysis. Clin Chim Acta. 2015;445:107-14. doi:10.1016/j.cca.2015.03.024 19. Souza EMC, Domingueti CP. Interference of iron deficiency in glycated hemoglobin testing and its clinical implications: a systematic review and meta-analysis. Biochem Med (Zagreb). 2026;36(3):030503. doi:10.11613/BM.2026.030503 20. Hag Elagib MA, Ahmed RA, Eltom H, Abdelrahman MF, Mohammed MMEA, Ahmed A, et al. Iron deficiency, iron deficiency anemia, and HbA1c levels in nondiabetic adults: a systematic review and meta-analysis. Cureus. 2026;18(8):e115191. doi:10.7759/cureus.115191 21. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370(9596):1453-7. doi:10.1016/S0140-6736(07)61602-X 22. World Health Organization. Haemoglobin concentrations for the diagnosis of anaemia and assessment of severity. Vitamin and Mineral Nutrition Information System (WHO/NMH/NHD/MNM/11.1) [Internet]. Geneva: WHO; 2011 [cited 2026 Oct 23]. Available from: https://www.who.int/publications/i/item/WHO-NMH-NHD-MNM-11.1 23. Mentzer WC Jr. Differentiation of iron deficiency from thalassaemia trait. Lancet. 1973;1(7808):882. doi:10.1016/s0140-6736(73)91446-3 24. Christy AL, Manjrekar PA, Babu RP, Hegde A, Rukmini MS. Influence of iron deficiency anemia on hemoglobin A1c levels in diabetic individuals with controlled plasma glucose levels. Iran Biomed J. 2014;18(2):88-93. doi:10.6091/ibj.1257.2014 25. Son JI, Rhee SY, Woo JT, Hwang JK, Chin SO, Chon S, et al. Hemoglobin A1c may be an inadequate diagnostic tool for diabetes mellitus in anemic subjects. Diabetes Metab J. 2013;37(5):343-8. doi:10.4093/dmj.2013.37.5.343 26. Rafat D, Rabbani TK, Ahmad J, Ansari MA. Influence of iron metabolism indices on HbA1c in non-diabetic pregnant women with and without iron-deficiency anemia: effect of iron supplementation. Diabetes Metab Syndr. 2012;6(2):102-5. doi:10.1016/j.dsx.2012.05.011 27. Bindayel IA. Influence of iron deficiency anemia on glycated hemoglobin levels in non-diabetic Saudi women. J Int Med Res. 2021;49(2):300060521990157. doi:10.1177/0300060521990157 28. AlQarni AM, Alghamdi AA, Aljubran HJ, Bamalan OA, Abuzaid AH, AlYahya MA. The effect of iron replacement therapy on HbA1c levels in diabetic and nondiabetic patients: a systematic review and meta-analysis. J Clin Med. 2023;12(23):7287. doi:10.3390/jcm12237287 29. Sundaram RC, Selvaraj N, Vijayan G, Bobby Z, Hamide A, Rattina Dasse N. Increased plasma malondialdehyde and fructosamine in iron deficiency anemia: effect of treatment. Biomed Pharmacother. 2007;61(10):682-5. doi:10.1016/j.biopha.2007.06.013 30. Kuang L, Li W, Xu G, You M, Wu W, Li C. Systematic review and meta-analysis: influence of iron deficiency anemia on blood glycosylated hemoglobin in diabetic patients. Ann Palliat Med. 2021;10(11):11705-13. doi:10.21037/apm-21-29