Prevalence and causes of anemia among older adults in India: findings from wave 2 of the Harmonized Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD)

TL;DR

This study finds that nearly half of older adults in India have anemia, with significant regional and gender differences, and identifies nutritional deficiencies as a major cause.

Contribution

The study provides the first nationally representative data on anemia prevalence and causes among older adults in India.

Findings

Anemia prevalence among older adults in India is 49.92%, with higher rates in women compared to men.

Nutritional anemia, particularly due to iron deficiency, is the most common cause of anemia in this population.

Regional disparities in anemia prevalence are significant, with the highest rates in Assam, West Bengal, Jharkhand, and Odisha.

Abstract

Anemia among older adults aged ≥ 60 years is a well-described risk factor that increases the risk of falls, cardiovascular diseases, and mortality. In India, objectively measured national estimates of anemia prevalence and the causes of anemia among older adults are lacking. The Harmonized Diagnostic Assessment of Dementia for the Longitudinal Aging Study in India (LASI-DAD) collected venous blood samples from a nationally representative sample of 3,252 individuals in wave 2 of the study. Out of these, 3,009 samples were used to estimate national prevalence and regional differences in anemia prevalence and its underlying causes. Anemia was defined as hemoglobin < 13 mg/dl in males and < 12 mg/dl in females and further categorized into nutritional and non-nutritional anemia based on several nutritional (ferritin, Vitamin B12, and folate), inflammatory (ferritin, C-reactive protein), and renal (serum creatinine) biomarkers. The overall national anemia prevalence was 49.92%. Anemia prevalence was significantly higher among women (53.9%) than men (45.8%). Among those with anemia, nutritional anemia was found in 63.5% of the respondents, with isolated iron deficiency anemia being the most common cause (51.8%). Among the non-nutritional category, anemia of chronic disease was the most common type (31.4%). About 10% of all anemia cases could not be classified into either category using the available data. Regional differences were seen with the highest prevalence of anemia in Assam, West Bengal, Jharkhand, and Odisha, and the lowest prevalence (< 30%) in Jammu & Kashmir and Haryana. Multivariate adjustment showed that age, sex, and region of residence are independently associated with anemia status. The study provides valuable insights into the overall anemia prevalence among older adults in India, its underlying causes, and regional differences to lay a strong foundation for making informed decisions toward anemia control in India.

The online version contains supplementary material available at 10.1186/s12939-025-02671-4.

Full text

Background

Anemia is recognized as a public health concern with a global prevalence of 24.3% across all age groups [1]. According to the Global Burden of Disease, Injuries, and Risk Factors study, the sub-Saharan African and South-Asian countries have the highest levels [1]. Children, adolescents, women of reproductive age, and older adults (≥ 60 years) have been identified to be at high risk of developing anemia [2]. A recent systematic review and meta-analysis of cross-sectional studies reported a pooled anemia prevalence of 68.3% among older adults in India [3]. Anemia can significantly impact an individual’s daily activities, especially older adults, as it adversely affects physical [4] and cognitive functioning [5, 6], increases the risk of falls [7], frailty [8], cardiovascular [9] diseases, and mortality [10].

The pathophysiology of anemia can either be multifactorial or result from a single cause [11]. In India, nutritional deficiencies [12] related to iron, vitamin B12, and folate are the predominant causes of anemia. The non-nutritional causes include the presence of chronic disease, such as chronic kidney disease [13], myelosuppression, inflammatory or autoimmune diseases, blood loss due to parasitic infections, and malignancies [2]. In the older population, in addition to the reasons mentioned above, low energy intake, nutrient malabsorption, anorexia, immunodeficiency, medications, and lower socioeconomic status, especially in India, play an important role in anemia development [14]. However, up to 30% of anemia cases among older adults may be unexplained and difficult to categorize [13].

Statement of the problem

Several previous studies have attempted to estimate the prevalence of anemia and identify its causes among older adults in India. However, the evidence is limited as it reflects a small non-representative sample population mostly recruited at local institutions [15, 16] or hospitals [17, 18], in urban [19–24] or rural [25–32] community settings, but limited to one state [33–35]. Moreover, the prevalence of anemia reported in these studies varies widely, ranging from 21% to 96%.

In India, preventive measures to curb anemia were first implemented in 1970 through the National Anemia Prophylaxis Program [36]. However, since then, efforts have primarily been directed toward reducing anemia prevalence among children and women of reproductive age. In 2016, the Longitudinal Aging Study in India [37] (LASI) was launched as part of the National Program for the Health Care of the Elderly to study aging patterns in a nationally representative sample in India. The Harmonized Diagnostic Assessment of Dementia for LASI (LASI-DAD) studied a nationally representative sub-sample of LASI and collected venous blood samples. In this study, we report comprehensive national estimates of anemia prevalence and causes of anemia among older adults aged 60 years and above in India. Published literature indicates that regional variation in dietary, socio-economic, lifestyle, and cultural patterns may affect the distribution of different anemia subtypes and also help identify geographical clustering [38–40]. Therefore, we also present regional differences in anemia prevalence in the LASI-DAD sample.

Methods

Participants

The Longitudinal Aging Study in India (LASI) is an ongoing study that collects socioeconomic and health-related data from over 73,000 adults aged 45 years and above and their spouses. Its sample was drawn from the 2011 Census, representing each state and union territory as well as the country as a whole. A sub-sample of LASI was drawn to collect data from 4,096 to 4,638 LASI respondents during Wave 1 and Wave 2 of the LASI-DAD study, respectively [41, 42]. Wave 1 of the LASI-DAD study was launched in 2017, and wave 2 in 2022 to study late-life cognition and estimate the prevalence of dementia among older adults aged 60 years and above in India [41, 42]. Key demographic variables, such as age, gender, and urbanicity, were obtained from the main LASI and used in both waves of the LASI-DAD study [41–43]. Information on the highest level of education was obtained from the respondents during the interview and recorded as less than primary education, primary education, middle school, secondary education, higher secondary school, diploma/certificate holders, graduate degree, postgraduate degree, and professional degree. The detailed study design and methodology have been published elsewhere [41, 42]. The study was conducted according to the Declaration of Helsinki. Ethics approval was obtained from the Institute Ethics Committee of All India Institute of Medical Sciences (AIIMS), New Delhi, all collaborating medical institutions, and the University of Southern California (UP-15-00684). The Health Ministry Screening Committee of the Indian Council of Medical Research provided administrative clearance to the LASI-DAD research study (2202–16741/F1).

Venous blood sample collection and processing

For Wave 2 of the LASI-DAD study, 3252 out of 4638 (70.1%) participants consented to the collection of a 17 mL venous blood specimen. Blood draws were conducted across India from November 2022 to April 2024. A written informed consent form was obtained from each respondent in the presence of a family member. A consent form was translated into the participant’s local language, and consent was obtained with the participant’s signature or a thumbprint in cases where the respondent could not read or write. For cognitively impaired participants, consent was obtained from a legal representative authorized to sign on their behalf. The samples were collected by trained phlebotomists at each respondent’s house and shipped under ambient temperature (2–8° Celsius) to the local Metropolis Healthcare Ltd. laboratory for processing within 4–6 h of collection. Detailed procedures for training of phlebotomists and other study staff have been previously described elsewhere [44]. The serum and plasma tubes were centrifuged at the local Metropolis Healthcare Ltd. laboratory before shipment to the central laboratory. Within 24 h of collection, these samples were then shipped under a cold chain from the local Metropolis laboratory to the Metropolis Healthcare Ltd. Laboratory in Delhi. A temperature logger was placed with each shipment to record the temperature changes during transportation. Deviation from cold chain temperature was observed during hot periods, especially in regions with tropical or high-temperature weather conditions. In such regions, the field investigators and phlebotomists carried additional ice packs to maintain sample integrity during venous blood specimen collection. The ice packs were also replaced at quality check points: before shipping the samples from the point of collection to the local laboratory, and then from the local laboratory to the central laboratory. At the central laboratory, the serum and plasma samples were aliquoted into 0.5 ml aliquots for long-term storage and performing various biochemical assays. The whole blood assays included Glycosylated hemoglobin (HbA1c), and complete blood cell count (CBC) with differential leucocyte counts. The serum-based assays comprised lipid profile, liver and kidney function tests, thyroid function tests, and other tests, including vitamin B12, folic acid, vitamin D, homocysteine, pro-B-type natriuretic peptide, high-sensitivity C-reactive protein (hsCRP), and lipoprotein-a estimation. We utilized standardized diagnostic criteria of different analytes to classify anemia and its subtypes, such as hemoglobin, serum ferritin, transferrin saturation, vitamin B12, folate, C-reactive protein, and chronic kidney disease estimated using the Chronic Kidney Disease Collaboration (CKD-EPI) equation that utilizes both serum creatinine and cystatin C [45]. The detailed methodology for venous blood sample collection, processing, transportation, with maintenance of the cold chain, and sample storage in LASI-DAD has been published elsewhere [44].

Equipment used for measuring analytes

For measuring analytes relevant to anemia, we employed standardized methods, quality assurance, and equipment for each venous blood sample [44]. Complete blood counts were measured using the Beckman Coulter DxH 8000 instrument. We used the Roche Cobas 8000 (i502 & e801) to quantify serum creatinine, cystatin C, uric acid, vitamin B12, and folic acid. hs-CRP was quantified by nephelometry using Atellica NEPH 630.

Measures

Primary outcome

Anemia was defined as hemoglobin < 13 mg/dl in males and < 12 mg/dl in females [46]. The criteria used to classify anemia into different categories are summarized in Table 1.

Table 1. Criteria used to classify anemia in the LASI-DAD studyClassification of anemiaDefinitions and measures Nutritional anemia  Iron deficiency anemiaSerum ferritin level < 30 ng/mL [47] or serum ferritin level of 30–100 ng/mL [48] and transferrin saturation [48] of < 20%. Vitamin B12 deficiency anemiaVitamin B12 levels < 200 pg/mL.[49] Folate deficiency anemiaSerum folate level < 3 ng/dL.[50] Multiple nutritional deficienciesMore than one nutritional deficiency Non-nutritional anemia Cases that did not fall under any category of nutritional deficiency anemia Anemia of chronic disease/inflammationSerum ferritin levels > 100 ng/ml or hsCRP > 3 mg/dl Anemia of chronic kidney diseasee-Glomerular Filtration Rate (eGFR) < 60 ml/min Multiple non-nutritional causesMore than one cause of non-nutritional anemia Unexplained non-nutritional anemiaNon-nutritional anemia cases that did not fit into the above criteria

Covariates

We used age, gender, educational level, urbanicity, and geographical region as covariates. For these analyses, the educational level was recategorized as – no education, less than high school education, completed high school, and college or a higher degree. The individual states were grouped into four regions based on geographical proximity and prevalence of anemia in the individual states to facilitate analysis of regional differences in anemia.

Statistical analysis

Descriptive statistics were calculated after accounting for sample weights to characterize the study cohort. Two-sample t-tests were used to compare the distribution of continuous variables, such as age, while the χ^2^ test was used to compare the distributions of categorical variables. Weighted frequencies of anemia and its subgroups were calculated after adjusting for sampling weights that account for the complex stratified sampling design. Unconditional multivariate logistic regression models were used (adjusting for sampling weights as a fixed effect) to evaluate the effect of age, sex, regions within India, urban/rural residence, and educational status on anemia. All analyses were conducted using SAS version 9.4.

Results

The mean age of the participants in this study was 70.2 years, and 50.5% were women. Out of 3252 respondents who consented to a venous blood draw, blood samples from 3009 respondents were deemed sufficient in quantity and quality to conduct bioassay analysis. Among 3009 participants with hemoglobin levels in Wave 2 of LASI-DAD, the overall weighted anemia prevalence was 49.9% (95% CI: 48% − 51.9%) (n = 1502). The weighted prevalence of anemia was significantly higher among women as compared to men (53.9% vs. 45.8%; p < 0.001). Though the prevalence of anemia was higher in rural areas as compared to urban areas, this difference was not statistically significant, and anemia status did not differ significantly with respondents’ educational status (Table 2).

Table 2. Characteristics of the LASI-DAD study populationSample characteristicsAnemia(n = 1502)No anemia (n = 1507)P value Age (years) 70.17 (± 0.18)69.08 (± 0.15)< 0.001 Gender < 0.001 Male682 (45.8%)806 (54.2%) Female820 (53.9%)701 (46.1%) Education status 0.48 No school752 (50.8%)728 (49.2%) Less than high school622 (48.3%)667 (51.7%) High school40 (51.1%)39 (48.9%) College and higher87 (54.2%)74 (45.8%) Regions* < 0.001 Region 1351 (37.8%)577 (62.2%) Region 2244 (46.8%)277 (53.2%) Region 3308 (41.0%)443 (59.0%) Region 4599 (74.1%)209 (25.9%) Urban/rural setting 0.09 Urban420 (47.3%)467 (52.7%) Rural1082 (51.0%)1040 (49.0%)*Regions:Region 1: Jammu and Kashmir, Punjab, Uttarakhand, Haryana, Delhi, Rajasthan, Uttar Pradesh, BiharRegion 2: Maharashtra, Gujarat, and Madhya PradeshRegion 3: Andhra Pradesh, Karnataka, Tamil Nadu, Pondicherry, Telangana, KeralaRegion 4: Assam, West Bengal, Jharkhand, Odisha, Chhattisgarh

Among LASI-DAD participants with anemia, nutritional causes, which include iron, Vitamin B12, or folate deficiency anemia, characterized 63.6% of the cases (Table 3). Among those with nutritional anemia, isolated iron deficiency anemia was the most common type (51.8%) (Table 3). The second most common cause of nutritional anemia was multiple nutritional deficiencies (26.2%), followed by isolated Vitamin B12 deficiency (12.6%) and isolated folic acid deficiency (9.4%) (Table 3). Among the remaining participants with non-nutritional anemia, chronic disease/inflammation was the most common cause (39.5%) (Table 3). Other causes of non-nutritional anemia were CKD-associated anemia (9.7%) and multiple non-nutritional causes (24.1%) that included both CKD and anemia of chronic disease/inflammation (Table 2). The remaining 146 participants (26.7%) with non-nutritional anemia could not be classified into any identifiable causes of anemia (Table 3).

Table 3. Distribution of the different anemia types among the anemic LASI-DAD study populationAnemia typeAnemia (n = 1502)(95% CI)Percentages (95% CI) Nutritional Anemia

955 (845, 1066) 63.6% (61.0%, 66.2%) ^↨^ Isolated iron deficiency anemia495 (456, 533)51.8% (48.4%, 55.1%) Isolated Vit B12 deficiency anemia120 (98, 142)12.6% (10.3%, 14.8%) Isolated folate deficiency anemia90 (71, 110)9.4% (7.5%, 11.4%) Multiple nutritional deficiency anemia250 (220, 281)26.2% (23.3%, 29.2%) Non-nutritional anemia 547 (457,** 636)****36.4% (33.8%**,39.0%) ^↨^ Anemia of chronic disease/inflammation216 (187, 246)39.5% (35.2%, 44.1%) Anemia of chronic kidney disease53 (39, 67)9.7% (7.1%, 12.2%) Multiple non-nutritional causes anemia132 (111, 151)24.1% (20.2%, 27.6%) Unexplained non-nutritional anemia146 (121, 172)26.7% (22.7%, 30.9%)

Nutritional anemia was more common in women than men (68.0% vs. 58.2%; p = 0.001) (Table 4). Among those with nutritional anemia, isolated iron deficiency anemia was also substantially higher among women as compared to men (60.2% vs. 40.1%; p < 0.001) (Table 4). In contrast, isolated vitamin B12 deficiency and isolated folate deficiency were lower in women as compared to men (vitamin B12 deficiency: 9.7% vs. 16.6%; p = 0.04 and folate deficiency 5.9% vs. 14.4%; p = 0.0008) (Table 4). However, the proportion of men and women with anemia due to multiple nutritional deficiencies was similar (Table 4). Non-nutritional anemia was less common among women than men (32.0% vs. 41.8%; p = 0.001). The percentages of CKD-associated anemia and anemia of chronic disease/inflammation were similar among men and women, while anemia due to multiple non-nutritional causes was more common among women (Table 4). The proportion of men with unexplained causes of non-nutritional anemia was substantially higher than that in women (30.2% vs. 22.9%; p = 0.0009) (Table 4).

Table 4. Sex-stratified distribution of different causes of anemia among those with anemia in LASI-DADAnemia typeMen (n = 682)Women (n = 820)P- value Nutritional anemia

397 (58.2%)

558 (68.0%)

0.001  Isolated iron deficiency anemia159 (40.1%)336 (60.2%)< 0.0001 Isolated Vit B12 deficiency anemia66 (16.6%)54 (9.7%)0.04 Isolated folate deficiency anemia57 (14.4%)33 (5.9%)0.0008 Multiple nutritional deficiency anemia115 (29.0%)135 (24.2%)0.84 Non-nutritional anemia

285 (41.8%)

262 (32.0%)

0.001  Anemia of chronic disease/ inflammation117 (41.1%)101 (38.5%)0.01 Anemia of chronic kidney disease27 (9.5%)26 (9.9%)0.86 Multiple non-nutritional causes anemia55 (19.3%)75 (28.6%)0.02 Unexplained non-nutritional anemia86 (30.2%)60 (22.9%)0.0009Note: These are column percentages with 95% CI of the column percentages and a significance level of 0.05

The prevalence of anemia was significantly different across the 22 states where participants were enrolled (p < 0.0001), with the overall anemia prevalence being > 70% in Assam, West Bengal, Jharkhand, and Odisha, and the overall anemia prevalence being < 30% in Jammu & Kashmir and Haryana. In general, states in East/Northeast India (Region 4) had a substantially higher anemia prevalence (74.1%) than the northern states (Region 1), where the lowest anemia prevalence was observed in India (37.8%) (Table 2). A more detailed evaluation of the subtypes of anemia between the four regions showed that nutritional anemia was significantly lower in Region 4 than in other regions in India (51.6% vs. 69.8%-73.8%; p < 0.0001) (Table 5). Among those with nutritional anemia, isolated iron deficiency anemia was slightly lower among participants in Region 4, though this difference was not statistically significant. In contrast, the percentage of isolated folate deficiency among those with nutritional anemia was almost 2–3 times higher in Region 4 as compared to other regions (15.5% vs. 4.4%-7.2%; p = 0.0001) (Table 5). Region 4 also had a substantially higher proportion of non-nutritional anemia as compared to other regions (48.4% vs. 26.2%-30.2%; p < 0.0001) (Table 5). Though there were no differences in the distribution of non-nutritional anemia types among the regions, the number of unexplained cases of non-nutritional anemia was slightly higher in Region 4 as compared to other regions (32.1% vs. 18.8%-22.6%; p = 0.05) (Table 5).

Table 5. Region-stratified distribution of different causes of anemia among those with anemia in LASI-DADAnemia typeRegion 1 (n = 351)Region 2(n = 244)Region 3(n = 308)Region 4(n = 599)P- value Nutritional Anemia 250 (71.2%)180 (73.8%)215 (69.8%)309 (51.6%)< 0.0001 Isolated iron deficiency anemia70 (28.0%)43 (23.9%)66 (30.7%)70 (22.7%)0.35 Isolated Vit B12 deficiency anemia118 (47.2%)99 (55.0%)118 (54.9%)160 (51.8%)0.003 Isolated folate deficiency anemia18 (7.2%)8 (4.4%)15 (7.0%)48 (15.5%)0.0001 Multiple nutritional deficiency anemia44 (17.6%)30 (16.7%)16 (7.4%)31 (10.0%)0.19 Non-nutritional anemia 101 (28.8%)64 (26.2%)93 (30.2%)290 (48.4%)< 0.0001 Anemia of chronic disease/inflammation41 (40.6%)27 (42.2%)36 (38.7%)112 (38.6%)0.95 Anemia of chronic kidney disease11 (10.9%)6 (9.4%)8 (8.6%)29 (10.0%)0.90 Multiple non-nutritional causes anemia28 (27.7%)19 (29.7%)28 (30.1%)56 (19.3%)0.08 Unexplained non-nutritional causes21 (20.8%)12 (18.8%)21 (22.6%)93 (32.1%)0.05Notes: Percentages are column percentages, at a significance level of 0.05* Regions:Region 1: Jammu and Kashmir, Punjab, Uttarakhand, Haryana, Delhi, Rajasthan, Uttar Pradesh, BiharRegion 2: Maharashtra, Gujrat, and Madhya PradeshRegion 3: Andhra Pradesh, Karnataka, Kerala, Tamil Nadu, Pondicherry, TelanganaRegion 4: Assam, West Bengal, Jharkhand, Odisha, Chhattisgarh

Multivariate analyses showed that participant age, sex, and region of residence continue to be independently associated with anemia status.

Between waves 1 and 2, 236 participants developed incident anemia in wave 2. The distribution of the incident and prevalent anemia cases was very similar in wave 2, with incident anemia being similar among women and men (51.3% vs. 48.7%; p = 0.20). There were no substantial rural/urban differences in anemia incidence. Regions 2 and 3 contributed to 18.2%-20.8% of cases of incident anemia, while Region 4 contributed to 28.4% of cases. Region 1 contributed a third of all incident anemia cases (33.3%) (p < 0.0001). Among those with incident anemia, nutritional anemia was more common in women (63.5%) than men (47.8%) (p = 0.02) (See Supplementary Table 1, Additional File 1). There were no regional differences in the distribution of incident anemia in wave 2 (See Supplementary Table 2, Additional File 2).

Discussion

The LASI-DAD study confirms the high national prevalence of anemia among older adults aged 60 years and above in India. We observed that nearly half of the older adults had anemia, with the overall and nutritional anemia prevalences being higher among women. This is the first study to evaluate the causes of anemia nationally and at the regional level. We show that there are substantial regional differences in anemia prevalence, with a higher prevalence in Eastern/Northeastern states, which can be attributed to a higher prevalence of non-nutritional anemia in these states.

The finding that nearly half of the older adult population in India is anemic supports previous observations that have shown a high prevalence of anemia (68.3%) among older adults in India [2]. Previous studies on older adults in India have shown highly variable estimates for anemia prevalence, ranging from 20.6% to 96% [21, 22, 51–52]. However, most studies on the geriatric population in India report prevalence estimates of more than 50% (ranging from 57% to 96%) [15–21, 24–28, 24–35, 53–55] Some studies have reported anemia prevalence similar to ours ranging from 45.5% to 49.6% [23, 30, 31]. Heterogeneity in prevalence estimates can be attributed to the variety of laboratory methods used to estimate anemia, smaller sample size (n = 60 to n = 982), hospital/institution-based samples, limited participant inclusivity (studies with only male or female participants), a convenience sampling design and/or sampling limited to individual states within India that limit the comparability and extrapolation of their results to a nationally representative study. Our study addresses all these limitations as we employed a two-stage stratified sampling technique nested with the Longitudinal Aging Study of India (LASI) cohort which allowed us to provide nationally representative prevalence estimates of anemia among people aged 60 years and above in India. Furthermore, we used standardized automated methods in a central laboratory for hemoglobin estimation and measuring other relevant analytes to determine the causes of anemia for all venous blood samples collected across India. A particular strength of this study is the ability to classify the underlying causes of anemia in a nationally representative population. Many community-based studies have reported the prevalence of anemia; however, only a few have studied the underlying causes [20–22]. In contrast, most studies that determined the underlying causes of anemia were hospital-based [55–63]. Our findings are consistent with previous reports that indicate that nutritional deficiency is the most common cause of anemia among older adults [22, 60, 62]. Similar to the variability seen in overall anemia prevalence in previous studies, there is considerable heterogeneity in the estimates of nutritional anemia in published literature. Studies have reported that nutritional anemia contributed to 26.8% to 56.8% of anemia cases [21, 22, 48, 49, 59–66]. Differences in methodology that allowed the identification of limited nutritional deficiencies only, variations in diagnostic criteria, a smaller sample size (n = 70 to n = 100), and recruitment of participants with pre-existing medical conditions are possible explanations for the observed differences from previous studies [56, 57, 59, 60, 62, 63].

Among the nutritional anemias, the present study found that iron deficiency anemia was the most common type, followed by vitamin B12 and folate deficiency anemia. In addition, anemia due to multiple nutritional deficiencies was the second most common form of nutritional anemia. These results are all largely consistent with previous studies that have evaluated different causes of nutritional anemia [55–62]. However, as stated previously, there is considerable heterogeneity in these estimates from previous studies due to small sample sizes [56–60, 62], hospital-based sampling [55–62], differences in criteria used to define specific nutritional deficiencies, such as using serum ferritin levels < 15 ng/mL to define isolated iron deficiency anemia [22], which is lower than the cut-off (< 30 ng/mL) used in our analysis and the lack of comprehensive evaluation of all nutritional deficiencies assessed in the studies [20, 21, 55, 59, 63]. An additional source of heterogeneity when identifying different types of nutritional anemia was using red blood cell (RBC) indices such as mean corpuscular volume (MCV) and RBC size to classify nutritional anemia [56, 63]. Given the high proportion of multiple nutritional deficiencies in the LASI-DAD population with anemia, the use of RBC indices such as RBC size and MCV to classify anemias resulted in an inability to classify a majority of individuals with anemia (data not shown), and hence RBC indices were not used in the categorization of anemia in this study.

The proportion of cases attributed to non-nutritional anemia in this study was also similar to previously published estimates [58, 60–62]. While other studies have published higher estimates of non-nutritional anemia, these studies were hospital-based studies where one would expect a higher prevalence of anemia due to pre-existing chronic diseases/inflammation among the recruited participants [55–57, 59, 63]. Anemia of chronic disease/inflammation was the most common cause of non-nutritional anemia and contributed to 40% of the cases. Previous community-based studies that have reported anemia prevalence have not comprehensively evaluated non-nutritional causes of anemia. However, the estimates of anemia of chronic disease/inflammation reported by hospital-based studies vary from our findings to some extent (21% to 56%)[55–63]. Our study found that 3.5% of all anemia cases were related to CKD. In other studies, CKD accounted for 11–20% of all anemia cases [56, 58, 60, 61]. The higher proportion of anemia related to chronic disease/inflammation and CKD in previous studies may be attributed to the likelihood of pre-existing chronic disease/inflammatory and risk factors such as diabetes, hypertension, and cardiovascular diseases among the participants recruited in hospital-based studies as compared to those in the community. Our study was able to categorize 90.3% of all anemia cases. Our findings are consistent with a few previous studies that reported a prevalence of unexplained anemia ranging from 8–11% [55, 57–61]. Though two previous studies reported a higher percentage (20-21.6%) of unexplained anemia [56, 62]), the comprehensive biochemical and hematological assessments used in LASI-DAD allowed us to identify a vast majority of the causes of anemia in this study.

Though the prevalence of anemia in our study was higher in rural areas than in urban areas, the difference was not statistically significant. These results are consistent with previous studies that have reported a similar trend [33, 35]. The most surprising finding in this study was the high prevalence of anemia in Region 4 (Assam, West Bengal, Jharkhand, Odisha, and Chhattisgarh) as compared to other regions in India. This finding is consistent with another study using data of 61,481 men from rural areas aged 15–54 years from the National Family Health Survey (NFHS)-5 [40], and showed that the districts with high anemia prevalence were from the eastern region including West Bengal, Odisha, Chhattisgarh, and Assam. However, NFHS also showed that the union territories of Jammu & Kashmir and Ladakh also had a high prevalence of anemia [40]. The study also reported the lowest prevalences from the northern states (Uttarakhand and Himachal Pradesh), some districts of north-eastern states (Nagaland and Mizoram), and southern states (Tamil Nadu and Karnataka). In contrast to the findings from our study, another study, which utilized the LASI data and asked people to self-report diagnosis of anemia within the last 2 years found [66] anemia prevalence to be highest in Region 1 (33.4%) followed by Region 4 (23.24%). However, since the prevalence estimate in the LASI study was based on self-reported cases of anemia, which are very sensitive to recall bias, knowledge and awareness regarding anemia, healthcare utilization, and completion of anemia treatment. In addition, the higher prevalence of self-reported anemia in LASI in Region 1 is likely biased as regions with relatively better healthcare facilities have easier accessibility to screening tests than others. Additional studies [27, 54, 59] from the eastern region of India have reported anemia prevalence ranging from 65% to 89.5%, which is consistent with the estimates from our study. Additional analysis of the distribution of different types of anemia in the four regions showed that non-nutritional anemia was significantly higher in Region 4 as compared to other regions in India. The reason for this finding is unclear at this time. Analysis of the prevalence of self-reported stroke, heart disease, and chronic kidney disease did not show any meaningful differences between the regions. Notably, the proportion of individuals with unexplained anemia was significantly higher in Region 4 (32.1%; p = 0.05) as compared to other regions (18.8%-22.6%). Given the higher prevalence of hemoglobinopathies in Eastern/North Eastern parts of India as compared to other regions [67], future studies that evaluate the prevalence of hemoglobinopathies in different states within India may provide additional insights into the possible reasons for the higher prevalence of anemia in Eastern/Northeastern states in India. The observed geographical differences can potentially have major implications for anemia control in India. Though the focus on reducing iron deficiency anemia is a nationwide goal and is supported by data from LASI-DAD and other studies, there may be an additional need to address the different causes of anemia using specific strategies in individual states to reduce anemia prevalence.

In wave 1 of LASI-DAD, the prevalence of anemia among older adults (n = 2758) was found to be about 40% [68]. The prevalence increased to 49.9% over the years among the LASI-DAD study participants. Incident anemia cases followed a pattern that was very similar to that observed with prevalent anemia with nutritional anemia (and iron deficiency anemia) being the most frequent cause of anemia.

Major strengths of this study include the nationally representative sampling design, high quality of the samples assayed, standardization of protocols, and representation of the geriatric population from both rural and urban communities on a national level. However, our study has certain limitations. Though we were able to identify iron deficiency anemia as the major cause of anemia in this population, this study could not identify the potential sources of iron deficiency. Insufficient dietary intake, occult blood loss due to malignancy or parasitic infections, and malabsorption are frequently observed among older adults. A major limitation of the LASI-DAD study was that we did not collect additional information to identify the causes of IDA in the study population. About 10% of the anemia cases were due to unexplained causes. One of the major causes that was not evaluated in this study was the contribution of hemoglobinopathies to anemia in this population. A recent study in North India showed that the prevalence of hemoglobinopathies was 8.5% and was almost three times higher in men (18.9%) as compared to women (6.9%). Therefore, evaluating hemoglobinopathies in this population may help explain the unexplained causes of anemia.

Conclusion

Our study findings have shown that approximately 50% of older adults are suffering from anemia, with nutritional deficiencies accounting for two-thirds of all anemia cases. This is a rapidly growing public health concern, especially for those with iron deficiency, anemia of chronic disease, and multiple underlying causes. The risk of anemia increases with age and the two waves of LASI-DAD show that the overall prevalence of anemia increased by 10%. Health policies to curb the anemia epidemic in India should also include older adults as a highly vulnerable group, besides children, adolescents, and women in the reproductive age group. India is experiencing a rapid epidemiological transition in disease patterns and will see an exponential increase in the population size of older adults in the coming decades [69]. The study results highlight the need to expand national anemia control programs, such as the Anemia Mukt Bharat, to include anemia control in older adults as well and/or integrate it into the National Program for the Health Care of the Elderly (NPHCE).

In addition to national anemia control programs focused on older adults, given the regional differences in prevalence and causes of anemia, the results of this study suggest that state-level anemia screening and treatment programs may be most effective in reducing the burden of anemia among older adults. Our study highlights the high overall prevalence of cases in the states of Eastern and North-Eastern regions, with a significantly higher proportion of anemia cases in these regions being attributed to non-nutritional causes. Specifically, these areas showed a higher prevalence of unexplained anemia, suggesting that factors such as hemoglobinopathies may account for a greater proportion of anemia in these states. If confirmed in other studies, this finding would suggest that targeted screening for hemoglobinopathies may be more effective in reducing anemia prevalence in Region 4 rather than a focus on nutritional anemia alone. The lower overall prevalence of anemia in Region 1 (North India) is an encouraging sign. Though the relative proportion of nutritional and non-nutritional causes of anemia was similar in Region 1 compared to Regions 2 and 3, the lower overall anemia prevalence in Region 1 suggests that national efforts to reduce anemia may be more effective in Region 1 than in other regions. A detailed understanding of the factors resulting in lower anemia prevalence in Region 1 may help design and implement effective anemia control programs in the rest of the county. This study highlights the importance of identifying and treating the specific causes of anemia that differ across states in India. This study supports that continued focus on nutritional anemia in general and particularly iron deficiency anemia in anemia control programs, as these are the dominant causes of anemia in almost all regions in India. However, adding additional region-specific components to the anemia prevention program may improve the program’s effectiveness.

In our study, 10% of the anemia cases were unexplained due to data limitations. It indicates the need for conducting pilot screenings for hemoglobinopathies, especially in high-prevalence states and regions, to provide a comprehensive map of other causes of anemia. The lack of serum ferritin measurements in LASI-DAD wave 1 did not allow us to estimate the change in percentage of nutritional and non-nutritional anemia across the two waves. Through our study, we demonstrate the feasibility of gathering high-quality, blood-based biomarker data through a logistically strong home-based collection and meticulous quality assurance. Overall, our study highlights the need to design health programs that are inclusive of the healthcare needs of older adults and are well-equipped to cater to the underserved aging population of India.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1