International Journal of

cardiology
ELSEVIER International Journal of Cardiology 56 (1996) 289-298

Association of trans fatty acids (vegetable ghee) and clarified butter

(Indian ghee) intake with higher risk of coronary artery disease in
rural and urban populations with low fat consumption

Ram B. Singh”‘“, Mohammad A. Niaz”, Saraswati Ghosh”, Raheena Beegom”,

Vipul Rastogi”, Jagdish P. Sharma”, Gopal K. Dubeb
“Heart Research Laboratory, Centre of Nutrition, Medical Hospital and Research Centre, Civil Lines, Moradabad 10 (UP) 244001,
India
hNKP Salve Institute of Medical Sciences, Nagpur, India

Received 20 March 1996; accepted 26 June 1996

Abstract

These cross-sectional surveys included 1769 rural (894 men and 875 women) and 1806 urban (904 men and 902 women)
randomly selected subjects between 25-64 years of age from Moradabad in North India. The total prevalence of coronary
artery disease based on clinical history and electrocardiogram was significantly higher in urban compared to rural men ( 11.O
vs. 3.9%) and women (6.9 vs. 2.6%), respectively. Food consumption patterns showed that important differences in relation
to coronary artery disease were higher intake of total visible fat, milk and milk products, meat, eggs, sugar and jaggery in
urban compared to rural subjects. Prevalence of coronary artery disease in relation to visible fat intake showed a higher
prevalence rate with higher visible fat intake in both sexes and the trend was significant for total prevalence rates both for
rural and urban men and women. Subgroup analysis among urban (694 men and 694 women) and rural (442 men and 435
women) subjects consuming moderate to high fat diets showed that subjects eating trans fatty acids plus clarified butter or
those consuming clarified butter as total visible fat had a significantly higher prevalence of coronary artery disease compared
to those consuming clarified butter plus vegetable oils in both rural (9.8, 7.1 vs. 3.0%) and urban (16.2, 13.5 vs. 11.O%) men
as well as in rural (9.2, 4.5 vs. 1.5%) and urban (10.7, 8.8 vs. 6.4%) women. Univariate and multivariate regression analysis
with adjustment for age showed that sedentariness in women, body mass index in urban men and women, milk and clarified
butter plus trans fatty acids in both rural and urban in both sexes were significantly associated with coronary artery disease. It
is possible that lower intake of total visible fat (20 g/day), decreased intake of milk, increased physical activity and
cessation of smoking may benefit some populations in the prevention of coronary artery disease.

Keywords: Visible fat; Milk; Clarified butter; Trans fatty acids diet; Risk factors

1. Introduction ment and rapid changes in diet and lifestyle, car-

diovascular diseases would emerge as the most
It has been predicted that with economic develop- important cause of mortality in India in the next two
decades [1,2]. Recent studies [3-81 have shown a
*Corresponding author. higher prevalence of coronary artery disease in both

0167-5273/96/$15.00 0 1996 Elsevier Science Ireland Ltd. All rights reserved

PIZ SO167-5273(96)02760-X
rural and urban populations of India, and in urban
areas the disease is as common as in developed
countries [6-S]. Studies from rural areas [3-51 have
demonstrated a lower prevalence rate compared to
urban but statistics indicate that coronary disease has
at least doubled in the last two decades between
1975-1995 in both rural (1.7 to 3.0%) and urban
(4.5 to 8.6%) population groups [5,9]. Coronary
disease is 3 times more common in urban north and
-5 times in Indian immigrants to Britain than in rural
north Indians (3.0%) [5,10,11]. The increase in the
prevalence of coronary artery disease in the last two
decades particularly in the urban areas has been
marked with the introduction of vegetable ghee tram
fatty acids into the diet in the 1960s and increased
consumption of Indian ghee (clarified butter) and
animal foods due to economic development [l-4].
Therefore, variations in the fatty acid composition of
the diet may explain the cause of differences in the
prevalence rates of coronary artery disease in differ-
ent population groups [ l-101. However, these levels
of fat intake are considered within permissible limits
in developed countries [ 1l-141. It is possible that
high levels of lipoprotein(a) and insulin resistance
observed in Indians [15] in association with in-
creased intake of trans fatty acids and clarified butter
might explain the cause of higher prevalence of
coronary disease even on modest increase in total fat
intake and other risk factors [4,12]. It seems that
Indians are in a better situation to demonstrate the
relation of trans fatty acids with coronary artery
disease because urban diets contain a higher amount
of trans fatty acids which is rarely consumed in rural
areas.
In the present study, we compare the visible fat
intake and coronary risk factors in a rural and urban
population in relation to the prevalence of coronary
artery disease using similar diagnostic tools with
adequate sample size and well-defined criteria, pos-
sibly for the first time, to our knowledge, from India.
2. Subjects and methods
The details of subjects and methods regarding the
urban sample have been described [ 161. In brief, the
sampling frame for urban subjects consisted of the
final population total of Moradabad city which is
roughly 0.43 million based on a 1Y9 1 census 117).
After excluding 16 suburban streets, 20 out of the
remaining 180 streets were randomly selected by
persons unconnected with the study by blindly
selecting 20 cards each enclosed in a sealed envelope
and blindly selected from 180 envelopes each repre-
senting a street. There are 68 villages in the
Moradabad Tahsil each having a population of
1000-5000 subjects with a total population of 0.1
million in the Tahsil. After excluding 6 suburban
villages, two villages were selected at random from
the 62 remaining villages. A colony or mohalla in a
rural village or urban area was defined as a street.
Each street contained 6-12 blocks and the block
containing 100-300 subjects was decided to be a
cluster.
2.1. Sample selection
Assuming from other studies [3,5] that the preval-
ence of coronary artery disease is at least 5% in the
community we would need a sample of at least 1500
subjects to estimate with 90% confidence and to
detect at 5% significance a relative risk of 1.54 for
the prevalence of coronary disease in Indians [18].
We decided to select 2000 subjects to increase the
level of confidence. Thus with a sample of 2000
subjects each, from rural and urban areas, we would
be able to estimate the prevalence to within 2% error
on either side which was deemed satisfactory as a
first estimate.
Of 20 streets randomly selected from an urban
area, and two villages, 2-3 blocks or clusters were
randomly selected from each street and from each
block 40-100 adults were randomly selected by
randomly selecting a card as described above. When
the random number fell to a subject who was
younger or not available, it was assigned to the next
person in the list. We contacted 2024 rural and 1991
urban subjects between 25-64 years of age. Of these
subjects, 255 (12.6%) rural and 185 (9.3%) urban
refused to give time for detailed history and exami-
nation. The remaining 1769 rural (894 males and 875
females) and 1806 urban (904 males and 902
females) between 25-64 years, inclusive, of age
were invited for the study. Our preliminary study [5]
from rural and urban populations showed a similar
response rate from both sexes. Detailed interviews
 R.B. Singh et al. I International
were performed with the help of a pretested and
validated questionnaire [5] prepared according to
guidelines of the World Health Organization [ 181.
The questionnaire was filled by the nutritionist and a
physician to collect information on dietary intake,
age, sex, education, occupation, socioeconomic
status, physical activity, alcohol intake, smoking,
hypertension, diabetes, and coronary artery disease
(Rose questionnaire) [ 181.
The social class of the family was graded I-V
according to housing condition, occupational and
other incomes of all the family members and number
of dependents [ 181. Per capita income was calculated
by finding out the total income of the family divided
by the total number of family members. Social class
I-III were subgrouped into higher and middle socio-
economic status and class IV and V into lower
socioeconomic status.
Dietary intakes were obtained based on World
Health Organization guidelines [ 181 by 7 days food
intake record of all adults in the household through-
out the year by using food measures, food models
and food portions. A cross-check questionnaire was
completed based on dietary record by asking probing
questions to assess the actual intake of foods men-
tioned in the record. Indian food composition tables
[ 191 were used to calculate 24 h nutrient intake based
on food intake record. Total visible fat intake was
classified into low (<20 g/day), moderate (20-30
g/day) and high (>30 g/day) and all subjects in
both groups were classified according to total visible
fat intake based on guidelines of the World Health
Organization [ 11, Indian Consensus Group [20] and
the Indian Council of Medical Research [21].
Body weights were measured by the health worker
independently in under-clothes to the nearest 0.5 kg.
Body mass index was calculated and obesity defined
as a body mass index of >27 kg/m* and overweight
as a body mass index >25kg/m*. Figures for criteria
laid down by the Indian Consensus Group [20] for
overweight (>23 kg/m’) were also calculated.
It is difficult to measure the amount of tobacco
consumed in India because it is consumed in various
forms. Cigarettes, beedies, Indian pipes, raw tobacco
and chewing tobacco are commonly consumed and
people use tobacco in more than one form. We
therefore categorized users of any form of tobacco as
tobacco users. Physical activity was assessed by
Journal of Cardiology 56 (1996) 289-298 291
asking about both occupational and spare time 1221
activities. According to Paffenberger et al. [22], a
person is considered as leading a sedentary lifestyle
if he walks less than 14.5 km a week, climbs fewer
than 20 flights of stairs a week or performs no
moderately vigorous physical activity on 5 days a
week.
Coronary artery disease was diagnosed if one or
more criteria [18] were satisfied (a) documented
history of chest pain suggestive of angina or infarc-
tion and previously diagnosed coronary artery dis-
ease. (b) Affirmative response to Rose questionnaire
after excluding any obvious cause of pain due to
local factors. (c) Coronary artery disease was also
diagnosed in the absence of criteria (a) and (b) but in
the presence of electrocardiographic changes, namely
Minnesota codes l-l-l through l-l-7 or 1-2-1
through l-2-7; the presence of major ST segment and
major T wave and Q wave changes or Q wave
changes in the absence of high voltage R wave
(Minnesota code 4-l -1 and 4-l-2 and 5-l and 5-2
were also diagnosed as coronary artery disease). A
1Zlead electrocardiogram was recorded in all sub-
jects.
2.2. Statistical analysis
The prevalence rates are given in percent and
numerical variables as mean + standard deviation.
Significance of various age-related trends was ex-
amined by using a chi-square test for trend (Mantel-
Haenzel). Significance of association of various risk
factors and food intakes was determined by regres-
sion analysis wherein odds ratios and 95% confi-
dence intervals were calculated by univariate and
multivariate analysis using overall CAD prevalence
as the dependent variable.
3. Results
We studied a total of 1769 rural (894 men and 875
women) and 1806 urban (904 men and 902 women)
subjects. The age structure, male and female ratio
and visible fat intake in rural and urban subjects
were comparable with available census data [ 171 of
Moradabad and national data on food intake for rural
and urban subjects [ 19,211. The overall prevalence of
292 R.B. Singh et al. I International Journal of Cardiology S6 (1996) 289-298

Table 1
Food consumption pattern in rural and urban subjects by sex (g/day). Values are given as the mean with the standard deviation in
parentheses
Foods g/day Men Women
Rural Urban Rural Urban
Number 894 904 875 902
Wheat, rice and millets 306 (36.6)* 265 (28.7) 284 (30.5)* 245 (25.4)
Roots and tubers 48 (8.9) 82 (12.5)* 40 (5.8) 74 (9.8)”
Pulses 24 (4.5) 41 (6.7) 18 (3.5) 38 (6.0)
Vegetables 41 (6.6) 72 (8.8) 30 (4.0) 70 (5.8)”
Fruits 22 (3.2) 85 (10.5)* 18 (3.2) 76 (9.5)“”
Milk and its products 186 (11.8) 285 (16.8)* 160 (12.0) 232 (15.8)*
Sugar and jaggery 22 (3.2) 30 (4.6)* 18 (3.5) 34 (s.‘)*
Total visible fat 18.0 (3.1) 29.0 (4.5)* 14.9 (2.8) 22.8 (4.1)*
Butter 2.2 (0.5) 3.2 (0.9) 2.3 (0.4) 2.6 (0.8)
Ghee (Clarified butter) 6.5 (1.2) 9.5 (1.6)* 4.6 (1.0) 7.0 f,1.3)*
Hydrogenated fat 4.1 (0.8) 12.2 (2.5)** 3.2 (0.6) 9.4 ( 1.sy*
Oils 5.2 (1.0) 4.1 (0.9) 4.8 (1.2) 3.8 (0.8)
Meat and eggs 10.7 (2.5) 32 (4.8)** 7.6 (1.5) 27.6 (3.6)**
Total fruit, vegetable and pulses 87 (7.5) 198 (21.6)** 66 (6.5) 184 (16.5)**
Fruit, vegetables, pulses/visible fat ratio 4.83 (1.2) 7.4 (1.6)* 4.68 (1.1) 7.86 (1.8)*
P value obtained by comparison of rural and urban subjects. *P<O.O5; **P<O.Ol.

coronary artery disease in the rural population was
3.2% and in urban subjects 9.0%. The prevalence
rate between 25-64 years of age was -3 times
higher in the urban compared to rural population in
both men (11.0 vs. 3.9%) and women (6.9 vs. 2.6%).
The prevalence rate increased with age in all age
groups in both sexes and, after the age of 55 years it
became comparable in both sexes.
While the consumption of wheat, rice and millets
was significantly lower in urban, the intake of all
other foods was significantly higher compared to
rural subjects (Table 1). Table 1 shows that the
consumption of total visible fat was significantly
higher in urban men and women compared to rural,
however the intake of butter and oils was similar. In
our study, rural subjects were consuming mainly
unrefined mustard oil and urban either ground nut,
soyabean or mustard oil which are available as
refined oils. Indian ghee (clarified butter) and vege-
table ghee (trans fatty acids) intake were significantly
higher in urban men and women compared to rural
subjects who eat vegetable ghee occasionally. The
intake of meat and eggs was greater in urban than
rural subjects.
Table 2 shows that about half of the rural men
and women were eating a low fat diet and these

Table 2
Prevalence of coronary artery disease in relation to visible fat intake
Visible fat (g/day) Men Women
Rural Urban Rural Urban
Subjects Prevalence Subjects Prevalence Subjects Prevalence Subjects Prevalence
Low (<20 g/day) 452 9 (2.0) 210 10 (4.7) 440 5 (1.1) 208 7 (3.3)
Moderate (20-30 g/day) 330 14 (4.2) 489 52 (10.6) 325 8 (2.4) 482 30 (6.2)
High (>30 g/day) 112 12 (10.7) 205 38 (18.5) 110 10 (9.0) 212 26 (12.2)
Total 894 35 (3.9) 904 100 (11.0) 875 23 (2.6) 902 63 (6.98)
Mantel-Haenzel x2 10.2 17.2 8.6 12.6
P value <0.004 <O.OOl co.05 <co.004
 R.B. Singh et al. I International Journal of Cardiology 56 (1996) 289-298 293

Table 3
Prevalence of coronary artery disease among a subgroup of rural and urban subjects consuming a moderate to high fat diet in relation to type
of fat intake
Type of fat Men Women
Rural Urban Rural Urban
Subjects Prevalence Subjects Prevalence Subjects Prevalence Subjects Prevalence
Ghee + tram fatty acids 102 10 (9.8) 210 34 (16.2) 97 9 (9.2) 215 23 (10.7)
Ghee 140 10 (7.1) 96 13 (13.5) 132 6 (4.5) 90 8 (8.8)
Ghee+oils’ 200 6 (3.0) 388 43 (11.0) 206 3 (1.5) 389 25 (6.4)
Total 442 26 (5.9) 694 90 (12.9) 435 18 (4.1) 694 56 (8.0)
Mantel- Haenzel x’ 9.6 14.5 7.8 11.8
P value ‘Co.01 <O.OOl CO.05 <o.oos
Ghee=clarified butter.

adults had a lower prevalence of coronary disease.
However urban subjects eating the same amount of
fat in each category had a relatively higher preval-
ence of coronary artery disease. The prevalence of
coronary artery disease in association with visible fat
intake was apparent in both sexes and the trend
indicated higher prevalence on higher intake of fat
both for rural and urban men and women (Table 2).
Among urban (694 men and 694 women) and rural
(442 men and 475 women) subjects consuming a
moderate to high fat diet, subgroup analysis was
performed to find the correlation of fat consumption
with prevalence of coronary disease. The prevalence
of coronary artery disease was higher in those rural
and urban men and women who were consuming
clarified butter and trans fatty acids or clarified butter
alone compared to the subgroup eating clarified
butter and vegetable oils (Table 3).
There was a significant positive association of age,
body mass index, total, saturated fat and cholesterol,
milk, meat and eggs, sugar and jaggery, tobacco,
sedentary lifestyle, higher and middle socioeconomic
status and education with a higher prevalence of
coronary artery disease as shown in Table 4. Tobac-
co intake was not associated with coronary artery
disease in urban women. Spearman’s rank correla-
tion was calculated to find the correlation of food
intake with coronary disease. There was a significant
positive rank correlation of coronary disease with
consumption of visible fat in both rural (men, I=
0.14, P<O.Ol; women, r=O.lO, PcO.05) and urban
subjects (men, r=0.12; women, ~0.10, P<O.Ol).
Milk intake also showed significant correlation in
rural (men, r=0.09; women, r=0.08, PcO.05) as
well as in urban subjects (men, r=0.08; women,
r=0.07, PcO.05). The consumption of meat and
eggs showed significant correlation in urban men
(r=0.09, PcO.01) but not in women and rural
subjects.
Multivariate and univariate regression analysis
was performed to find out the relation of food intakes
and other risk factors with coronary disease. Odds
ratio and 95% confidence intervals have been calcu-
lated before and after incorporation of age in the
regression equation as shown in Table 5. The results
showed that after adjustment of age and sex, seden-
tariness in rural and urban women were significant
risk factors of coronary disease. Clarified butter plus
trans fatty acids and milk intake were significant risk
factors of coronary disease in both rural and urban
subjects in both sexes. Body mass index was a
significant risk factor in urban men and women.
4. Discussion
This study showed that the overall prevalence of
coronary artery disease was -3 times higher in the
urban population compared to rural subjects (9.0 vs.
3.2%). The prevalence rates of coronary disease
were significantly higher in subjects consuming
moderate and high visible fat diets compared to a
low visible fat diet (Table 2). In one study in 1975,
the prevalence of coronary artery disease was sig-
nificantly higher among urban compared to rural
subjects (4.5 vs. 1.7%) [9]. In a preliminary study
Table 4 a
Food intakes and risk factors in subjects with coronary artery disease. Values are the mean with the standard deviation in parentheses b
9
Risk factors Men Women D;s
Rural Urban Rural Urban 9
CAD Without CAD Without CAD Without CAD Without F-
,
CAD CAD CAD CAD 2
B
Number of subjects 3.5 859 100 804 23 852 63 839 3
Age b-s) 47.0 (12)* 39.0 (11.5) 48.1 (13) 39.5 (14) 46.6 (8.0)* 38.5 (11.5) 48.0 (9.6)* 37.6 (12.5) k
Body mass index (kg/m’) 22.9 (3.0)’ 21.0 (3.5) 23.8 (3.7)* 21.7 (2.8) 22.6 (2.5)* 21.0 (3.6) 23.5 (2.7)* 21.3 (3.6) k
Total fat (kcallday) 25.0 (2.6)* 15.2 (3.7) 27.5 (3.2)* 25.4 (3.5) 24.3 (2.4)* 15.0 (3.1) 27.0 (2.8)* 25.1 (3.71 2
Saturated fat (k&/day) 12.2 (2.2)* 5.1 (1.2) 13.6 (2.5)* 10.1 (2.7) 12.0 (3.0)* 4.8 (2.2) 13.8 (2.4)* 10.0 (2.5)
Cholesterol (mg/day) 165 (15)** 61 (8) 205 (21)* 162 (25) 154 (13)** 56 (14) 196 (18)* 156 (23) 3k
Milk (g/day) 258 (16)* 165 (14) 325 (20)* 252 (17) 240 (16)* 152 (22) 305 (18.2)* 217 (14.0) q
Meat and eggs (g/day) 22 (3.7)* 9.7 (5) 38 (5)* 25 (4.8) 20 (3.5)* 8.5 (2.5) 34 (5.2)* 23 (4.7) 2
Sugar and jaggery (g/day) 29 (4.0)* 19 (3.2) 38 (4.7)* 28 (3.8) 26 (4.1)* 15 (3.8) 36 (4.2)* 27 (5.0) &
Total fruit, vegetable, pulses 90 (7.0) 84 (7.6) 200 (22) 194 (23) 70 (8) 64 (10) 186 (21) 182 (20) 5
Total fruit, vegetable, pulses/visible fat ratio 5.0 (0.9) 4.6 (1.1) 6.8 (1.2) 6.6 (1.3) 4.6 (0.9) 4.2 (1.1) 8.0 (1.2) 7.2 (1.3) 5
Smoking 20 (57.1)* 280 (32.5) 30 (30)* 15.8 (19.6) 3 (13.0)* 48 (5.6) 2 (3.1) 13 (1.5) ‘3
Sedentariness** 33 (94.2)** 93 (10.8) 95 (95.0)** 353 (43.9) 22 (95.6)** 206 (47.6) 60 (95.2)“” 475 (56.6) 2
Higher and middle socioeconomic status 32 (91.4)** 262 (30.5) 92 (92.0)* 573 (71.2) 21 (91.3)*** 206 (24.1) 58 (92.0)* 580 (69.1) 8
More education (>5 years) 25 (71.4)* 465 (54.1) 95 (95.0)* 669 (83.2) 18 (78.2)* 510 (59.8) 55 (87.3)* 656 (78.1)
P value obtained by comparision of coronary artery disease with rest of the subjects in both rural and urban groups. *P<O.OS; **P<O.Ol. %
 R.B. Singh et al. I International Journal of Cardiology 56 (1996) 289-298 295

Table 5
Association of coronary artery disease prevalence with risk factors and food intakes by univariate and multivariate logistic regression
analysis with/without adjustment with age
Risk factor Men Women
Rural Urban Rural Urban
Odd’s ratio (95% C.I.) Odd’s ratio (95% C.I.) Odd’s ratio (95% C.I.) Odd’s ratio (95% Cl.)
Age 1.48 (1.27-l .67)** 1.52 (1.32-1.76)** 1.26 (1.07-1.47)* 1.32 (1.12-1.56)
Smoking
Unadjusted 1.36 (1.06-1.68)* 1.41 (1.26-1.65)* 0.91 (0.57-1.24) 0.96 (0.68- 1.22)
Adjusted 1.25 (0.88-1.89) 1.28 (0.92-l .92) 0.96 (0.62-l .42) 0.99 (0.72-I .36)
Sedentariness
Unadjusted 1.68 (1.05-2.62) 1.70 (1.08-2.75) 3.36 (2.23-5.88)* 4.12 (2.86-5.12)**
Adjusted 1.76 (1.08-2.85) 1.78 (1.10-2.82) 3.55 (2.45-5.15)* 4.34 (2.91-5.56)**
Body Mass Index
Unadjusted 1.08 (0.98-1.12) 1.32 (1.18-1.45)* 1.04 (0.92-1.08) 1.28 (1.11-1.41)
Adjusted 1.06 (0.95-1.10) 1.25 (1.12-1.38)* 1.01 (0.88-1.06) 1.25 (1.08-1.38)*
Milk
Unadjusted 1.52 (1.32-1.78)* 1.56 (1.34-l .75)* 1.28 (1.12-1.46)* 1.36 (1.15-1.51)”
Adjusted 1.41 (1.24-1.67)* 1.41 (1.26-1.66)* 1.25 (1.09-1.41)* 1.28 (1.12-1.48)*

Clarified butter plus hydrogenated fat

Unadjusted 1.65 (1.35-1.91)* 1.51 (1.28-1.69)** 3.12 (2.51-6.12)* 3.52 (2.12-4.56)*
Adjusted 1.58 (1.31-1.91)* 1.48 (1.25-1.64)** 2.98 (2.41-6.01)* 3.40 (2.08-4.42)*
Flesh foods and eggs
Unadjusted 1.07 (0.91-1.67) 1.12 (0.92-1.72) 0.98 (0.52-1.46) 1.05 (0.68-l .72)
Adjusted 1.03 (0.82- 1.58) 1.10 (0.88-1.71) 0.92 (0.47-1.41) 0.99 (0.62-1.65)
C.I., confidence interval; *P<O.O5; **P<O.Oll

[5], the prevalence of coronary artery disease (8.6 vs.
3.0%) and coronary risk factors were 2-3 times
higher among urban than rural populations.
The higher prevalence of coronary disease in
urban compared to rural subjects indicate that some
component of diet and lifestyle may be important.
Within subgroups consuming moderate and high
visible fat diets, the prevalence of coronary disease
was significantly higher in subgroups consuming
Indian ghee (clarified butter) plus tram fatty acids
and clarified butter as total visible fat compared to
those eating clarified butter plus vegetable oils. An
overall association of higher prevalence of coronary
artery disease with greater amounts of total visible
fat and clarified butter and tram fatty acids intake
was observed in all the subgroups and the trend was
significant for total prevalence rates both for rural
and urban subjects in both sexes (Table 2 and Table
3). The relation of clarified butter plus trans fatty
acids intake with coronary disease remained signifi-
cant after adjustment of age and other risk factors in
a univariate and multivariate analysis (Table 5).
A number of studies have examined the relation-
ship between tram fatty acids in the diet and serum
lipids and lipoprotein levels, however there is scant
evidence on its relation with coronary disease.
Recent studies have found that trans fatty acids
increase low density lipoprotein cholesterol [24,25]
and lipoprotein(a) [26,27] and reduce high density
lipoprotein cholesterol [24,25] and may be associated
with increased mortality due to coronary artery
disease [23]. It is possible that higher lipoprotein(a)
levels observed in Indians [15] and insulin resist-
ance, at least in part, may be because of rapid
increase in the consumption of trans fatty acids and
also possibly due to clarified butter intake. Indian
ghee (clarified butter) which is prepared by heating
the butter contains a substantial amount of oxidized
cholesterol which is more atherogenic [ 121. It has
been suggested that the oxidized form of low density
296 R.B. Singh et al. I International
lipoprotein cholesterol is more atherogenic than
native low density lipoprotein cholesterol [28,29].
In such category of fat intake, the urban popula-
tion showed a relatively higher prevalence of cor-
onary artery disease compared to rural subjects
(Table 2 and Table 3). It is possible that higher
intake of clarified butter and trans fatty acids within
the same category of fat intake and other dietary
factors such as a higher intake of milk (lactose) [30],
sugar and jaggery (sucrose) [2], meat and eggs
(saturated fat and cholesterol) [2] in urban and low
physical activity may be responsible for the higher
prevalence of coronary artery disease (Table 1). A
lower prevalence in rural subjects within each sub-
group may be because they mainly consume mustard
oil and cereals which are rich in n-3 fatty acids and
vitamin E known for a beneficial effect on coronary
disease [29,31]. In brief, the urban diet is more
atherogenic and thrombogenic and predisposes ox-
idative stress compared to rural diet which is more
cereal based and has less of these adverse effects
(Table 1). Among vegetable oils, urban subjects
consume mainly refined oils of ground nut, sun-
flower and soyabean [21]. The diets of Indian
immigrants to Britain appear to be highly rich in
several of the adverse factors of the Indian urban diet
causing greater risk of coronary disease [ 11,131,
although these authors reported that diet was not
associated with coronary risk. Table 3 shows that
moderate to high intake of clarified butter (>20
g/day) was also associated with a higher prevalence
of coronary artery disease compared to the subgroup
consuming also vegetable oils. It is possible that
adverse effects of clarified butter are independent of
trans fatty acids intake and addition of vegetable oils
to the diet lowers the risk of coronary artery disease.
There is evidence that fruits, vegetables and
legumes because of their antioxidant vitamin E and
C and beta-carotene and flavonoids content may have
a beneficial effect on coronary disease [ 14,28,29]. In
our study, the fruit, vegetable and pulse intake was
only one half of World Health Organization advice
(400 g/day) for prevention of chronic disease 121.
However deficiency of fruit, vegetable and pulse
consumption was not associated with risk of cor-
onary artery disease. Rather, fruit intake was 4 times
higher in urban compared to rural subjects. However
Journal of Cardiology .56 (1996) 289-298
it is not clear why increased intake of fruits failed to
provide benefit to urban subjects. It seems that apart
from dietary fat, low physical activity may be the
cause of a higher prevalence of coronary disease in
urban subjects despite a higher intake of fruits.
Similarly higher physical activity may be a potential
protective factor in the rural subjects.
Age, body mass index and the consumption of
total and saturated fat and cholesterol, milk, flesh
foods, eggs, sugar and jaggery were significantly
higher in subjects with coronary disease compared to
the rest of the subjects in both rural and urban
subjects in both sexes. The prevalence of higher and
middle socioeconomic status and education and
sedentary lifestyle as well as tobacco consumption in
men and rural women were also significantly associ-
ated in patients with coronary artery disease. How-
ever univariate and multivariate regression analysis
showed that after adjustment of age, milk and
clarified butter and trans fatty acids intake were the
risk factors of coronary artery disease in both rural
and urban men and women. Body mass index was
the risk factor in urban in both sexes whereas
sedentariness was the risk factor of coronary disease
in rural and urban women. The climate in
Moradabad is very hot in summer (43°C) and very
cold in winter (6°C) which is associated with a
higher admission rate of patients with coronary
artery disease.
The findings of this study suggest that relatively
higher consumption of tram fatty acids [32,33],
clarified butter [ 121 and milk [30] in conjunction
with sedentary lifestyle and higher body mass index
are significant risk factors of coronary artery disease
in Indians. It is possible that some rural and urban
populations of India can benefit by lower consump-
tion of visible fat (20 g/day) and by increased
physical activity and cessation of tobacco consump-
tion for prevention of coronary artery disease.
Acknowledgments
Acknowledgements are due to the Sandoz Founda-
tion of Gerontologic Research, Austratia for financial
support.
 R.B. Singh et al. I lntemational Journal of Cardiology
References
[I] Enas EA, Yusuf S, Mehta JL. Prevalence of coronary artery
disease in Asian Indians. Am J Cardiol 1992; 70: 945-949.
[2] World Health Organization Study Group. Diet, Nutrition and
Prevention of Chronic Diseases. Geneva: World Health
Organization, 1990.
[3] Gupta R, Gupta VP, Ahluwalia NS. Educational status,
coronary heart disease and coronary risk factors prevalence
in a rural population of India. Br Med J 1994; 309: 1332-
1336.
[4] Wander GS, Khurana SB, Gulati R, Sachar RK, Gupta RK,
Khurana S, Anand IS. Epidemiologic study of coronary heart
disease in a rural Punjab population. Prevalence and correla-
tion with various risk factors. Indian Heart J 1994; 46:
3 19-329.
[5] Singh RB, Ghosh S, Niaz MA, Gupta S, Bishnoi I, Sharma
JP, Agarwal P, Rastogi SS, Beegom R, Chibo H, Shoumin Z.
Epidemiologic study of diet and coronary risk factors in
relation to central obesity and insulin levels in the rural and
urban populations of north India. lnt J Cardiol 1995; 47:
245-255.
[6] Beegom R, Singh RB. Prevalence of coronary heart disease
and its risk factors in south and north India. Acta Cardiol
1995; 50: 227-240.
[7] Gupta R, Prakash H, Majumdar S, Sharma S, Gupta VP.
Prevalence of coronary heart disease and coronary risk
factors in an urban population of Rajasthan. lnd Heart J
1995; 47: 331-338.
[S] Singh RB, Niaz MA, Ghosh S, Beegom R, Bishnoi I,
Agarwal P, Singh R, Srivastav B, Rastogi SS, Postiglione A,
Shoumin Z. Epidemiological study of coronary heart disease
and its risk factors in an elderly urban population of north
India. J Am Co11 Nutr 1995; 14: 628-634.
[9] Gupta SP, Malhotra KC. Urban rural trends in the epidemiol-
ogy of coronary heart disease. J Assoc Phys India 1975; 23:
885-893.
[lo] Singh RB, Niaz MA. Coronary risk factors in Indians. Lancet
1995; 346: 778-779.
[l l] Mckeigue PM, Marmot MG, Adelstein AM. Diet and risk
factors of coronary artery disease in Asians in north-west
London. Lancet 1985; ii: 1086-1090.
[12] Jacobson MS. Cholesterol oxides in Indian ghee possible
cause of unexplained high risk of atherosclerosis in Indian
immigrant populations. Lancet 1987; ii: 656-657.
[13] Mckeigue PM, Shah B, Marmot MG. Relation of central
obesity and insulin resistance with high diabetes prevalence
and cardiovascular risk in south Asians. Lancet 1991; 337:
382-386.
[14] Singh RB, Ghosh S, Niaz MA, Singh R, Beegom R, Chibo
H, Shoumin Z, Postiglione A. Dietary intakes, plasma level
of antioxidant vitamins and oxidative stress in relation to
coronary artery disease. Am J Cardiol 1995; 76: 1233-1238.
[15] Bhatnagar D, Anand IS, Durrington PN, Pate1 DJ, Wander
GS, Mackness Ml, Creed F, Tomenson B, Chandrashekhar
56 (1996) 289-298 297
Y, Winterbotham M, Brit RP, Keil JE, Sutton GG. Coronary
risk factors in people from Indian subcontinent living in West
London and their siblings in India. Lancet 1995; 345: 405-
409.
[16] Singh RB, Niaz MA, Ghosh S, Rastogi SS, Beegom R,
Agarwal P, Sharma CLN. Serum cholesterol and coronary
artery disease in a population with lower fat intake. The
Indian Lifestyle and Heart Study. Brit Med J 1996; (in
press).
[ 171 Directorate of Census operations. A portrait of population -
Utter Pradesh, New Delhi (India). Controller of Publications,
Government of India Press, 1991.
[ 181 Rose G, Blackbum H, Gillum RF, Prineas RJ. Cardiovascu-
lar survey methods. Geneva: World Health Organization,
1982.
[19] Narsingrao BS, Deosthale YG, Pant KC. Nutrient Com-
position of Indian Foods. Hyderabad, India: National lnsti-
tute of Nutrition, 1989.
[20] Indian Consensus Group. Indian consensus for prevention of
coronary artery disease: a joint scientific statement of Indian
Society of Hypertension and International College of Nut%
tion. J Nutr Environ Med 1996; (in press).
[21] Expert Group of the Indian Council of Medical Research.
Nutrient Requirements and Recommended Dietary Allow-
ances for Indians. New Delhi, India: Indian Council of
Medical Research, 1990.
[22] Singh RB, Ghosh S, Niaz MA, Rastogi V. Validation of
physical activity and socioeconomic status questionnaire in
relation to food intakes for the five city study and a proposed
classification for Indians. J Asso Phys lndia 1997; (in press).
[23] Kromhout D, Keys A, Menotti A. Dietary saturated and trans
fatty acids, cholesterol and 25-year mortality from coronary
heart disease: the Seven Countries Study. Prev Med 1995:
24: 308-315.
[24] Judd JT, Clavidence BA, Muesing RA, Wittes J, Sunkin ME,
Podczasy JJ. Dietary trans fatty acids: effects on plasma
lipids and lipoproteins of healthy men and women. Am J
Clin Nutr 1994; 59: 861-868.
[25] Mensink RP, Katan MB. Effect of dietary tram fatty acids on
high density and low density lipoprotein cholesterol levels in
healthy subjects. N Engl J Med 1990; 323: 439-445.
[26] Nestel P, Noakes M, Belling B. Plasma lipoprotein lipid and
Lp(a) changes with substitution of elaidic acid for oleic acid
in the diet. J Lipid Res 1992; 33: 1029-1036.
[27] Mensink R, Zock PL, Katan MB, Homstra G. Effect of
dietary cis and tram fatty acids on serum lipoprotein (a)
levels in humans. J Lipid Res 1992; 33: 1493-1501.
[28] Steinberg D, Parthasarthy S, Carew TE, Khoo JC, Witzhim
JL. Beyond cholesterol modifications of low density lipopro-
tein that increase its atherogenicity. N Engl J Med 1989; 320:
915-924.
[29] Witztum JL. The oxidation hypothesis of atherosclerosis.
Lancet 1994; 344: 793-795.
[30] Segall JJ. Dietary lactose as a possible risk factor for
ischaemic heart disease: review of epidemiology. lnt J
Cardiol 1994; 46: 197-207.
298 R.B. Singh et al. I International Journal of Cardiology S6 (1996) 289~-298

[31] de Largeril M, Renand S, Mamelle N, Salen P, Martin JL, [32] Enic MC. Tram fatty acid in diets and datahaseh. Cereal
Monjand I, Guidoilet J, Tonboul P, Dealye J. Mediterranean Foods World 1996; 41: 58-63.
alphalinolinic acid-rich diet in secondary prevention of [33] Mann GV. Metabolic consequences of tram fatty Audi.
coronary heart disease. Lancet 1994; 343: 1454-1458. Lancet 1994; 343: 1268-1271.