|
Introduction
Chronic non-communicable diseases are the leading causes
of both disability and death worldwide and they strike
hardest at the world's low- and middle - income populations
[1]. In the Indian context, several studies have documented
the prevalence of chronic diseases in India too [2].
An analysis of chronic diseases with regard to socioeconomic
and demographic factors has been attempted by quite
a few researchers like, Omran (1971) [3], Lowery et
al. (1996) [4], Berendregt and Bonneux (1998) [5], Lopez
(2006) [6], Gupta et al. (1995) [7], Das, Sanyal and
Basu (2005)[8], Choudhury et al. (2009) [9] etc.
Various hypotheses have been put forward to explain
the rising trend of chronic diseases, and consequence
of urbanization is one of them. Though biological factors
might have influences on chronic diseases, a majority
of chronic diseases are due to lifestyle behaviours
[8]. On the basis of this we hypothesise the prevalence
of chronic diseases in urban areas of Nagaland.
Nagaland, the sixteenth state of the Indian union, is
one of the smallest states of India. The state consists
primarily of tribal population with 96% household heads
belonging to schedule tribal communities [10]. The major
tribes are the Angami, Ao, Chakhesang, Chang, Khemungan,
Konyak, Lotha, Phom, Pochury, Rengma, Sangtam, Sema,
Yimchunger and Zeliang. Kohima is the state capital
and Dimapur is the largest commercial town of the state.
The health indices of Nagaland seem to be better than
the national averages with infant mortality rate as
low as 26 infant deaths per 1000 live birth [11], and
life expectancy at birth 67.33(67.94) years for males
(females) [12]. If life expectancy at birth is more
than 55 years, then death due to chronic disease like
cardiovascular diseases, cancer and diabetes become
more prevalent and frequent [13]. As such there is every
possibility of prevalence of different chronic diseases
in urban areas of Nagaland. Keeping this in mind, this
paper assesses and evaluates the influence of various
socio-economic, demographic and cultural factors as
well as some risk behaviours on prevalence of chronic
diseases among a representative sample of urban population
of Nagaland.
Materials and Methods
The data used in this study is a primary one collected
through a household survey conducted in Kohima and Dimapur
towns of Nagaland during May - August 2010. The primary
objective of the survey is to gather information on
chronic diseases prevailing in urban areas of Nagaland
viz., in Kohima and Dimapur town. The municipal area
of Kohima and Dimapur had a total population of 77,030
and 107614 respectively in 2001. There are 19 wards
under Kohima town and 21 wards under Dimapur Municipal
Corporation. The sample selection and implementation
procedures were designed to ensure that the survey provides
statistically valid estimates for population parameters.
We have used Stratified Random Sampling technique taking
wards as strata [14]. It is more or less known that
the wards are homogenous with respect to socio-economic,
demographic and cultural factors in both the towns.
From each ward, households were selected by Simple Random
Sampling. The latest available voters list of both the
towns was used for random selection of the household.
For this purpose of data collection from the respondents,
a structured schedule was prepared in accordance with
local health related problems in the areas. The schedule
was pre-tested in order to test the validity of questions
with regard to the objective of the study. Basic information
was collected on each person including age, sex, marital
status, religion, tribe, education, occupation, household
income, food habit, type of residence, cast and family
type. Information is also collected on the prevalence
of certain chronic diseases (cardiovascular, diabetes,
cancer, chronic respiratory disease, cirrhosis of liver,
renal failure, asthma) and on certain risk behaviours
(physical exercise, consuming tobacco and alcohol, smoking,
body mass index etc.).
Altogether 4640 respondents were interviewed from both
the towns, out of which 2457 were male and 2183 were
females, from 958 households. Identification of the
chronic disease afflicted persons was based upon the
information provided by the respondent or elder family
members of the household, but not clinically tested.
The independent variable is divided into three categories
viz. socio-economic, cultural and demographic. From
socio-economic characteristic we have included annual
family income, occupation and education. From demographic
characteristics, variables like age, gender, marital
status and residence type were included. Regarding cultural
factors, we have considered, religion, caste, food habit
and type of family. Some other risk behaviours like,
chewing tobacco, addiction to smoking and consumption
of alcohol, level of physical exercise and body mass
index are also included for analysis.
The logistic regression analysis is used to examine
the strength of association between each covariate and
the dependent variate (presence of chronic disease).
The dependent variable is dichotomous i.e., presence
or absence of chronic disease in the respondent.
Let, Y denote the dichotomous
outcome variable and
be a set of independent variables. Then the form of
the logistic regression model is
 ;
where
… (1a)
A transformation of
is the logit transformation which is defined in terms
of as Hosmer and Lemeshow [15].
 
… (1b)
Where 
is the Y intercept, 
(i= 1,2,…,k) are regression coefficients, and
are a set of predictors.
Observations
We have observed from our data that out of the entire
sample of size 4640 very few cases of chronic disease
have been reported under age 25 years. Therefore for
the analysis we have considered only those respondents
who have attained age 25 years at the time of interview.
Accordingly, the logistic regression analysis was carried
out on 2328 respondents, out of which 1218 were male
and 1110 were female. In the reduced sample 15.55% respondents
were affected with chronic diseases.
Figure 1: Scatter diagram showing age and proportion
of persons affected by chronic diseases
From Figure 1 it has been observed that as age increases
the proportion of persons suffering from chronic diseases
also increases.
As mentioned earlier, we have used logistic regression
analysis technique to examine the association between
various factors and the prevalence of chronic diseases.
To check the appropriateness of the fitted model we
have compared the actual and the predicted outcomes
derived from logistic regression analysis. The Analysis
has been carried out by SPSS -11.5 software. The overall
correct percentage is 72.5, which is found to be very
satisfactory.
The Hosmer and Lemeshow test provides a formal test
for whether the predicted probabilities for a covariate
match the observed probabilities. The test shows large
p-value (p = 0.830) indicating a good match to describe
the relationship between the covariates and the outcome
variable.
In Table 1.1, Table 1.2 and Table 1.3, some of the significant
factors are presented along with their p-values for
chronic disease, cardiovascular disease and diabetes
respectively. The odds ratio corresponding to 95% confidence
intervals is also being presented.
Results and Discussion
It is clear from Table 1.1, that there is a significant
relationship between age and prevalence of chronic diseases.
For our analysis we have categorized age into two categories
viz. 25-50 years and above 50 years. This categorization
is done, as out of total chronic diseases reported in
the reduced sample, 62.71% chronic diseases are for
persons age 50 years and above. Taking persons above
50 years of age as reference category, it is seen that
persons in the age group 25-50 years are approximately
4 times less likely (odds ratio = 0.270) to have chronic
disease compared to those above 50 years of age. Similar
findings have been found by Gupta et al. in an investigation
on the prevalence of coronary heart disease and coronary
risk factors in an urban Indian population [7].
It has also been observed that there is a weak association
(p-value = 0.071) between gender and prevalence of chronic
disease. Taking female as a reference category, we find
that males are one and half times less likely (odd ratio
= 0.681) to have chronic diseases compared to females.
Table 1.1: Results of Logistic Regression for chronic
diseases
Note: ® Reference category, * Significant at
5% probability level, ** Significant at 10% probability
level.
For the variable marital status, we have considered
three categories viz. never married, widow/separated
and currently married. Taking currently married as the
reference category, we have observed that a never married
person is almost 6 times less likely (odd ratio = 0.172)
to have chronic disease as compared to the reference
category. However, a widow/separated person is almost
3 times more likely (odd ratio = 2.700) to have chronic
disease as compared to a currently married person. This
may be possible because without their spouse they may
have to take care of various problems in the family
which leads to physical and mental stress.
For the variable types of residence, we have considered
three categories - kachcha house, semi- pucca and pucca.
It has been observed that persons living in kachcha
houses are approximately three and half times less likely
(odd ratio = 0.293) to have chronic disease compared
to persons living in pucca house. Kachcha houses are
made up of mud and thatched roof which provides very
poor living conditions. People living in such environment
may have higher prevalence of communicable diseases
[16].
We have classified the per capita annual income into
three categories viz., less than Rs.10000, Rs. 10000
to Rs. 30000 and above Rs. 30000. This classification
has been done in accordance to the classification of
the Organization for Economic Co-operation and Development
(2003), which classifies India in the per capita annual
income < $745 (approx. Rs.30000) group. Considering
per capita annual income of more than Rs. 30000 as the
reference category it is observed that persons with
per capita income less than Rs. 10000 are one and half
times less likely (odd ratio = 0.623) to have chronic
disease. If per capita annual income is between Rs 10000
to Rs 30000, the chance of having chronic disease is
reduced approximately by 30% as compared to the reference
category. Thus economically advanced persons have more
chance of acquiring chronic diseases. People with lower
family income may have higher prevalence of other communicable
diseases. Similar results have been reported by Ghosh
and Arokiasamy in an investigation on morbidity in India
[17].
We also intended to examine the relationship between
occupation and prevalence of chronic disease. It is
observed that housewives have almost three times less
chance (odd ratio = 0.308) of having chronic disease
as compared to retired persons. However, a self employed
person is two and half times less likely (odd ratio
= 0.399) to have chronic disease compared to a retired
person. Persons working in the private sectors are three
times less likely (odd ratio = 0.318) to have chronic
disease compared to retired persons. Government employees
are approximately three times less likely to have chronic
disease compared to retired persons. The other category
includes both unemployed and elderly persons, which
also is found to be significant (p-value = 0.002).
It has been observed that variables like education,
religion, food habit, caste, family type etc. have no
significant effect on prevalence of chronic diseases.
For the variable physical exercise it has been observed
that persons doing regular exercise are three and half
times less likely (odd ratio = 0.280) to have chronic
disease compared to the reference category. There is
evidence that regular physical exercise increases the
high-density lipoprotein and decreases both body weight
and blood pressure which are beneficial to cardiovascular
health [18]. As far as the risk factors are concerned,
physical activity can interact in various ways that
influence the risk of several chronic diseases [1].
It has been observed that, persons who consume tobacco
regularly are one and half times at high risk of having
chronic diseases (odd ratio = 1.572) compared to persons
not consuming it. For the variable smoking we have considered
persons who smoke cigarettes daily or at least twice
a week as smokers and others as non smokers. Taking
non smokers as reference category we have observed that
smokers are approximately one and half times (odd ratio
= 1.418) at high risk of acquiring chronic disease compared
to non smokers. This fact is widely acknowledged as
smoking has been identified as a major coronary heart
disease risk factor [18]. For the variable consumption
of alcohol, we have considered those persons who consume
alcohol regularly or at least once a week as alcoholic
and others as non alcoholic. It is clear from the table
that alcoholic persons are more than two and half times
(odd ratio = 2.440) at high risk of having chronic disease
compared to non alcoholic persons. Similar results have
been observed by Lowry et al. (1996) [4]. Also, high
alcohol intake (75g or more) per day is an independent
risk factor for hypertension and all cardiovascular
diseases [19].
For the characteristics body mass index, we have considered
three categories viz. underweight (BMI<18), normal
(BMI = 18-25) and overweight (BMI>25). If we take
overweight as a reference category, then persons with
BMI<18 (underweight) are two and half times less
likely (odd ratio = 0.373) to have chronic diseases
compared to overweight persons. Further, persons with
normal weight are 50% less likely (odd ratio = 0.501)
to have chronic disease compared to overweight or obese
persons. Obesity may be mediated by other cardiovascular
disease risk factors, including hypertension, diabetes
mellitus, and lipid profile imbalances [20]. Overweight
and obesity have a significant association with chronic
disease [21].
It is observed from our data that in the reduced sample
of persons of age 25 years and above, 6.4% persons are
afflicted by cardiovascular diseases (CVD), 4.6% suffer
from diabetes, 0.9% are afflicted by cancer and 3.7%
are afflicted by other chronic diseases. Out of all
persons suffering from chronic diseases, 40.89% suffer
from CVD, 29.28% are from diabetes, 6.07% are from cancer
and 23.76% are from other chronic diseases. These observations
suggest that in our study prevalence of CVD is the most
common and frequent followed by diabetes. Similar observations
have been reported by Choudhury et al. in an investigation
on prevalence of chronic diseases in Guwahati city [9].
Table 1.2 presents the results of the logistic regression
analysis for cardiovascular diseases (CVD) with respect
to different characteristics. It has been observed from
the analysis that CVD in particular and chronic diseases
taken together show similar types of results with few
notable exceptions. There is a weak association (p-value
= 0.071) between gender and prevalence of chronic disease
taken together, but we found no significant association
between gender and the prevalence of CVD. Also, the
variable occupation has found to be associated with
chronic disease, but there is no significant effect
of occupation on prevalence of CVD. There is no significant
relationship between food habit and prevalence of chronic
diseases taken together, but we found a significant
association between food habit and prevalence of CVD.
Table 1.2: Results of Logistic Regression for Cardiovascular
diseases
Note: ® Reference category, * Significant at 5%
probability level, ** Significant at 10% probability
level.
It is seen that vegetarians are two times more likely
(odd ratio = 2.106) to have CVD as compared to non-vegetarians.
This may be possible because most of the vegetarian
people found in the survey are from non Naga community
and they usually take food high in fats resulting increasing
cholesterol level. In a study of dietary pattern of
Japan and Finland, it is found that Japanese have low
fat diets resulting in low serum cholesterol and low
incidence of coronary heart disease [18].
Moreover, persons belonging to a joint family have almost
one and half times less chance (odd ratio = 0.635) of
having CVD as compared to nuclear family. There is a
weak association (p-value = 0.098) between BMI and prevalence
of cardiovascular disease in the urban setup of Nagaland.
Persons with BMI<18 (underweight) are half times
less likely (odd ratio = 0.45) to have chronic diseases
compared to overweight persons.
The results of logistic regression analysis for diabetes
are presented in Table - 1.3. It is observed that age,
marital status, annual family income, physical exercise
and body mass index have significant effect on prevalence
of diabetes as well as chronic disease taken together.
It is seen that persons in the age group 25-50 years
are approximately three and half times less likely (odd
ratio = 0.284) to have chronic disease compared to those
above 50 years of age. Although diabetes may occur at
any age, surveys indicate that prevalence rises steeply
with age. [18] However, variables like gender, residence
type, occupation, chew tobacco, smoking, consumption
of alcohol were found to have no significant effect
on diabetes unlike chronic diseases taken together.
However, excessive intake of alcohol can increase the
risk of diabetes by damaging the pancreas and liver
and by promoting obesity [19]. We observe a weak association
(p value = 0.058) between a person who belongs to a
joint family and prevalence of diabetes. Persons belonging
to a joint family are approximately one and half times
more likely (odd ratio = 1.590) to have diabetes compared
to the reference category nuclear family.
Table 1.3: Results of Logistic Regression for Diabetes
diseases
Note: ® Reference category, * Significant at 5%
probability level, ** Significant at 10% probability
level.
Conclusions
Chronic disease is primarily a mass disease. The strategy
should therefore be based on mass approach focusing
mainly on the control of underlying causes in whole
populations, not merely in individuals. A small change
in risk factor levels in total population can achieve
the biggest reduction in chronic disease mortality.
As there is a large proportion of chronic diseases cases
are observed in urban areas of Nagaland, first of all,
the overall burden of chronic disease risk factors should
be lowered through population-wide public health measures,
such as community level campaigns against cigarette
smoking, unhealthy diets, and physical inactivity etc.
In spite of the tremendous advancement in the field
of preventive medicine, the health care services in
tribal communities of Nagaland are still poor and need
to be strengthened to reduce the occurrence of these
potentially fatal chronic diseases.
References
1. World Health Organization. Global Status Report on
Non Communicable Diseases, 2010.
2. Registrar General of India. Summary statistics of
causes of death India, 2008.
3. Omran A.R. The Epidemiological Transition: A Theory
of The Epidemiology of Population Change. Milbank Memorial
Fund.1971; 509-538.
4. Lowry R, Laura K, Collins J.L, Kolbe L.J. The Effect
of Socioeconomic Status on Chronic Disease Risk Behaviours
among US Adolescents. JAMA.1996; 276(10).
5. Barendregt J, Bonneux L. Degenerative Diseases in
an Aging Population: Models and Conjectures. Ph.D. Thesis.
Erasmus University, Rotterdam, 1998.
6. Lopaz A.D, Mathers C.D, Ezzati M, Jamison D.T, Murray
C. Global Burden of Disease and Risk Factors. Oxford
University Press, New York, 2006.
7. Gupta R, Prakash H, Majumdar S, Sharma S, Gupta V.P.
Prevalence of coronary heart disease and coronary risk
factors in an urban population of Rajasthan. Indian
Heart Jr. 1995; 47(4):331-8.
8. Das S.K, Sanyal K, Basu A. Study of urban community
survey in India: growing trend of high prevalence of
hypertension in a developing country. Int. J. Med. Sci.
2005; 70-78.
9. Choudhury L, Choudhury A, Deka M. Influence of socio-economic
and demographic factors on the risk of chronic diseases
in Guwahati city - The premier city of North East India.
Demography India. 2009, Vol. 38(1).
10. Registrar General of India. Census of India, 2011,
New Delhi, India.
11. Registrar General, India. Sample Registration System
Bulletin. 2011, 45(1).
12. Choudhury, L, Barman P, Sarma R. Estimation of life
expectancy at birth for the smaller north eastern states
of India (2001-2005). Indian Journal of Public Health
Research and Development; 4(1): 143-146.
13. Preston S.H. Mortality Patterns in National Populations
with Special Reference to Recorded Causes of Death.
Academic Press. 1976.
14. Gupta A, Gupta R, Sarma M, Rastogi S, Gupta V.P,
Kothari K. Prevalence of diabetes, impaired fasting
glucose and insulin resistance syndrome in an urban
Indian population. Diabetes Res Clin Pract. 2003; Jul
61(1), 69-76.
15. Hosmer D.W, Lemeshow S. Applied Logistic Regression.
New York: John Wiley & Sons, 1989.
16. Yesudian C.A.K. Health Services Utilisation in Urban
India: A Study. Mittal Publication, Delhi, 1988.
17. Ghosh S, Arokiasamy P. Morbidity in India Trends,
Patterns and Differentials. Journal of Health Studies;
2009, II.
18. Park K. Park's Textbook of Preventive and Social
Medicine. 18th ed. Banarsidas Bhanot publication, 2005.
19. World Health Organization. Primary prevention of
CHD. 1985.
20. Harrison T.R. Principles of Harrison's Internal
Medicine, 17th ed. Mc Graw Hill Medical publication,
2004.
21. Valocikova I, Valocik G, Kristofova B, Druzbacka
L. Obesity paradox and chronic kidney disease. Bratisl
Lek Listy 2011; 112(7).
|
