ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846
-A A +A

Non-communicable Diseases, Affluence, and Gender

Pratima Yadav (pratimay7@gmail.com) is with the Institute for Human Development, Delhi. Veena S Kulkarni (vkulkarni@astate.edu) teaches at the Department of Criminology, Sociology and Geography, Arkansas State University, United States. Vani S Kulkarni (vaniskulkarni2001@gmail.com) teaches at the Department of Sociology, University of Pennsylvania,
United States. Raghav Gaiha (raghavdasgaiha@gmail.com) is (Hon) Professorial Fellow, the Global Development Institute, University of Manchester, England.

Whether the burden of non-communicable diseases has shifted to older men and women, and whether it varies by marital status and affluence of the household has been examined. The analysis is based on the 60th and 71st rounds of the National Sample Survey for 2004 and 2014. Even though comparisons of prevalence and shares between men and women are relied on in the analysis, some glaring disparities emerge. The growing menace of NCDs in the context of a rapidly increasing older population calls for bold policy initiatives, which are currently either underfunded or limited in coverage and uncoordinated. A drastic overhaul of the health system and behavioural changes are thus emphasised.

The authors are grateful to the anonymous reviewer, Jere Behrman, and N Chandramohan for their constructive suggestions. The views expressed are those of the authors.
 

A recent Lancet study (Jan et al 2018: 2047) draws attention to the devastation that “non-communicable diseases (NCDs) wreak and are likely to if timely action to prevent its continuing rapid rise is not taken.” Old-age morbidity is a rapidly worsening curse in India. The swift descent of the elderly (60+ years) in India into NCDs (for example, cardiovascular diseases [CvDs], cancer, chronic respiratory diseases [CrDs], and diabetes) could have disastrous consequences in terms of impoverishment of families, excess mortality, lowering of investment, and consequent deceleration of economic growth. Indeed, the government has to deal simultaneously with the rising fiscal burden of NCDs and substantial burden of infectious diseases (Bloom et al 2014a, 2014b). As a recent Lancet report (Ghebreyesus 2018) points out, failure to devise a strategy and make timely investment now will jeopardise achievement of sustainable development goal 3 (SDG) and target 4 of a one-third reduction in premature mortality from NCDs by 2030.

Four NCDs (CvDs, cancer, CrDs, and diabetes) accounted for 42% of all deaths in India in 2010. These diseases contributed 22% of disability-adjusted life years (or DALYs—the combination of years lived with serious illness and those lost due to premature death) in India in 2010 (Bloom et al 2014b). So, the cost in terms of lives lost is horrendous. Besides, NCDs hamper growth in different ways. They reduce the supply of labour and redirect resources from productive investments to healthcare, and thus drain the public and private budgets, raise business costs, and undermine competitiveness. In fact, based on the Environmental Policy Integrated Climate (EPIC) model of the World Health Organization (WHO), the potential cumulative losses to India’s economy during 2012–30 are projected to be $6 trillion (Bloom et al 2014b), nearly thrice of India’s gross domestic product (GDP) in 2017.1

The present study addresses the following questions: (i) Has the burden of NCDs shifted to old men and women? (ii) Does the burden vary by their marital status? (iii) Does it vary with the affluence of men and women? An important point to note is that there is no detailed analysis of burden of NCDs by gender. So our study is the first of its kind.

Literature Review2

The broad theme of ageing and morbidity in a global context has been the subject of several studies, of which the review by Prince et al (2015) is perhaps one of the most comprehensive and insightful. Drawing upon a large body of data, they offer a rich empirical account. Focusing on ageing and NCDs, they emphasise that the worldwide epidemic of chronic diseases is driven by population ageing. Disorders with a strong age-dependent relation are likely to increase in prevalence in parallel with the absolute and relative numbers (relative to the total population size) of the older people (≥60 years).3

Another major contribution is “The Lancet Taskforce on NCDs and Economics” (Lancet 2018) with several individual contributions on different aspects of NCDs. How households cope with the high costs of treatment associated with NCDs remains a major concern. For all health conditions investigated, NCDs are associated with substantial economic burden on patients and their households from all strata, particularly in the poorest populations. Direct medical expenses for medicines, outpatient visits, diagnostics, and hospitalisation are the main contributors to out-of-pocket (OOP) costs. Transport costs are substantial in some cases (for example, about 40% of total medical expenditure is spent on patients receiving a kidney transplant in India) (Jan et al 2018).

Health insurance has a role to play in protecting individuals and households from catastrophic expenditure associated with NCDs and, in doing so, facilitating access to healthcare. However, the protection offered by insurance is far from adequate. Moreover, over 84% of India’s population is not covered under any health insurance scheme (Yadav 2018).

Turning to studies on India, an informative study of pattern of morbidity among the elderly is Yadav et al (2017). Although the data used is not so recent, it offers a detailed account of NCDs by age. The morbidity analysis is based on Study on Global Ageing and Adult Health (SAGE) 2007. Binary logistic models are applied to obtain the odds of different types of morbidity among the elderly population in urban India. The dependent variables in the model are chronic lung diseases, diabetes, depression, cataracts, arthritis, stroke, angina, asthma, and oral health.

The odds of having diabetes, for example, are 2.1 times higher in the age group of 60–69 years, compared to the age group of 50–59 years. Those consuming alcohol are 1.9 times more vulnerable to diabetes, relative to those who do not. Those with 10 years or more of education are 2.41 times more likely to suffer from diabetes compared to illiterates. Those belonging to castes other than Scheduled Castes (SCs) and Scheduled Tribes (STs) are 2.1 times more likely to be victims of diabetes. The odds of diabetes are 2.5 times more among individuals in the middle and richer wealth quintiles, and 2.7 times more among the richest wealth quintile than among the poorest quintile (Yadav et al 2017).

The older age group of 70–79 years is more prone to chronic lung disease than the age group of 50–59 years. A person consuming alcohol also is more vulnerable to this disease (the odds being 2.3 times higher) than someone who does not. Besides, tobacco consumption increases the odds of this disease by 2.03 times (Yadav et al 2017).4  Updates of NCD burden and risk exposure in different states in India in the context of epidemiological transition are given in a Lancet study (India State-Level Disease Burden Initiative Collaborators 2017; ICMR et al 2017). Our review is confined to NCDs and risk exposure during the period 1990–2016.

The all-age prevalence of most leading NCDs increased substantially in India from 1990 to 2016, but the age-standardised prevalence increased only for diabetes, cerebrovascular disease, ischaemic heart disease (IHD), and skin diseases. This suggests that the overall increase in NCD prevalence has been a mixed phenomenon, with ageing of the population being a significant contributor together with additional increases due to changes in risk exposure for the causes that have an age-standardised increase in prevalence.

The major risk factors for IHD, cerebrovascular disease, and diabetes have been rising across epidemiological transition level (ETL) groups on the basis of the ratio of DALYs from communicable, maternal, neonatal, and nutritional diseases (CMNNDs) to those from NCDs and injuries combined in 2016. Dietary risks, high systolic blood pressure, high fasting plasma glucose, high total cholesterol, and high body mass index together account for a quarter of the DALYs in India in 2016, which is more than twice their share in 1990. Exposure to air pollution in India is among the highest in the world, contributing to both NCDs and communicable diseases.

Salient Features of National Sample Surveys

Our study is based on the NSS household/individual data on health and morbidity for 2004 and 2014. Both NSS rounds (60th and 71st) follow a stratified multistage design and first stage units (FSUs), which are villages in rural areas and blocks in the urban sector. FSUs have been selected following probabilities proportional to size with replacement (PPSWR) technique. Further, both rural and urban FSU samples are drawn as two independent subsamples and equal numbers of samples are allocated to both areas. The households are the ultimate stage units in both the sectors.

Both communicable diseases (for example, malaria, tuberculosis or TB, and HIV/AIDS) and NCDs (for example, heart diseases, hypertension, respiratory diseases, diabetes, and cancer) are covered by outpatient and inpatient illness reporting. As other salient features of the NSS are well known, no further elaboration is necessary.5

Analysis

As a detailed econometric analysis is reported in Yadav et al (2018), here we confine ourselves to a comparison of mean differences in the burden of selected NCDs between 2004 and 2014 by gender, using the t-test, and changes in shares of these NCDs and their covariates (for example, age, rural and urban, marital status, caste, and affluence) by gender. It is emphasised that no causal inferences can be drawn from these comparisons. However, some useful insights emerge, with the caveat that confounding variables could vitiate some comparisons.

 

All non-communicable diseases: Let us first consider the results for all NCDs for men and women in Table 1 (p 47).

In sharp contrast to women who recorded a significant rise, overall prevalence of NCDs among men fell significantly during 2004–14. As in the case of women, the highest prevalence of NCDs was observed among men in the older age group, 70–79 years, in 2004. Also, the highest prevalence moved to the oldest men and women, 80 years+, in 2014. The burden of NCDs shifted starkly from men ≤ 60 years to those ≥ 60 years, but the majority was still accounted for by men ≤ 60 in both 2004 and 2014, with a large diminution in the more recent year. This mirrors a similar shift among women. Thus, the burden of NCDs shifted to the old.

Men recorded significant reductions in prevalence in each age group of ≤ 50 years during 2004–14, but women did only in two younger groups, 0–20 years and 21–30 years. In all older age groups (above 50 years), men recorded significantly higher prevalence rates in 2014, while women did so in all age groups above 30 years.

Both men and women in rural areas were the majority in total NCDs in 2004 and 2014, with reductions in their shares in 2014. The mean prevalence of NCDs was higher for men in urban areas as compared to those in rural areas in 2004 and 2014. It was also higher relative to urban women in 2004. But, there was a reversal in 2014, when prevalence among urban women was higher than among urban men.

In a striking contrast, while prevalence of NCDs among both rural and urban women rose significantly between 2004 and 2014, it fell significantly among both rural and urban men.

Considering men by marital status in 2004, we find that the highest prevalence of NCDs occurred among widowers and the divorced/separated. These rates were higher than among widows and divorced/separated women in 2004. But in 2014, the prevalence rates among widowed and divorced/separated men were lower than those among women from the same group. The majority of total NCDs was accounted for by currently married men, as also by currently married women in 2004, but with a lower share. While the total share of currently married men rose, that of currently married women declined in 2014.

There is a contrast between changes in prevalence rates of NCDs among men and women by marital status during 2004–14. The prevalence declined significantly among both never married and currently married men, while it rose significantly among currently married women and widows. One exception was that both never married men and women experienced significant reductions in NCD prevalence during this period.

Male graduates and above had the highest prevalence of all NCDs in 2004, much higher than that of female graduates. The former continued to have the highest prevalence in 2014, while among women, the highest prevalence moved to illiterates. The largest share in total NCDs, however, was that of males with middle-higher secondary education in both 2004 and 2014, with a small increase in the more recent year. Among women, the largest share in total NCDs was that of illiterates in both 2004 and 2014.

Mean prevalence of NCDs among men declined significantly in all educational categories, except among illiterates, during 2004–14, while among women, the prevalence rose among
illiterates and those with primary or lower education, but declined among those with middle-higher secondary education and graduation and above.

The caste hierarchy mirrors socio-economic deprivation with the STs at the bottom and the Others at the top. Men belonging to Others had the highest prevalence rate of NCDs, and SCs/STs the lowest in 2004. This pattern was reproduced in 2014, with reductions in prevalence. An interesting contrast between men and women is that the prevalence among the latter in Others was lower in 2004 but higher in 2014.

The highest concentration was of “Others” among men in 2004, which declined in 2014. The men from Other Backward Classes (OBCs) accounted for the largest share in 2014. The “Others” among women displayed the largest share in 2004, but were replaced by OBC women in 2014.

A comparison of mean prevalence between 2004 and 2014 reveals a striking contrast. Women in all caste categories experienced significant increases in prevalence rates, while men experienced significant reductions.

There was a significant affluence gradient to prevalence of NCDs among men, with a sharp increase in the prevalence from the lowest expenditure quintile to the highest in 2004. This is similar to what women experienced. A similar pattern is reproduced among both men and women in 2014, but with one reversal. While the prevalence among the most affluent men was higher than among the most affluent women in 2004, the latter recorded a higher prevalence in 2014.

Again, the contrast between men and women is striking. During 2004–14, men in each expenditure quintile recorded significant reductions in the prevalence rate of NCDs, while women recorded significant increases.

Hypertension: Turning to Table 2, we get some insights into how prevalence of hypertension has changed among women and men during 2004 and 2014, and selected covariates. The first important finding is that the prevalence of hypertension among both women and men rose significantly during this period.

Among men, the mean prevalence of hypertension rose significantly during 2004–14, as also among women. The highest prevalence occurred among old men, 60–69 years, in 2004 but shifted to the older group, 70–79 years, in 2014. The highest share of total hypertension cases6 was contributed by men in the age group of 31–50 years in 2004, but shifted to the old, 60–69 years, in 2014. This was part of a shift of the overall burden of hypertension from those below 60 years to those ≥ 60 years. This shift is similar to what women experienced.

As among urban women, the prevalence of hypertension among men in urban areas was higher than in rural areas and rose significantly during 2004–14. However, urban women were more prone to hypertension than urban men in both years. Urban men accounted for the majority in total cases of hypertension among males in 2004, but this ceased to be so in 2014, as the share of rural men was higher. As in the case of women, the mean prevalence among men was significantly higher in both urban and rural areas during 2004–14.

Going by marital status, unlike women, divorced/separated men experienced highest prevalence of hypertension in both 2004 and 2014. The highest concentration of hypertension cases occurred among currently married men, as also among currently married women. The two groups continued to have the highest concentration of hypertension cases in 2014, with a large increase among currently married men and a decrease among currently married women since 2004.

For never married men and never married women, the prevalence declined significantly from 2004 to 2014. However, among the currently married and widowed, the mean prevalence rose significantly during 2004–14. The mean prevalence among divorced/separated men and women did not change significantly during this period.

The highest prevalence among men was found in graduates in both 2004 and 2014 with a slight increase (not significant statistically). Among women, the highest prevalence was found among those with middle-higher secondary education in 2004, and it shifted to those with graduate and above education in 2014. The highest concentration of hypertensive men in total cases of hypertension was found among those with middle-higher secondary education in both 2004 and 2014. In sharp contrast, the highest share among women was that of illiterates in both years.

As among women, men in different educational categories experienced significant increases in the prevalence of hypertension during 2004–14. As among women, men in Others had the highest prevalence in 2004 and also the largest share of total cases of hypertension. However, in 2014, OBC men had the highest prevalence rate and the largest share of total hypertension cases. Among women, the highest prevalence remained among Others with a significant rise in 2014 but the concentration shifted from Others to OBCs. In both years, the prevalence among men was lower than among women in most cases except that OBC women had a lower prevalence than OBC men in 2014. The mean prevalence among men in different castes rose significantly, as among women, during 2004–14.

Going by our measure of affluence, a gradient is observed between different levels of affluence and prevalence of hypertension among both men and women in both 2004 and 2014. The lowest prevalence was among the least affluent/first quintile, and highest among the most affluent/fifth quintile in both years. The concentration of hypertension was also highest among both men and women in the most affluent/fifth quintile in both years.

The mean prevalence among both men and women rose significantly in each expenditure quintile during 2004–14. As in 2004, the mean prevalence of hypertension was higher among most affluent women compared with most affluent men in 2014.

Diabetes: As shown in Table 3, the overall prevalence of  diabetes rose significantly among both men and women during 2004–14, with the prevalence much higher among men in 2014 relative to women. Unlike women, highest prevalence of diabetes occurred among men in the age group of 51–59 years in both years. Among women, the highest prevalence occurred in the age group of 60–69 years in 2004 but shifted to the younger in the age group of 51–59 years in 2014. The highest concentration of male diabetics occurred in the age group of 31–50 years in 2004 but they were replaced by older men in the age group 60–69 years in 2014, though with a reduced share. As among women, a shift of the overall burden from men below 60 years to those older occurred in 2014. However, both among men and women, those below 60 years remained in majority.

As among women, urban men recorded higher prevalence than rural men in both 2004 and 2014. Urban men also accounted for the larger share of diabetics in both 2004 and 2014 with a slight increase in the share in the more recent year. Both rural and urban men, as also women in these locations, recorded significantly higher prevalence during 2004–14.

Gender differences by marital status reveal some striking differences. While currently married men had highest prevalence of diabetes in both 2004 and 2014, divorced/separated women had highest prevalence in both years. Indeed, their prevalence rates were markedly higher than those of currently married men. However, the concentration of diabetics was highest among currently married men in both 2004 and 2014, with a more-than-moderate increase in 2014. Although, concentration of female diabetics was also highest among currently married, the shares were markedly lower than those of male diabetics. Mean prevalence in each marital status rose significantly during 2004–14 except among divorced/separated men and women.

Both among men and women, the highest prevalence was among graduates and above in 2004 and 2014. Both prevalences rose significantly during this period. However, the highest share of men was found among middle-higher secondary during this period. But, the highest share of women was found among illiterates, which rose during this period. Both men and women in each educational category saw significant increases in the mean prevalence of diabetes during 2004–14.

There were some dissimilarities between men and women by caste too. The highest prevalence occurred among men belonging to Others in both 2004 and 2014. This contrasts with highest prevalence among women belonging to OBCs in both years. The concentrations also differed. While women belonging to OBCs were the majority in both years, with a small increase in 2014, men belonging to Others were the majority in 2004, but they were replaced by OBC men in 2014. However, a striking similarity was that both men and women belonging to different castes saw significant increases in the mean prevalence of diabetes during 2004–14.

As among women, there is a gradient between mean prevalence and expenditure quintiles in both 2004 and 2014. The least affluent men had the lowest prevalence and the most
affluent the highest prevalence in both years. The concentration was also highest among most affluent men and women in both 2004 and 2014. Both men and women recorded significant increases in the prevalence in each expenditure quintile during 2004–14, barring the lowest quintile.

Heart attack: There are some striking similarities and dissimilarities between men and women, as shown in Table 4. The prevalence of heart attack declined slightly among men, but significantly among women during 2004–14. The highest prevalence of heart attacks occurred among men in the age group of 51–59 years in both 2004 and 2014, but with a diminution in 2014. Among women, however, the highest prevalence was among the oldest in 2004, but it shifted to the youngest (0–20 years) in 2014. The highest concentration was among both men and women below 60 years in both 2004 and 2014, but with substantial reductions. Thus, there was a marked shift of the overall burden of heart attacks from those below 60 years to those 60 years and above during 2004–14.

The mean prevalence of heart attacks declined significantly in different age groups of men and women during 2004–14. Among men, it declined in the age group of 31–50 years, as also among women in this age group; it also declined among men in the older group of 60–69 years. In contrast to men, youngest women in the age group of 0–20 years recorded a significant increase.

Urban men were more vulnerable to heart attacks than rural men in both 2004 and 2014. They were also more vulnerable than rural and urban women in both years. While urban and rural men accounted for equal share of heart attacks in 2004, the rural share was higher in 2014. Among women, rural women accounted for the majority share in both years, with a higher majority in 2014.

While mean prevalence declined significantly among urban men, and rose significantly in rural men, it declined only among urban women during 2004–14. Men and women by marital status reveal a contrast. The highest prevalence of heart attacks was experienced by divorced/separated men in 2004, but shifted to currently married men in 2014. The highest concentration, however, was of currently married men in total cases of heart attacks in both 2004 and 2014, with a much higher concentration in 2014. Among women, the highest prevalence was shared by currently married and widows in 2004 but divorced/separated women had much higher prevalence in 2014. The majority share, however, was that of currently married women in both years with a diminution in 2014.

Currently married women and widows experienced a significant decline in the prevalence during 2004–14. However, prevalence of heart attack among never married women rose significantly during 2004–14 as also among divorced/separated (significant at ≤  10%). Never married and divorced/separated men experienced significant reductions.

Among men, the highest prevalence was among graduates and above in 2004, but it shifted to those with primary or lower education in 2014. Among women, highest prevalence was among those with middle–higher secondary education in 2004, but it shifted to primary and below in 2014. Among men, the highest share of total heart attacks (in respective groups) was among those with middle–higher secondary education in 2004, and it remained so in 2014, but among women the share of illiterates was highest in both years.

The mean prevalence rose significantly among illiterate men and those with primary or lower education (significant at ≤ 10%) while it declined among those with higher levels of education such as middle–higher secondary and graduation and above during 2004–14. Among women in these groups too, the prevalence reduced significantly.

In the caste hierarchy, men belonging to Others exhibited the highest prevalence of heart attacks in both 2004 and 2014, while among women it was so in 2004, but SCs/STs replaced Others with the highest prevalence in 2014. Men belonging to Others accounted for the largest share of total heart attacks among them in 2004, but OBCs became the dominant group in 2014. Women belonging to Others were the largest group in 2004, but they were replaced by OBC women in 2014. In 2004, OBC men experienced a significant increase in the prevalence of heart attacks, while Others recorded a reduction during 2004–14. Among women, while SCs/STs experienced a significant rise, OBCs and Others saw a significant reduction.

There is a gradient between prevalence of heart attack and affluence among men except that the lowest prevalence is observed in the third expenditure quintile in 2004. This gradient disappears in 2014 as the lowest and fourth quintiles have the same prevalence, but the highest prevalence is in the fifth quintile. The highest prevalence occurred among women in the third expenditure quintile in 2014, and the next highest among those in the first quintile. However, the largest concentrations were observed among the most affluent men and women in both 2004 and 2014. The mean prevalence changes varied between men and women during 2004–14. Among men, the prevalence rose significantly in the first quintile and declined in the fifth quintile, while among women it rose in the first three quintiles but declined in the fourth and fifth quintiles.

Cardiovascular diseases: Comparisons of prevalence and overall burden of CVDs between men and women during 2004–14 are given in Table 5.

The overall prevalence of CVDs rose significantly among both men and women during 2004–14. The highest prevalence was observed among men in the age group of 51–59 years in 2004, but it shifted to the older group of 70–79 years in 2014. Among women, the highest prevalence was among the oldest (80+ years) in both 2004 and 2014. The highest concentration of CVDs among men was in the age group of 31–50 years in 2004, but shifted to the much older group of 60–69 years in 2014. Among women, the highest concentration was in the age group of 31–50 years in 2004 and shifted to the older age group of 60–69 years in 2014, with a large reduction in the highest share in the more recent years. Men below 60 years bore a majority of the overall burden of CVDs in 2004, but their share reduced to well below 50% in 2014. The share of women below 60 years also fell during this ­period, with the difference that they barely retained their ­majority in 2014.

During 2004–14, the mean prevalence of CVDs significantly rose among men in all age groups except those between 0–20, 21–30, and 51–59 years old. Among women, it declined in the age group of 31–50 years, and rose in the age groups of 60–69 years, 70–79 years, and 80+ years during 2004–14. The mean prevalence among women in the age group of 51–59 years
remained unchanged.

Both urban and rural men accounted for equal shares in 2004, but the rural share was larger in 2014. Among women, while the rural share in total CVDs was in majority in 2004, the urban share became larger in 2014. While the mean prevalence of CVDs was significantly higher among both urban and rural men, it was higher among rural women and lower in urban women during 2004–14 (significant at ≤ 10%). The share of ­rural men was higher than that of urban men in 2014. The share of rural men rose in 2014, while that of urban men
declined. Rural women were in majority in 2004, but ceased to be so in 2014.

Turning to marital status, the highest prevalence was among divorced/separated men in 2004, replaced by widowers in 2014. Among women, currently married had the highest prevalence in 2004, but they were replaced by widows in 2014. Currently married men had the largest share of total males with CVDs in 2004, and it rose further in 2014. Currently married women reflected similar outcomes in 2004 and 2014.

Unlike divorced/separated men, both currently married men and widowers recorded significant increases in the prevalence, while those never married showed a significant reduction during 2004–14. Similarly, currently married women and widows recorded significant increases in the prevalence rate of CVDs during 2004–14.

Among men, graduates and above experienced highest prevalence of CVDs in both 2004 and 2014. In sharp contrast, women with middle-higher secondary education recorded highest prevalence in 2004, but they were surpassed by women with primary or lower education in 2014. Men with middle–higher secondary level of education accounted for the largest share of total CVDs in both 2004 and 2014, while women displayed a different outcome as illiterate women had the highest share in both the years.

The mean prevalence among men rose significantly in all educational categories except among graduates and above who recorded a significant reduction during 2004–14. There was a contrast among women as the mean prevalence rose significantly among illiterates, those with primary or lower education, and graduates and above, while it reduced significantly (at ≤ 5%) among those with middle–higher secondary education.

Using the caste affiliations, men belonging to Others recorded the highest prevalence of CVDs in 2004, but not in 2014 as OBCs recorded the highest prevalence. Among women, Others recorded the highest prevalence in both 2004 and 2014. Men belonging to Others recorded the largest share of total CVDs in 2004, but were replaced by OBC men in 2014. Among women, Others recorded the highest prevalence share in 2004 but this was replaced by OBCs in 2014.

The mean prevalence rose significantly among men belonging to SCs/STs and OBCs, while it rose significantly among women in all castes.

Both among men and women, there was a marked gradient between prevalence of CVDs and affluence measured in terms of expenditure quintiles. In other words, the least affluent had the lowest prevalence and the most affluent the highest in both 2004 and 2014. The shares of the most affluent men and women were also largest in both 2004 and 2014, with a slight reduction in the share of women in total CVDs in 2014.

The mean prevalence among men rose significantly in all expenditure quintiles as also among women in all quintiles except the fifth.

Discussion

We now discuss our major findings from a broader policy perspective. The present study provides detailed evidence on NCDs and their covariates by gender. This is particularly relevant in the present Indian context, as the elderly population ≥ 60 years is growing three times faster than the population as a whole (Agarwal et al 2016; Alam et al 2012; Alam and Yadav 2014).

The projected marked future shift in the share of older Indians in the population is taking place in the context of changing family relationships and severely limited old-age public ­income support, hence raising a variety of social, economic, and healthcare policy challenges (Beard and Bloom 2014; WHO 2015).

From this perspective, the shift in the overall burden of major NCDs such as diabetes, hypertension, heart attack, and CVDs, during 2004–14 from the young to the old (60+ years) raises an important policy concern. This shift is shared by both men and women. Continuing neglect and failure to anticipate these demographic and epidemiological shifts—from infectious diseases to NCDs—may result in enormously costlier policy challenges.

The overall prevalence of NCDs was higher among men in 2004, but lower than that among women in 2014. This gender disparity of course varied with specific NCDs. The prevalence of hypertension, for example, was higher among women than among men in both 2004 and 2014. This disparity is further revealed by marital status. Widowhood for women meant higher prevalence of hypertension than among men in both 2004 and 2014. In contrast, never married are better off than currently married. Thus, it follows that marriage is not necessarily a barrier against NCDs, especially women, as they are subject to discrimination in allocation of food and medical care. What are their exit options? Few, if any, as argued persuasively by Heath (2013). Below a threshold of bargaining power, assertion of autonomy by women for a fair allocation is likely to be counterproductive.

Social and family norms that restrict women’s access to healthcare are not as rigid as generally believed. Greater awareness of equity and better recognition of women’s contribution to household and social welfare could enhance their access to healthcare.

Our finding that both rural men and women were better off than urban men and women is somewhat intriguing as availability of and access to medical facilities are better in urban locations. However, our findings could be justified by widely acknowledged views of greater environmental pollution and unhealthy diets (for example, processed and fried foods) in urban areas more than offsetting the advantage of better medical facilities in urban areas. What could make matters worse for urban populations is their sedentary lifestyle. As rapid urbanisation of the population is irreversible, policies should stress the importance of healthy diets and more effective taxation of cigarettes/other tobacco products and alcohol in various forms, and better
regulatory mechanisms for limiting environmental pollution.7

Both among men and women, prevalence of all NCDs and specific NCDs (for example, diabetes) was highest among graduates and above (with a few exceptions such as CVDs) in both 2004 and 2014. However, both men and women with middle–higher secondary and graduation and above level of education experienced reductions in the prevalence of all NCDs and heart attacks during 2004–14, while in other cases (for example, hypertension and diabetes) there were increases (significant only among women). One important issue is that the better-educated also tend to be more affluent and thus more prone to sedentary lifestyles and unhealthy diets that offset their advantage in terms of easier affordability of expensive treatments for chronic conditions.

The caste hierarchy reflects different degrees of social exclusion and affluence. For example, while STs are least affluent and most isolated, Others are the most affluent and not isolated at all. So, while men belonging to SCs/STs displayed lowest prevalence of, say, all NCDs in both 2004 and 2014, women did so only in 2014. In contrast, both men and women belonging to Others displayed the highest prevalence in 2004, while women did so in both years. The lowest prevalence among SCs/STs reflects healthier lifestyles and diets and extensive use of indigenous knowledge systems, while the affluence of Others is characterised by sedentary lifestyles, unhealthy diets, and more frequent use of alcohol and cigarettes/tobacco that more than offset their advantage in terms of affordability of expensive treatments.

This is further highlighted by the affluence gradient to prevalence of all NCDs and hypertension, and with some variation in diabetes and CVDs among both men and women during 2004–14. In other words, prevalence was largely higher across more affluent quintiles in these years. While among men, the prevalence of all NCDs declined significantly, it rose significantly among women during 2004–14. As far as hypertension is concerned, both men and women recorded higher prevalence in each quintile during this period. Worse, the concentration of all NCDs, hypertension, diabetes, heart attacks, and CVDs remained highest among the most affluent/fifth quintile. This further strengthens the case for more physically active lifestyles and healthier diets.

Concluding Observations

The growing menace of NCDs in a context of rapidly increasing old population calls for bold policy initiatives. Although such initiatives are not lacking, they are either underfunded or limited in coverage and uncoordinated (Chatterjee 2017).

A National Health Policy was announced in 2017. It proposed raising public health expenditure progressively to 2.5% of the GDP by 2025 and advocated a major chunk of resources to primary healthcare, followed by secondary and tertiary healthcare. This policy and the National Institution for Transforming India (NITI Aayog) action agenda have set targets for reduction of premature deaths and morbidity due to major NCDs in India. There are two serious concerns, however. One is that scant attention is given to where the resources will come from. ­Another glaring omission is that little is said about the rapid rise in the share of the old in the total population and associated multi-morbidities of NCDs. In the context of declining family support and severely limited old-age income security, catastrophic consequences for destitute individuals afflicted with these conditions cannot be ruled out (Jan et al 2018).

Although NSS data do permit analysis of hospitalisation by NCDs, a few observations are in order. First, disaggregation of hospital care into inpatient and outpatient revealed a mixed pattern. While inpatient care reduced from under 6% of the population, outpatient care rose slightly from about 9% to under 11% during 2004–14. Second, as far as inpatient care is concerned, private hospitals overtook public hospitals and maintained their dominance but their share diminished. This domination became more glaring in urban areas. The domination of private hospital care among outpatients was far more pronounced as it catered to three-quarters of them in both 2004 and 2014. Commercialisation of hospital care fills an important gap but without insurance cover it is likely to impoverish those suffering from NCDs.8 As reported by Hooda (2015), private hospitals charge exorbitant fees for chronic conditions. In 2014, the cost of hospitalisation in a private facility was 4.2 times as compared to that in a public facility. In the same year, illustrative evidence shows that for cancer and CVDs, the average expenditure in a private hospital was close to about 3.5 to 4 times the corresponding expenditure in a public hospital (Hooda 2015). So, rapid expansion of private hospitals without appropriate regulatory mechanisms and comprehensive health insurance is likely to aggravate the risk of impoverishment.

A related question is coverage of health insurance.9 Both rural and urban populations experienced a surge in insurance during 2004–14: in rural populations, it spiked from 0.4% of the population to 14.1%; and in urban populations from 3.1% to 18.1%. However, a little over 15% of the all-India population was covered (Patel et al 2015). As NCDs typically require
extended hospitalisation in advanced stages, health insurance is vital for low-income households to be able to afford it.

Public health system must devote additional resources towards active population-based surveillance. Besides financing, there is a need to develop institutional mechanisms for engagement of adequate human resources for surveillance and disease management (Mishra et al 2016; Chatterjee 2017). Specifically, health systems need to be reorganised to better provide coordinated and informed geriatric services. Another shift required is patient technology to support self-management of conditions—especially for the old. Integration of care in creative ways such as treatment centres for multi-morbidity clusters is thus a priority (Lancet 2018a).

Behavioural changes are no less important and perhaps also no less challenging. Lack of physical activity and unbalanced high-calorie diet promote weight gain. The culprits are sugar and dairy fat. As of 2010, 15% of women and 12% of men were obese in India (Bloom et al 2014a). Obesity is a risk factor for CVDs and diabetes and can aggravate symptoms of chronic obstructive pulmonary disease such as emphysema and bronchitis.

Both alcohol consumption and smoking must be curbed. Taxation is one instrument. Given the prevailing tax system with widely varying taxes, it is imperative to streamline the system through the goods and services tax (GST)—a contentious issue. For example, alcohol is still not covered by the GST.

The higher the level of OOP costs for long-term therapy relative to the costs of other competing household needs, the lower is the incentive for individuals to adhere to treatment. Encouraging the individuals to prioritise spending on long-term treatment and prevention of NCDs is particularly challenging in resource-poor settings (Jan et al 2018).

The Government of India launched Ayushman Bharat—Pradhan Mantri Jan Arogya Yojana, a health insurance scheme, on 23 September 2018. It is arguably the most ambitious social health insurance programme ever launched anywhere in the world, though without details of funding and strengthening of primary health centres. Whether it is mere election rhetoric or a substantive scheme is hard to assess on the basis of evidence at hand at present.

Notes

1 We have updated India’s gross domestic product (GDP) to arrive at this ratio, compared to the 2012 GDP estimate in Bloom et al (2011).

2 For a more comprehensive literature review, see Yadav et al (2018).

3 With increasing age, numerous underlying physiological changes occur, and the risk of chronic diseases rises. By the age of 60 years, the major burdens of disability and death arise from age-related losses in hearing, seeing, and moving, and non-communicable diseases (NCDs), including heart disease, stroke, chronic respiratory disorders, cancer, and dementia. In fact, the burden of these diseases on old people is considerably higher in low- and middle-income countries (WHO 2015).

4 In India, tobacco consumption takes different forms, some more harmful than others. Beedi smoking accounts for about half of Indian tobacco consumption. Beedis are more harmful than cigarettes because they deliver more nicotine, carbon monoxide, and tar. Another popular tobacco use is flavoured chewing tobacco called gutka. It is also more harmful than cigarettes as it brings on cancer faster (Bloom et al 2014b; Agarwal et al 2016).

5 For an important recent contribution with these two rounds of the National Sample Survey (NSS) to assess self-reported health status by caste and religion, see Borooah (2018).

6 Cases/patients are used synonymously. Shares are computed of men and women suffering from NCDs.

7 (i) In the case of cigarettes, on which excise duty, value added tax (VAT), and other state-level taxes were being levied previously, the highest goods and services tax (GST) rate of 28% is now being levied in addition to GST Compensation Cess and National Calamity Contingency Duty (NCCD). (ii) The government has kept potable alcohol or alcohol for human consumption out of the ambit of GST with states retaining the right to impose taxes on it. Alcohol taxes in India vary according to state laws. So there is no uniformity in alcohol taxes in India, which results in the price difference. What happens is that it is not only the central government that taxes alcohol but also the
local licensing fees and taxes play a huge role. So in the states that have a liberal government, the taxes are low and hence the price of alcohol is relatively low, but for others it is pretty high.

8 In an admirably meticulous analysis, Hooda (2015) attributes commercialisation of hospital care to economic liberalisation and legislation in 2000 permitting 100% foreign direct investment (FDI) in the hospital sector through automatic route. Consequently, foreign investment in hospital sector increased to ₹ 3,995 crore in 2013–14 from a meagre ₹ 31 crore in 2001–02.

9 Hooda’s (2015) view that provisioning of health insurance promotes privatisation of health care seems contentious, as these are likely to be jointly determined. Besides, it is arguable that FDI flows are attracted by the widening gap between demand for and supply of public hospital care.

References

Agarwal, A et al (2016): “Population Aging in India: Facts, Issues, and Options,” Discussion Paper No 10162, Institute of Labour Economics (IZA), Bonn.

Alam, M et al (2012): “Report on the Status of Elderly in Select States of India, 2011,” United Nations Population Fund (India), https://india.unfpa.org/sites/default/files/pub-pdf/AgeingReport_2012_F.pdf, viewed on 5 October 2019.

Alam, M and P Yadav (2014): “Financial Implications of a Universal Old Age Pension Scheme in India: An Exploration,” report submitted to ­International Development Research Centre (IDRC), Canada.

Beard, J R and D E Bloom (2014): “Towards a Comprehensive Public Health Response to Population Aging,” Lancet, Vol 385, No 9968, pp 658–61.

Bloom, D E et al (2011): “The Global Economic Burden of Non-communicable Diseases,” Geneva: World Economic Forum and Harvard School of Public Health.

­— (2014a): “Economics of Non-communicable Diseases in India: The Costs and Returns on Investment of Interventions to Promote Healthy Living and Prevent, Treat, and Manage NCDs,” Geneva, World Economics Forum, Harvard T.H. Chan School of Public Health, https://iogt.org/wp-content/uploads/2015/03/WEF_EconomicNonCommunicableDiseasesIndia_ExecutiveSummary_2014.pdf, viewed on 5 October 2019.

­— (2014b): “China and India’s Descent into Chronic Disease: Killing Themselves Slowly,” Milken Institute Review, 2nd Quarter, pp 24–33.

Borooah, V K (2018): “Caste, Religion, and Health Outcomes in India, 2004–14,” Economic & Political Weekly, Vol 53, No 10, pp 65–73.

Chatterjee, P (2017): “The Health System in India: The Undeserved Majority,” Lancet, Vol 390, No 10111, pp 2426–27.

Ghebreyesus, T A (2018): “Acting on NCDS: Counting the Cost,” Lancet, Vol 391, No 10134, pp 1973–74.

Heath, R (2013): “Women’s Access to Labour Market Opportunities, Control of Household Resources, and Domestic Violence: Evidence from Bangladesh,” World Development, Vol 57, pp 32–46.

Hooda, S K (2015): “Private Sector in Healthcare Delivery Market in India: Structure, Growth and Implications,” Institute for Studies in Industrial Development (ISID), Working Paper 185, New Delhi.

ICMR, PHFI, and IHME (2017): India: Health of the Nation’s States—The India State-Level Disease Burden Initiative, New Delhi: Indian Council of Medical Research, Public Health Foundation of India, and Institute for Health Metrics and Evaluation.

India State-Level Disease Burden Initiative Collaborators (2017): “Nations within a Nation: Variation in Epidemiological Transition across the States of India, 1990–2016,” Lancet,
Vol 390, No 10111, pp 2437–60.

Jan, S et al (2018): “Action to Address the Household Economic Burden of Non-communicable Diseases,” The Lancet Taskforce on NCDs and Economics 3, Lancet, Vol 391, No 10134, pp 2047–58.

Lancet (2018): “Making More of Multimorbidity: An Emerging Priority,” Vol 391, No 10131, p 1637.

(2018a): “The Lancet Taskforce on NCDs and Economics,” Vol 391, No 10134, pp 2029–78.

Mishra, U S et al (2016): “Surveillance of Chronic Diseases: Challenges and Strategies for India,” Indian Council for Research on International Economic Relations (ICRIER), Working Paper No 332, New Delhi.

Patel, V et al (2015): “Assuring Health Coverage for All in India,” Lancet, Vol 386, pp 2422–35.

Prince, M J et al (2015): “The Burden of Disease in Older People and Implications for Health Policy and Practice,” Lancet, Vol 385, No 9967,
pp 549–62.

WHO (2015): World Report on Ageing and Health, Geneva: World Health Organization, https://apps.who.int/iris/bitstream/handle/10665/
186463/9789240694811_eng.pdf;jsessionid=3132839A9EEE6B8134A5E59FBB009276?sequence=1
, viewed on 5 October 2019.

Yadav, A et al (2017): “A Study of Morbidity Pattern among Elderly Population in Urban India,” Journal of Social Health and Diabetes, Vol 5, pp 100–06.

Yadav, P (2018): “Essays in Health Economics from a Health Insurance Perspective,” unpublished manuscript.

Yadav, P, V Kulkarni and R Gaiha (2018): “Growing Burden of Non-communicable Diseases in India,” University of Pennsylvania, Population Centre Working Paper (PSC/PARC), 2018–20, https://repository.upenn.edu/psc_publications/20, viewed on 5 October 2018.

Updated On : 25th Nov, 2019

Comments

(-) Hide

EPW looks forward to your comments. Please note that comments are moderated as per our comments policy. They may take some time to appear. A comment, if suitable, may be selected for publication in the Letters pages of EPW.

Back to Top