New Study Shows Life Expectancy At Birth Shrunk By 2.6 Years In 2020

The study, based on NFHS-5 data, showed greater loss of life expectancy among marginalised groups than upper caste Hindus

By :  Menaka Rao
Update: 2024-07-20 10:13 GMT

A new paper published in the journal Science Advances said that life expectancy at birth shrunk by 2.6 years in 2020, the first year of the COVID-19 pandemic.

This study is led by Sangita Vyas and Aashish Gupta. Vyas is an economist, demographer, and Assistant Professor at CUNY Hunter College in New York, while Gupta is a sociologist and Marie Sklodowska-Curie fellow at Oxford University.

The study, using the latest National Family Health Survey (NFHS 2019-’21), pegged India’s mortality during the COVID-19 pandemic’s first year higher than the official figures as well as the World Health Organisation estimates. It has recorded an estimated 11.9 lakh (1.19 million) excess deaths during 2020, eight times more than official figures and 1.5 times more than the WHO’s estimates, which were rejected by the Indian government .

The study emphasised that mortality increased among younger ages groups, especially female children, signifying reasons other than COVID-19. As per the study, the loss of life expectancy in females was a year more than males. It also studied the effect of caste and religion, and estimated a higher decline of life expectancy for Muslims (5.4 years), Scheduled Tribes (4.1 years) and Scheduled Castes (2.7 years) as compared to upper caste Hindus (1.3 years).

IndiaSpend spoke to Gupta and Vyas about their findings. Excerpts:

Your paper says that there is a decline in life expectancy of 2.6 years between 2019 to 2021 - from 69.1 to 66.6. 2020 is the first year of the Covid-19 pandemic. What does a decline of life expectancy signify?

Sangita Vyas: Life expectancy is a summary measure of mortality in a particular year. So if someone were exposed to mortality rates in 2019, they would be expected to live a certain number of years. In 2020, because mortality rates increased, our paper shows that someone exposed to 2020 mortality rates would be expected to live 2.6 years less.

In this paper, we're not able to identify Covid deaths by themselves. We are looking at changes in mortality from any cause. This 2.6 decline is coming about by an increase in all-cause mortality. The 2.6 year decline is larger than the decline in life expectancy in any other high-income country in the same period.

Are there any other lower and middle income countries we can compare this statistic to?

Sangita Vyas: India’s life expectancy decline is similar to or larger than declines seen in the same period in other large lower and middle income countries including Brazil, Russia and Mexico. There is not really much that's known about mortality in many lower or middle income countries like India, as most of these countries do not have a complete civil registration system. Not all births and deaths are recorded, and it is really hard to know in a comprehensive way.

How are the Indian life expectancy numbers in your paper comparable to high income countries?

Sangita Vyas: In high income countries, most of the mortality increase was among older people, as compared to middle aged persons or children. But in India we see a different pattern. We see mortality increasing at all ages, and not just for the older people. So what that actually suggests to us is that it's probably not just Covid mortality that we see reflected.

Probably the reason why we see mortality increases for children in India is because of the other effects of what happened because of the severe lockdown. There were changes to people's livelihoods. The entire health system was disrupted. which meant that many children were not not born in hospitals. So there were all these disruptions that had something to do with the increases in mortality that we see for the younger ages.

Are these mortality figures that show deaths across ages also indicative of maybe a less robust health system? During the Covid-19 pandemic, it was to be expected that older people would die in larger numbers, but the spike in deaths in the younger age groups is perhaps surprising?

Sangita Vyas: If what we know about the actual disease is correct, based on the research done in richer countries with a different population, we wouldn't see much of an increase in mortality in children. The mortality among children is likely not from Covid but probably caused by other impacts that were going on during the lockdown.

So far, the research on mortality during the COVID-19 pandemic focused on excess deaths, or death rates.

Aashish Gupta: Excess mortality compares death rates. While valuable, we should remember that death rates are not exactly comparable across place and time. For example, Kerala's death rate right now is higher than that of Uttar Pradesh. The reason is that Kerala has more older people. When you have a greater proportion of older people in your population, your death rate becomes high.

So what you can do is you can create age-standardised death rates. That's comparable across countries. The reason why demographers like life expectancy as a measure is because it gives an intuitive understanding of changes across place and across time. So what life expectancy does is that if a population experiences so and so age specific mortality rates, what would the average age of death be?

It is not a prediction and it is not a projection. It is just an estimate for a hypothetical cohort based on prevailing mortality conditions. Another feature of life expectancy is that it is more sensitive to mortality changes at younger ages than older ages.

Then you are also looking for a metric that is looking at factors other than Covid-19-related deaths?

Aashish Gupta: Yes, that is definitely the case. There is 17% change in mortality. But it matters more in India than other countries, because these mortality impacts were at younger ages. A 17% increase in mortality in the US would lead to a different calculation of life expectancy, and would lead to a lower calculation of life expectancy at birth. In India, the decline would be greater.

Do we have an understanding of how this particular disaster of Covid-19 pandemic is comparable to other disasters or wars, famines etc.

Aashish Gupta: We already know this is the biggest crisis in post-Independence India. 2021 (during the Delta wave) would be a bigger crisis, one would imagine. But the NFHS data does not allow us to calculate that period.

Women are generally more resilient and live longer. Their life expectancy is better than men. But the life expectancy declined more for women than men by one year. Why do you think the pandemic affected females more?



(Page 26 of study)


Sangita Vyas: So another unique feature of what we saw in India is that mortality increased by more and life expectancy declined by more for women as compared to men. This is completely in contrast to all global patterns in every country in the world. In most countries, life expectancy declined more for men than for women.

What this suggests is that it probably has something to do with gender inequality in India.This has been documented even before Covid. Research suggests that women are less likely to go to a doctor than men when they are sick. During the pandemic, they were probably less likely to go and get tested, less likely to seek health care when they were sick from Covid.

In the serosurveillance data, there is actually no difference in the prevalence of Covid between men and women. Among reported deaths of Covid-19, male deaths are more likely to be captured than female deaths. The indirect effects of the pandemic, like reduced maternal health care, and impacts on livelihoods may also have contributed to a greater decline for women compared to men.

Aashish Gupta: Also, India was one of the few places in the world where life expectancy for women was lower than for men until 1985. So given that history, it is perhaps not surprising that in times of crisis, female mortality would be higher.




Could you explain your paper’s estimates for all-India excess deaths during 2020? Your extrapolated estimate is eight times higher than the official Covid-19 deaths, and 1.5 times more than WHO’s extrapolated estimate. Why are your numbers so different?

Sangita Vyas: As per our paper, the mortality was 17% higher in the pandemic months of 2020 - that is, April through December of 2020, compared to April to December in 2019.

Remember that our data is coming from a set of regions that are representative of only one quarter of India's population. But if we extrapolate this to the whole population of India, there would have been 1.19 million excess deaths in 2020 nationally. Our estimate of all-India excess deaths is eight times the official number of Covid-19 excess deaths, because many deaths during that period were not registered and recorded as a Covid death.

Our estimates are also one and a half times the WHO's estimate of excess deaths in India. So, why are we getting something this is different from the WHO?. The NFHS carried out surveys among households between 2019 and 2021, which was disrupted by the pandemic. We are only using the set of households that were interviewed in 2021. The surveyors asked respondents in these households to report on deaths in the past four years, and from that we know who died in 2017, 2018, 2019 and who died in 2020.

When you compare that data to WHO in each month, we get similar data for every month except March, April and May of 2020. They are probably finding less mortality in these months because this is the lockdown period. WHO data is using registered deaths. The registrations system was disrupted during the lockdown period and fewer deaths were registered because offices were closed and people could not go there to register. Our data is not subject to the same bias, because we are using household reporting on deaths.

So, while we are calculating 1.9 lakhs excess deaths, the WHO estimates 832,000 deaths in 2020.

Your earlier studies also showed large disparities in life expectancy of lower castes and Muslims as compared to upper caste Hindus before the pandemic. This paper shows that these disparities have exacerbated during the pandemic. Relative to the decline in life expectancy of 1.3 years for high caste Hindus, the loss for Muslims was 5.4 years, for Scheduled Tribes was 4.1 years, and for scheduled castes was 2.7 years.

Sangita Vyas: We see declines that are larger for marginalised groups in India, that is dalits, adivasis and Muslims. In particular we see the largest decline among Muslims. The decline is so large among Muslims that in 2020, Muslims had the lowest life expectancy at birth among all of the groups. But in 2019, they had a life expectancy that was higher than Dalits. The relative rank changed a lot.

This is consistent with reports of increasing marginalisation of Muslims. Part of these declines could come from increasing marginalisation during the pandemic, or the differences in disruptions to livelihoods, economic status, health care access, exposures, or other factors. More research is needed to disentangle all of these and understand why we saw disparities increase during the pandemic.

We knew that Covid did not affect young children as much. Yet your paper shows that the observed excess mortality is higher than expected. Can you explain this further? And why do you think younger children are more affected in India as compared to higher income countries?

Sangita Vyas: The expected increase in mortality is based on observed international infection fatality rates and seroprevalence in India. This is using other people's research. We are seeing an observed increase in mortality that is greater than the expected increase in mortality, which means we are seeing statistically elevated morality for younger ages. This is a meaningful increase.



(Page 28 of study)


Aashish Gupta: I do not feel we have had the last word on Covid and children. In malnourished populations, there could be an impact. In high income countries where most of the research has been done on children, that segment was found to be less vulnerable.

Besides that, healthcare disruptions, immunisation disruptions, and a general decline of people’s ability to put food on the table and take care of their families would cause higher mortality. We also found out that 0-5 ages among girls contributes to higher child mortality.

We welcome feedback. Please write to respond@indiaspend.org. We reserve the right to edit responses for language and grammar.

Tags:    

Similar News