Keywords

1 Introduction

There is a considerable variation in trend in adult stature with changes in age and this trend is generally negative with the advancement of age. The negative effect in stature in human body becomes conspicuous in post adulthood phase, i.e., after when one attains around 40 years of age. For some adults it may be visible even in late thirties. There is considerable variation of it among different populations in the world (Harvey 1974; Roche et al. 1981; Malina et al. 1982) and in India (Sidhu et al. 1975; Singh 1978; Bagga 2010, 2013). In India, this type of studies has been carried out for adults who have already attained 60 years and the adults in their twenties (Sharma et al. 1975). It has been found that magnitude of differences between young adults (around 20 years) and late adults (around 70 years) is generally 7–10 cm, though it varies among different communities widely. It may seem to be quite a considerable difference. But Miall et al. (1967) found the decline in stature about 6–7.2 cm among the Welsh women. Among Indian population, very recently, Bagga (2013) studied on Maratha women of 30–70 years and found that it declines up to 3.6 cm.

The decline in the stature with the advancement of age is mainly associated with the changes in the vertebral column, i.e. mainly compression of inter-vertebral discs and hypnosis. This decrement is related with the advancement of age. So, to study the decadal changes, along with the total changes, may reveal features including the intensity of decline in the stature over ages. This change in the length of vertebrae is also associated with osteoporosis and vertebral diseases which cause degenerative changes in vertebral column. Besides this, a few studies in India also stated that socio-economic status has effect on the intensity of degeneration in the stature of human irrespective of gender.

Though there have been studies on the decline in the stature of adults in India, most of the studies, except Bagga’s study, dealt with very old data. Even Bagga’s study consists of very small sample size and no conclusion can be drawn from such a small sample. In this context, our study provides an opportunity to investigate among adult Indian population through national level data. The objective of the study is to find (1) the decadal changes of the height of women and men of 20–49 years of age and (2) the extent of influence of the different socio-economic variables on such changes.

2 Materials and Methods

For this study, we have used the National Family Health Survey (NFHS-III) data conducted by the International Institute for Population Sciences (IIPS), Mumbai, in 2005–2006 (IIPS 2007). IIPS collected unit level data on reproductive aged men of age (15–54) years and women of age (15–49) years from 29 states in India. However, to maintain parity we have taken age range of (15–49) years for both males and females. The sample sizes thus consist of 94,417 women and 52,460 men in the age-group 20–49 years. It may be noted that each round of NFHS is a cross-section data. The background explanatory variables are (1) type of place—rural and urban areas, (2) educational attainment of women grouped into four categories—illiterate (those who can neither read nor write), primary (literate up to class IV standard), middle (Class V to Class X standard) and high school & above (Class XI and above), (3) religion which is classified into four categories, namely Hindu, Muslim, Christian and Others, (4) ethnicity having four categories such as Scheduled Castes (SC), Scheduled Tribes (ST), Other Backward categories (OBC) and Others, (5) Occupations of the women are clubbed into five major groups like not working; professionals, managers, technicians; engaged in service or sales; engaged in agriculture related works and skilled, unskilled or manual labourers, and (6) wealth index of the families. Wealth index represents the economic status of the households. It is an indicator of the level of the wealth, which is consistent with expenditure and income measure (Rutstein 1999). It is based on 33 household assets and housing characteristics like type of windows, sources of drinking water, types of toilet facility, flooring, roofing, ownership of a mattress, a pressure cooker, a chair, a cot/bed, a table, an electric fan, a radio/transistor, television, telephone, a computer, a car, etc. Each household was assigned a score for each asset and the scores were summed for each household and individuals were ranked according to the score of the household and the scores were divided into five quintile groups starting from lower strata to higher strata like poorest, poorer, medium, richer and richest.

Since the ages of males and females span 30 years, the data set has been divided into three consecutive time periods with 10 years span for each period taking age ranges (20–29), (30–39) and (40–49) years. Height has been considered as the dependent variable. To measure the decadal changes of mean height, the mean height of youngest group (20–29 years) has been subtracted from elder groups (30–39 and 40–49 years) and also the mean height of the middle group (30–39 years) has been subtracted from the mean height of the eldest group (40–49 years), so that the differences between the two consecutive age-groups as well as between the two extreme groups can be compared. Besides correlation between height with age, education and wealth index, we have carried out a regression analysis to see how the socio-economic variables influence the height or rather changes in the height. It was done for each decadal age-group separately for males and females. Thus six regression equations have been found. Here height is the dependent variable and place of residence, education, religion, ethnicity, occupation and wealth index have been considered as independent or explanatory variables. Symbolically we can write

$$ {\mathrm{y}}_{\mathrm{i}} = \upalpha + {\upbeta}_1{\mathrm{x}}_{1\mathrm{i}} + {\upbeta}_2{\mathrm{x}}_{2\mathrm{i}} + {\upbeta}_3{\mathrm{x}}_{3\mathrm{i}} + {\upbeta}_4{\mathrm{x}}_{4\mathrm{i}} + {\upbeta}_5{\mathrm{x}}_{5\mathrm{i}} + {\upbeta}_6{\mathrm{x}}_{6\mathrm{i}} + {\upbeta}_7{\mathrm{x}}_{7\mathrm{i}} + {\upvarepsilon}_{\mathrm{i}}. $$
(1)

where y is the dependent variable, i.e., height and the independent variables are x1 = Place of residence, x2 = Education, x3 = Religion, x4 = Caste/tribe, x5 = Respondent’s occupation, x6 = Wealth Index and x7 = Age in years. α is the intercept term and the regression coefficients are β1, β2, β3, β4, β5, β6 and β7, corresponding to the variables x1, x2, x3, x4, x5, x6 and x7.

We have taken binary data for all explanatory variables but age. The binary variables take only 0 for base and 1 for the other category. The base categories are ‘Rural’ for Place of Residence, ‘Primary educated or less’ for Education, ‘Hindus or Muslims’ for Religion, ‘SC, ST or OBC’ for Castes, ‘Other than not working, professionals, technicians or managers’ for Occupation and ‘Poor or middle income persons’ for Economic level. The details of variables, sample sizes, etc. used in the analysis are given in the Appendix. The statistical package for the social sciences (SPSS, version 16.0) has been used for all the analysis.

3 Results

Table 1 and Fig. 1a, b give a vivid picture of the mean heights of women of ages (20–29) years, (30–39) years and (40–49) years, along with decadal changes of the mean heights by zones and states in India. The adult decadal growth in height is found to be negative with a reduction of 0.12 cm from 20–29 years to 30–39 years aged women and 0.31 cm from 30–39 years to 40–49 years, the total reduction being 0.43 cm. Small positive changes have occurred only in eight states out of 29 states in India. As many as 21 states witnessed negative changes. The growth is negative in all the zones. The highest total change occurs in South zone (−1.32 cm) and the lowest total change occurs in North zone (–0.10 cm). Out of total eight states in India, where positive changes have been observed, four states, namely Haryana, New Delhi, Punjab and Rajasthan belong to the north zone of India. The other positive growths are seen in Nagaland, Orissa, Uttar Pradesh and Goa. Almost same trend is seen for both the decadal changes. In India, it is also seen that magnitude of reduction in height due to decadal change from 30–39 to 40–49 years is more than 20–29 to 30–39 years. Since there are six zones in India and for each zone two changes are observed, we have altogether 12 changes for the zones. Out of these 12 changes, only 1 case shows positive growth from 20–29 years to 30–39 years in the central zone and the growth is only 0.11 cm.

Fig. 1
figure 1

(a) State wise changes in the mean height of adult female and male between the age-groups (20–29) years and (40–49) years in India. (b) Zone-wise changes in the mean height of adult female and male between the age-groups (20–29) years and (40–49) years in India

Full size image
Table 1 Zone and state wise changes in the mean height of adult women between (20–29) years and (40–49) years in India
Full size table

Table 2 gives almost similar picture for men so far as positive and negative trends in the height, but here positive changes are found to be lesser in number. Also, the amounts of changes are seen to be more than those of women. The total difference is 1 cm, i.e., the change from 20–29 years to 40–49 years is less by 1 cm on the average for all men taken together. When seen zone-wise, the highest difference is −1.79 cm in west zone. The lowest difference is observed in North-east zone. Out of 29 states, the averages in the heights increased for 6 states, namely Arunachal Pradesh, Meghalaya, Jharkhand, Orissa, New Delhi and Punjab, and for the other 23 states the changes are either negative or remain more or less same. The magnitude of difference of this change for men is a bit more than that of women.

Table 2 Zone and state wise changes in the mean height of adult men between (20–29) years and (40–49) years in India
Full size table

Table 3 describes the total difference and decadal changes in the mean height of women in respect of different socio-economic variables. It is seen that total difference is negative in older aged women than in the younger aged women and the magnitude of difference is more or less double in urban areas (−0.59 cm) than in rural areas (–0.31 cm). In case of relationship with the status of education, illiteracy and primary educated women show positive increments at younger ages and negative growth in older age. The highest educated women have substantially higher heights than other women and do not show much negative trend. Among the Christian and other ethnic groups of women, positive changes are observed, while among Hindus and Muslims, always negative changes occur for all the age-groups. In case of ethnicity, highest negative changes have been found among other backward classes followed by scheduled castes and the lowest difference is observed among scheduled tribes. Occupation-wise the highest total difference is found among service holders (–0.90 cm) and it is followed by professionals (–0.81 cm) while lower magnitudes of differences are observed among non-working women (–0.28 cm) and women engaged in agriculture (–0.24 cm) as well as for skilled/unskilled women labourers (–0.34 cm).The trend is same in both the decades but magnitude is higher in later period than younger ages. Regarding wealth Index, magnitude of negative changes is the highest among the richest women and the lowest among the poorest women.

Table 3 Decadal and total changes in the mean height of women of age (20–49) years in relation to different socio-economic variables
Full size table

Table 4 also describes the relationship between decadal changes in stature with different socio-economic variables among Indian men of aged (20–49 years).Secular total change is negative irrespective of all socio-economic variables. High magnitude of total negative changes has been observed in case of urban areas, Hindu religious group, not working and professional occupation holders and richest wealth index families. The same trend is more or less observed in case of decadal changes also.

Table 4 Decadal and total changes in the mean height of men of age (20–49) years in relation to different socio-economic variables
Full size table

Table 5 and Fig. 2 show the correlation between adult height with age, wealth index and educational level of men and women. The result shows that height is significantly positively correlated with these three socio-economic variables either negatively or positively. It is also seen that adult height is significantly negatively correlated with the age. Thus it proves that there is a negative trend in the heights with the advancement of adult age.

Fig. 2
figure 2

Trend in the mean height of men and women over decadal age-groups

Full size image
Table 5 Correlation between adult heights with age, wealth index and educational level of men and women in India
Full size table

Table 6 contains the decadal age-group wise results of the linear regression of height with different socio-economic variables separately for female and male data in India. The six fitted regression equations are as follows:

  • Female height (40–49 years)

    $$ {\fontsize{8.5}{10}\selectfont{\begin{array}{l}\widehat{\mathrm{y}} = 154.6-0.668\ {\mathrm{x}}_1 + 0.615\ {\mathrm{x}}_2 + 0.758\ {\mathrm{x}}_3 + 0.983\ {\mathrm{x}}_4-0.122\ {\mathrm{x}}_5 + 1.309\ {\mathrm{x}}_6-0.088{\mathrm{x}}_7.\\ {}\kern1.75em (0.000)\kern0.75em (0.000)\kern1.75em (0.000)\kern1.75em (0.000)\kern1.75em (0.000)\kern1.75em (0.188)\kern1.75em (0.000)\kern1.75em (0.000)\end{array}}} $$
    (2)

  • Male height (40–49 years)

    $$ {\fontsize{8.5}{10}\selectfont{\begin{array}{l}\widehat{\mathrm{y}} = 165.0-0.518\ {\mathrm{x}}_1 + 1.155\ {\mathrm{x}}_2-0.462\ {\mathrm{x}}_3 + 1.451\ {\mathrm{x}}_4 + 0.329\ {\mathrm{x}}_5 + 1.784\ {\mathrm{x}}_6-0.063\ {\mathrm{x}}_7.\\ {}\kern1.75em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.072)\kern1.50em (0.000)\kern1.50em (0.001)\end{array}}} $$
    (3)
    Table 6 Linear regression of height with different socio-economic variables for each decadal group of ages among adult females and males in India
    Full size table

  • Female height (30–39 years)

    $$ {\fontsize{8.5}{10}\selectfont{\begin{array}{l}\widehat{\mathrm{y}}=151.6-0.445\ {\mathrm{x}}_1 + 0.719\ {\mathrm{x}}_2 + 0.178\ {\mathrm{x}}_3 + 0.763\ {\mathrm{x}}_4-0.055\ {\mathrm{x}}_5 + 1.441\ {\mathrm{x}}_6-0.002{\mathrm{x}}_7.\\ {}\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.065)\kern1.50em (0.000)\kern1.50em (0.440)\kern1.50em (0.000)\kern1.50em (0.096)\end{array}}} $$
    (4)

  • Male height (30–39 years)

    $$ {\fontsize{8.5}{10}\selectfont{\begin{array}{l}\widehat{\mathrm{y}} = 163.7-0.239\ {\mathrm{x}}_1 + 1.229\ {\mathrm{x}}_2 - 0.959\ {\mathrm{x}}_3 + 1.302\ {\mathrm{x}}_4 + 0.260\ {\mathrm{x}}_5 + 1.775\ {\mathrm{x}}_6-0.031\ {\mathrm{x}}_7.\\ {}\kern1.50em (0.000)\kern1.50em (0.038)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.112)\kern1.50em (0.000)\kern1.50em (0.081)\end{array}}} $$
    (5)

  • Female height (20–29 years)

    $$ {\fontsize{8.5}{10}\selectfont{\begin{array}{l}\widehat{\mathrm{y}}=150.7-0.514\ {\mathrm{x}}_1 + 1.099\ {\mathrm{x}}_2 + 0.101\ {\mathrm{x}}_3 + 0.846\ {\mathrm{x}}_4-0.065\ {\mathrm{x}}_5 + 1.623\ {\mathrm{x}}_6-0.002{\mathrm{x}}_7.\\ {}\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.237)\kern1.50em (0.000)\kern1.50em (0.330)\kern1.50em (0.000)\kern1.50em (0.860)\end{array}}} $$
    (6)

  • Male height (20–29 years)

    $$ {\fontsize{8.5}{10}\selectfont{\begin{array}{l}\widehat{\mathrm{y}} = 163.4-0.275\ {\mathrm{x}}_1 + 1.635\ {\mathrm{x}}_2-1.230\ {\mathrm{x}}_3 + 1.366\ {\mathrm{x}}_4 + 0.971\ {\mathrm{x}}_5 + 1.984\ {\mathrm{x}}_6-0.039\ {\mathrm{x}}_7.\\ {}\kern1.50em (0.000)\kern1.50em (0.011)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.000)\kern1.50em (0.112)\kern1.50em (0.000)\kern1.50em (0.021)\end{array}}} $$
    (7)

  • (Figures in parentheses represent level of significance)

The results of the linear regressions can easily be understood if the values of the regressors are known. When we look at the regression results we see that some relations give different or opposite results than those obtained from taking the simple group means. The place of residence is consistently negatively related with height in the regression equation and the coefficient is significant in all the cases. Observe that we have taken the value 1 for urban and 0 for rural and the negative coefficient of place of residence clearly indicates that rural adults have more height if the effect of other variables is eliminated. The mean values of height in the rural and urban cases give the opposite results. The mean height of urban adults is always more than the mean height of rural adults for each combination of age-group and gender. Other regression coefficients, except religion, more or less give expected results. Age is seen to have a negative relation with height both for the regressions and for group averages. Wealthier or more educated persons have higher heights on the average. Caste is also positively related with height. This means that General Caste Hindus, Christians, etc. have higher heights than SC, ST and OBC people. However, occupation is not significantly related with height for most cases. Religion needs special mention here, because it is positively related with height for females, but negatively related with height for males. This result conforms to the result of the corresponding group means.

4 Discussion

The paper investigates the changes in height vis-à-vis changes in age-groups of adult men and women in India taking three age-groups, namely (20–29), (30–39) and (40–49) years. The reduction in the average height is 0.12 cm from 20–29 years to 30–39 years aged women and 0.31 cm from 30–39 years to 40–49 years, the total reduction being 0.43 cm. So the study shows negative changes in the heights over the successive age-groups. The decadal changes are found to be negative in all the zones of India though it varies zone-wise greatly. Among the women, south zone shows the highest (1.32 cm) and north zone shows the lowest (0.10 cm) change. In most of the states, negative growth occurs but in a few states, positive growth occurs for both the gender groups. In case of men, the highest (1.79 cm) and the lowest (0.28 cm) changes are observed in west and north-east zones, respectively. When male and female heights are compared, the magnitude in the total change is found to be more in males more than in females. The changes in the heights have also been seen among the different socio-economic groups. It is seen in almost all cases that negative changes occur regardless whether it is found for men and women separately or found for all adults in India. It is also an interesting feature that in urban areas, among the richest and richer families, maximum negative increments occur, while among the manual labourers, and scheduled tribes, low magnitude of negative increment occurs. But it is firmly confirmed that height reduces with the advancement of age. Thus it propagates the idea that in human body, post adulthood changes do occur in height. It is supported by many findings (Miall et al. 1967; Roche et al. 1981; Malina et al. 1982; Kirchengast 1994). The most supporting relevant work staking Indian data are (Bagga 1998, 2013; Bagga and Sakurkar 2013).This type of study was mainly done in India or around the world during 1980s and 1990s and in that period, the difference was 5–7 cm (approximately) from younger to older generation, but in our study, the difference is found to be only 0.43 cm. It may be due to the fact that we have taken a smaller span of total years (20–49 years) compared to the time span (20–70 years) taken by them. But, even then, the change in height found by us is too less compared to the changes found by them. It is true that the degeneration starts after 40 years (Roche et al. 1981; Noppa et al. 1980; Sussame 1977; Cline et al. 1989). To understand the changes in the height over age, the span of age must be from 30 to 70 or 80 years. But here, the terminal point of age is 49 years only. As the data is from secondary sources, the male data is available up to 54 years and female data is available up to 49 years. So we have taken all data from 20 years to 49 years to maintain the parity between male and female data. It needs further investigations to see when the degeneration starts and how much degeneration occurs. The effect of the socio-economic variables also needs to be further explored. Is it true that more changing lifestyle in a broad sense which includes changing food habits also result into more degeneration of height? We have in fact seen that more negative changes occur among the urban people, and richer and richest families as age increases.