Should the anemia cut-off in India be revised?
Anemia prevalence has increased in India. A new study suggests that India should revise the cutoff of normal hemoglobin level.
Anemia has been a hot topic of discussion lately, with the government excluding anemia assessment from the NFHS-6 survey. The methodology of the NFHS has been blamed for the incorrect assessment of anemia. It has also been argued that the cut-offs used for estimating anemia need to be revised as the WHO cutoffs are based on a predominantly white adult population. Based on the data from the Comprehensive National Nutrition Survey (CNNS) 2016-18, a study published in The Lancet Global Health suggested that the cutoff for anemia for children and adolescents in India should be 1-2 g/dl lower. Using the revised cutoff, the prevalence of anemia would drop to 10.8% from 30% as per WHO cutoffs. This raises two important questions - Is NFHS overestimating anemia? Are the findings of the Lancet paper reliable?
NFHS and Anemia
NFHS uses capillary blood for hemoglobin assessment. Capillary blood can give a falsely lower value of hemoglobin and thus overestimate the prevalence of anemia. A study conducted in Uttar Pradesh showed that in non-pregnant women capillary blood samples overestimated anemia compared to the venous samples (59.2% vs 35.2%). Mild and moderate categories of anemia were reported to be overestimated when using capillary blood to assess anemia. However, there are studies reporting the reverse finding too. A study conducted on 60 adults in San Francisco reported capillary blood to be marginally higher than venous blood.
We still lack high-level evidence (systematic review or meta-analysis) to help us decide which method of blood sample collection leads to a higher estimation of anemia. While the debate is still on, it is important to note that the blood collection method was the same in NFHS-4 and NFHS-5. Therefore, the rising prevalence of anemia can not be blamed on the method of sample collection. Even if it is being overestimated, a rising trend means the current interventions need to be reassessed.
Do we need a separate hemoglobin cut-off level for Indians?
The currently used anemia cut-offs were published in 1968 by a WHO study group on nutritional anemia. The cut-offs were based on five studies performed on adult white populations in Europe and North America. Data from other countries, races, and ages (ie, infants, young children, adolescents, and elderly people) were not available to the panel. Cut-offs for mild, moderate, and severe anemia were defined later for children less than five years of age, children between 5-11 years of age, children 12-14 years of age, pregnant women, non-pregnant women (15 years of age and above) and men (15 years of age and above). These cut-offs have not been revised since then and have been included in numerous subsequent WHO publications and are being followed globally.
A study done in Indonesia found the WHO cutoffs to be valid for the Indonesian population. Another study carried out in 8 countries - Brazil, China, India, Italy, Kenya, Oman, United Kingdom, and the United States - found the normal hemoglobin level in pregnant women to be similar to the normal value defined by WHO for pregnant women. The paper ended by concluding that “population-specific ranges for maternal Hb in pregnancy are not required.”
However, considering the small sample sizes of these studies, the findings can not be generalized. A re-examination of the existing hemoglobin cutoffs is warranted in representative healthy populations of children and adults in India.
The Lancet study on anemia - How reliable are the findings?
A study published in the Lancet by Sachdev et al proposed the re-examination of hemoglobin cut-off values that have been given by WHO to define anemia. The study analyzes age and sex-specific percentiles of hemoglobin and reports new cut-offs to define anemia by using data from the Comprehensive National Nutrition Survey (CNNS). The CNNS was a large-scale nationally representative survey conducted in 30 states of India during 2016–2018. It included three population groups – preschoolers (0–4 years), school-age children (5–9 years), and adolescents (10–19 years) – from rural and urban areas. The survey included nearly 38000 children aged 0–4 years and 5–9 years each, and nearly 36000 adolescents aged 10–19 years.
The Lancet study used data from this survey and followed strict exclusion criteria to identify a healthy population for hemoglobin estimation. Participants without iron, folate, vitamin B12, and vitamin A deficiencies were included. They excluded participants with inflammation, abnormal hemoglobins like hemoglobin A2 and hemoglobin S and a history of smoking. Children with low albumin, parasitic infection, and zinc deficiency were also excluded to create separate subgroups for analysis.
The participants finally included were considered healthy. The hemoglobin values obtained for this supposedly healthy population were then compared with existing WHO cutoffs for different age groups of males and females up to 19 years of age. The cutoffs identified in this study to define anemia were found to be lower at all ages by 1–2 g/dL as compared to WHO cutoffs. As the CNNS questionnaire and individual-level data are not available in the public domain, it becomes difficult to blindly believe these findings. However, data presented in the Lancet study shows that the population analyzed to estimate normal hemoglobin was not actually healthy.
The authors mention “Children in the primary sample were undersized compared with WHO standards, with means ranging from −0·8 Z score to −1·2 Z score for various anthropometric indices.”. Mild malnutrition is defined as a difference of more than - 1 Z score, moderate as -1 to -2 Z scores, and a difference of more than -3 Z scores signifies severe malnutrition. Therefore, the children assumed to be healthy were actually underweight (wasted) and shorter for their age (stunted) when compared to healthy standards. The same is noticed in the final sample that was analyzed. (as seen in Supplementary File Table 1)
It is known that anemia leads to poor growth in children. Anemia and malnutrition are known to be associated with each other as seen in studies conducted in middle and low-income countries. Therefore can a population of malnourished children be considered healthy and their hemoglobin level normal?
Can a hemoglobin of 9 g/dl be normal?
Even if we choose to ignore the available evidence of the impact of low hemoglobin levels on pregnant women and children, there is ample evidence to show how low hemoglobin negatively affects adults. It is important to note that a child or adolescent with a hemoglobin of 9 g/dl will continue to have a similar level of hemoglobin as an adult if left untreated.
Besides being an independent risk factor for heart failure, low hemoglobin in adults with heart failure (less than 13 g/dl for males and 12 g/dl for females) is associated with a higher risk of death. Similarly, hemoglobin levels of less than 11g/dL lead to an increase in hospitalization and death in patients with chronic kidney disease. Low hemoglobin (less than 11g/dl) is also associated with poor quality of life and impacts disease outcomes in patients with cancer.
Not just chronic diseases, anemia can also have a poor prognosis when it coexists with infectious diseases. It not only makes individuals susceptible to infections, but it can also increase the risk of morbidity and death in patients with infections. A study published in Nature, also showed that anemia, defined as per WHO cutoffs, significantly increased the risk of death in patients admitted with COVID-19 infection.
The extent of symptoms seen with anemia depends on its severity. While mild anemia (less than normal but more than 10 g/dl) often has no overt manifestations, it has been seen to affect aerobic fitness and cognitive function. This is especially significant in younger populations, who have to live longer with uncorrected mild anemia and operate at a recognizably lower level of mental and physical ability.
Implications of revising the hemoglobin cut-off based on the Lancet study
Changing the cut-off will drop the prevalence of anemia in India and give a false idea of the health status of the country. The Anemia Mukt Bharat program would then become obsolete as anemia will no longer be a ‘severe public health problem’. The existing interventions under the program would then be phased out or de-funded leading to the worsening of the problem of anemia which will go unnoticed due to a revised cutoff.
Furthermore, conditions like sickle cell anemia and thalassemia will be missed as the new cut-offs will not warrant further investigations to identify the cause of what previously would have been labeled as anemia. This could lead to a delay in the diagnosis and treatment of these diseases, thus, increasing the complications associated with them. People who previously fell in the mild to moderate category of anemia will then be considered normal. This will only lead to an unhealthier population as anemia is a risk factor for various diseases ranging from depression to heart failure.
The authors of the Lancet study believe that changing cutoffs would enhance “the management of anemia at the individual and population level.” However, we fear that it will only affect the poor and the vulnerable population who fall under the ambit of population-level care through national health programs and policies. The rising burden of anemia should be tackled by generating good-quality evidence through research. The push from government officials to change the cutoff along with the decision of excluding anemia from NFHS-6 points more towards an attempt to improve just the perception of health and not the actual health of people in the country.
About the author:
Divya Shrinivas is the managing editor at Nivarana.
This article was originally published on Nivarana.org