Dietary Patterns and Metabolic Risk Factors for Cardiovascular Disease among University Students in Ghana
Aaron O. Amankwaa
Reginald A. Annan
Dietary intake is a strong modifiable index for Metabolic Risk Factors (MRF) for atherosclerotic cardiovascular disease. Poor dietary intake has been shown among Ghanaian university students but little is known about how that is related to MRF. This study examined dietary patterns and MRF among KNUST students. Data were collected from 66 randomly selected undergraduate students (31 males and 35 females) from 18-27 years old. Dietary intake was assessed using an 81 item food frequency questionnaire and patterns of dietary intake by K-means cluster analysis. The dietary patterns (clusters) were related to metabolic risk components including fasting blood lipids, fasting blood sugar, blood pressure and waist circumference. In this population, 34.3% of females and 9.7% males had low HDL; 22.9% females and 9.7% males had high LDL; and 11.4% females and 6.5% males had high total cholesterol. Three unique dietary (clusters) patterns were identified, termed: broad, convenience and ordinary. The broad cluster, characterised by a varied dietary intake of most food groups and good intake of fruits and vegetables was associated with more physically active (41.9%) participants than the others. The ordinary cluster, characterized by whole grains, meat and dairy products but low intake of fruits and vegetables, had more participants who drank alcohol (38.9%) and ate out (32.9%). The convenience, characterised by very little to no fruits and vegetables intake, frequent skipping of breakfast (46.2%) and high beverage consumption was associated with higher serum triglyceride (p = 0.021) and total cholesterol (p = 0.005) among male participants. Between the three, the broad cluster seemed the best regarding types and frequencies of foods consumed and association with MRFs whereas the convenience cluster seemed the worst. In conclusion, our findings showed significant prevalence of MRF for CVDs in this young population and distinct dietary patterns, which were associated with these risk factors.
Received: January 10, 2014;
Accepted: March 03, 2014;
Published: April 15, 2014
Cardiovascular Disease (CVD), including Coronary Heart Disease (CHD), stroke
and heart failure, is associated with 16.7 million deaths in the world every
year (Keyhani et al., 2013) with 80% of all
CVD deaths occurring in developing countries (WHO, 2012). In 2001, 9.2% of all
deaths in Africa were caused by CVD (Abanilla et al.,
2011) but it is estimated that CVD will cause 23 million deaths in the world
by 2030 (WHO, 2012) In Ghana, CVD has been reported to
be a leading cause of death in some parts of the country (Aikins,
2007). CVD was the predominant cause of death in Accra, Ghana, in 1991 and
2001 (Aikins, 2007). Hypertension, heart failure and
stroke account for 17.9% of acute medical admissions at Komfo Anokye Teaching
Hospital (KATH), a major referral hospital in Ghana (Plange-Rhule
et al., 1999). More recently, the national prevalence of hypertension
(one of the major causes of CVD in Ghana) has been estimated at 28.7% (Aikins,
Unhealthy diet, smoking and physical inactivity are among lifestyles that are
now known to be major risk factors for CVDs (WHO, 2012;
Berg et al., 2008). Stroke, CHD and peripheral
artery disease are directly affected by unhealthy diet (Anderson,
2003). These lifestyle factors predispose individuals to metabolic risk,
including low serum HDL cholesterol, elevated serum triglyceride, high waist
circumference, high blood pressure and high fasting blood glucose which ultimately
predispose people to the development of hypertension and atherosclerosis (Afoakwah
and Owusu, 2011; Onwuanyi et al., 1998).
These Metabolic Risk Factors (MRFs) are now estimated in patients to examine
risk of developing diet-related CVDs (Lorenzo et al.,
Generally, dietary patterns characterized by high intake of fatty foods and
refined carbohydrates and a low consumption of fruits and vegetables are commonly
reported to be the causes of MRFs for CVDs (Berg et
al., 2008). In many developing countries, there is currently a shift
in patterns of dietary intake towards more westernised foods (Bourne
et al., 2002) characterised by increased intake of saturated fat,
salt and refined foods and decreased fruits, vegetables and fibre intake ultimately
increasing the incidence of diet-related chronic diseases (Slattery
et al., 1998; Hu et al., 2000). Increasing
consumption of westernized foods high in fats, sugars and salt is a major contributor
to the increasing prevalence of CVD in Ghana (Ministry of
Health, 2010). In Ghana, the percentage of adults who consume sufficient
amounts of fruits and vegetables is less than 5% (Government
of Ghana, 2012).
Young Ghanaian university students are not left out in the current observed
trend. University students are highly exposed to unhealthy eating habits, because
of lack of time for cooking, stress and poor awareness. High consumption of
fried foods, eggs sandwich for breakfast and rice for lunch and supper have
been reported among students in Ghana (Danquah et al.,
2010). Because diet-related chronic diseases begin early in life, their
prevention must also begin early (Hong, 2010) and dietary
patterns of university-age adults are of great concern to achieve this.
There is a call on governments in Africa to draw effective policies to control
the high prevalence of CVD deaths on the continent. The Ghana Health Service
in heeding to this call has initiated interventions among school youth in the
country with emphasis on healthy dietary behavior, low consumption of energy
dense foods, salts, trans fatty acids and sugar to ensure primary prevention
of CVD (Government of Ghana, 2012). However, limited
data exist on dietary intake among Ghanaian university students and its associated
effects to serve as justification for specific interventions. The study sought
to characterise the diet and determine the relationship between dietary patterns
and metabolic risk factors for CVD among KNUST students.
This study was a cross-sectional study conducted between January and May, 2013
at Kwame Nkrumah University of Science and Technology (KNUST), a tertiary institution
with about 20000 under-graduate students located in Kumasi, Ghana. The participants
of the study were selected randomly from the University hostel facilities. A
total of 66 students gave their consent to participate in the study.
An approved research information leaflet and consent form was provided to prospective
participants to decide if they would like to be part of the study. Participants
were required to sign the written consent form, which was documented systematically
using code numbers. Students aged 18 years and above who gave their consent
were eligible for the study. Pregnant female students were excluded from the
A detailed questionnaire was then used to collect demographic data of eligible
subjects. The questionnaire also captured information about health, lifestyle
factors and family history of CVD. All consented participants were invited to
the Clinical Analysis Laboratory (CAnLAB) of the Department of Biochemistry
and Biotechnology, KNUST the next day for physical examination and donation
of fasting blood sample after completion of the health history and dietary questionnaires.
This study was approved by the Committee on Human Research Publication and Ethics
(CHRPE) of the School of Medical Sciences, KNUST and Komfo Anokye Teaching Hospitals.
Dietary assessment: Information on food intake pattern over a period
of three months was assessed with a self-administered Food-Frequency Questionnaire
(FFQ) designed to reflect the foods commonly consumed by KNUST students. Food
items on the FFQ were obtained from results of a focus group discussion conducted
among 10 KNUST students (6 female and 4 male students). The FFQ included a list
of 81 food items with 8 frequency categories of intake ranging from daily
Assessment of metabolic risk factors for CVD: Metabolic risk factors
for CVD were characterized by the components of the metabolic syndrome defined
by the National Cholesterol Education Program, Adult Treatment Panel III (Lorenzo
et al., 2007). These components include Waist Circumference (WC),
Blood Pressure (BP), High Density Lipoprotein cholesterol (HDL-c), serum triglycerides
(TG) and fasting blood glucose (FBS) (Lorenzo et al.,
2007). Waist circumference (cm) of each subject was measured using a plastic
tape measurement at a level just above the navel with participants wearing no
shoes. A manual aneroid sphygmomanometer with an appropriate size of cuff was
used to take two measurements each of systolic and diastolic blood pressures
(at least 20 minutes apart) in mmHg. The average of the two measurements was
used for analysis.
Five milliliters of blood sample was collected from each subject by venipuncture
after an 8-12 h overnight fast and the serum was separated. The separated serum
was analyzed for total cholesterol and triglycerides using enzymatic methods
designed by fortress diagnostics Limited, United Kingdom. Serum HDL cholesterol
was measured after phosphotungstic acid-magnesium chloride precipitation of
LDL and VLDL. The semi-micro assay method was used in the precipitation procedure.
The concentration of the HDL cholesterol was determined using the CHOD-PAP (Total
cholesterol) method. The concentration of low-density lipoprotein cholesterol
(LDL) was calculated using the Friedwald formula. The method was repeated for
all samples. Serum samples from specimens in fluoride tubes were also analyzed
for glucose using the glucose oxidase method designed by fortress diagnostics
Limited, United Kingdom.
Data analysis: Statistical analysis of data was done using SPSS software,
version 16.0.1 (SPSS Inc., Chicago IL, USA). Scores were provided from the responses
of the FFQ data as follows: never = 1, occasionally = 2, once per month = 3,
once per two weeks = 4, once per week = 5, two-three times per week = 6, four-six
times per week = 7 and daily = 8. The major food patterns of students were identified
using K-means cluster analysis. The criteria below were adopted to categorize
metabolic risk factors as follows: Waist Circumference: Men ≥102 cm, Women
≥88 cm; Blood pressure: Systolic ≥130 mmHg, Diastolic ≥85 mmHg, Total
Cholesterol: ≥5.2 mmol L-1, HDL cholesterol: Men ≤1.0 mmol
L-1, Women ≤1.3 mmol L-1, LDL cholesterol: >2.6
mmol L-1, Serum triglycerides: ≥1.7 mmol L-1; Fasting
Blood glucose: ≥6.1 mmol L-1 (Lorenzo et
al., 2007). The ranges above are standards used to determine if the
measures are within normal levels or not. Where necessary, analysis was stratified
by gender or controlled for possible confounders. One-way ANOVA was used to
determine the significance of the mean values of metabolic risk factors between
the major dietary patterns. Chi-squared test was used to determine significance
of prevalence metabolic risk factors for CVD among the major food clusters.
Descriptive characteristics of study population: Sixty-six participants
took part in the study, 31 (47) were males and 35 (53%) were females. Study
participants were from 18-27 years old, with mean age of 21 (±2.2) years,
whilst the males slightly older than the females (p = 0.171) (Table
1). All participants were non-smokers, 5 (7.6) drank alcohol, 52 (78.8%)
participants skipped breakfast frequently and 18 (27.3%) ate at restaurants
or fast foods 3 times or more per week. Thirty one (47%) participants engaged
in 30 min or more of moderate physical activity on most days of the week; the
males had a higher rate than the females (p = 0.007). There was no significant
sex difference in the mean DBP but the mean SBP was significantly higher in
the males (110 mmHg) than in the female (104 mmHg) students (p = 0.003). The
means FBS was 4.67 (±0.84) mmol L-1, serum TC was 3.78 (±0.77)
mmol L-1, TG was 0.61 (±0.31) mmol L-1, HDL was
1.53 (±0.38) mmol L-1 and LDL was 1.97 (±0.77) mmol
L-1. There were no significant differences between males and females.
Prevalence of metabolic risk factors for cardiovascular disease among participants
is presented in Table 2. Prevalence of high FBS was 1.5%,
high TC was 9.1%, high TG was 1.5% and low HDL was 22.7% Prevalence of high
LDL was 16.7%, high DBP 12.1 and high WC being 3%. None of the participants
SBP considered high. The frequencies of high FBS, TC, TG, LDL, DBP and WC were
not statistically significant by gender but the prevalence of low HDL was significantly
(p = 0.017) higher in females (34.3%) than in males (9.7%).
||Characteristics of participants by sex
||Prevalence of metabolic risk factors for CVD by sex
||Relating dietary patterns and some background characteristics
Patterns of dietary intake: Three distinct dietary (clusters) patterns
were identified by K-means cluster analysis; broad,
convenience and ordinary
(Table 3). The food clusters were named according to the types
of foods and their frequencies consumed. The broad cluster, on the average,
was characterized by consumption of tomatoes stew and cocoa powder drinks (2-3
times per week) and weekly consumption of bread, fried fish, tinned fish, oranges,
fried ripe plantain, jollof
or fried rice, fried and boiled egg, kenkey
fruit juice and cakes or sweet biscuits consumption. In the convenience cluster,
cocoa powder drinks (such as MiloTM, RichocoTM and ChocolimTM)
were the only food items consumed once per week on the average. All other foods
were consumed less than once per week. The ordinary cluster was associated with
2-3 times per week consumption of tomatoes stew. Cocoa powder drinks, fried
egg, sugar bread, fruit juice, fried chicken, kenkey, jollof/ fried rice, butter
bread and fried fish were consumed once a week on the average. All other food
items were consumed less than once a week.
Among the three clusters, the broad group was associated with better consumption
of varied food groups and fruits and vegetables compared with the others. The
ordinary group had fewer foods and food groups where as the convenient group
was mainly cocoa drink (Table 3).
Association between dietary clusters and lifestyle characteristics:
Relating the dietary cluster with some lifestyle factors (Fig.
1), it is shown that the convenient dietary cluster was associated with
significantly more males (45.2 versus 32.3 for broad and 22.6% for ordinary),
breakfast skippers (46.2 versus 28.8 for broad and 25% for ordinary: p<0.028)
and more subjects (37.9%) compared with the other clusters. The ordinary cluster
was associated with more alcohol drinkers and those who ate outside although
these differences were not significant. The broad cluster on the other hand
was associated with significantly more participants who exercised regularly
(41.9%) compared with the convenient (32.3%) and ordinary cluster (25.8%).
||Relating dietary patterns by mean values of metabolic risk
Association between dietary clusters and prevalence of metabolic risk:
The prevalence rates of the metabolic risk factors for cardiovascular disease
are presented in Table 4. Although, statistically not significant,
there were more participants in the convenient cluster who had high LDL (24%
convenient versus 15.9% ordinary and 9% broad cluster), high DBP (20% convenient
versus 5.3% ordinary and 9.1% broad cluster) and high TC (16 convenient versus
5.3 ordinary and 4.5% broad cluster) than the other clusters. There were also
more participants in the broad cluster who had high HDL than the other clusters.
The total number of participants who suffered from at least one metabolic risk
factor in each cluster was 80% for the convenience
cluster, 63.6% for the broad
cluster and 52.6% for the ordinary
Relating dietary clusters by means of metabolic risk indices: There
was a significant effect of dietary pattern on triglyceride level for the three
clusters [F (2,63) = 4.117, p = 0.021] Table 5. Post hoc comparisons
showed that the mean triglyceride value for the convenience
cluster (0.74±0.40) was significantly higher than the ordinary
cluster (0.51±0.16) (p = 0.034). However, the broad
cluster (0.54±0.24) did not differ significantly from the two clusters.
There was no significant difference in the HDL cholesterol, LDL cholesterol,
fasting blood glucose, systolic blood pressure, diastolic blood pressure and
waist circumference between the three dietary patterns.
Relating dietary clusters by means of metabolic risk indices stratified
by gender: Dietary pattern had a significant effect on total cholesterol
and triglyceride level among males [F (2,28) = 4.078, p = 0.028; F (2, 28) =
6.373, p = 0.005 respectively]. Among the males, mean TG and TC were significantly
lower among the broad cluster than the rest (Table 6). Post
hoc comparisons showed that the mean total cholesterol for the convenience
cluster (3.93±0.86) was significantly higher than the broad
cluster (3.13±0.46) (p = 0.021) but the ordinary cluster
(3.61±0.45) did not differ significantly from the two clusters.
||Relating dietary patterns by mean values of metabolic risk
factors, stratified by gender
Again, post hoc comparison showed that the mean triglyceride level for the
convenience cluster (0.86±0.37) was significantly higher
than both the broad cluster (0.48±0.24) (p = 0.012) and the
ordinary cluster (0.47±0.17) (p = 0.02). The ordinary
cluster did not differ significantly from the broad cluster. Among
the females, there was no statistical difference in the mean values for MRFs
between the dietary patterns.
In this study, there was as significant prevalence of metabolic risk for cardiovascular
disease in this young population and this was higher in females. Close to 10%
had high total cholesterol, 16.7% had high LDL and about a quarter had low HDL.
It is notable that these parameters are already of concern in this young population.
Moreover, prevalence of most of the MRF were higher in females, with low HDL
recording significant differences between females and males and this shows that
the female had higher risks. Hyperglycaemia, hypertension, abdominal obesity
and lipid disorders are generally more common in older people (Nkum
and Micah, 2012) and the low prevalence observed in this study could be
attributed to the age group studied but we have observed that these problems
are occurring in much younger ages now.
Three unique dietary patterns were identified in this study of which no significant
difference in the prevalence of metabolic risk factors for CVD were identified.
None of the three patterns seemed to meet dietary guidelines, which would help
reduce cardiovascular disease risk (Kelly and Stanner,
2003), such as consumption of 5 portions of fruits and vegetables a day,
low dietary fat and energy dense food and consumption of unsaturated fatty acids.
In this study, fruits and vegetables were consumed less than 2-3 times per week
across the three clusters. More than 50% of subjects skipped breakfast regularly,
another critical health-promoting recommendation (Ruxton
and Kirk, 1997). Our results suggest a generally poor dietary lifestyle
and possible dietary inadequacy consistent with other studies among Ghanaian
university students (Danquah et al., 2010).
The ordinary dietary pattern (cluster) was also comparable to a typical Ghanaian
diet, which is characterized by frequent (4.2 days per week) consumption of
maize containing foods such as banku and kenkey, wheat/bread (3days), rice (2.8
days) and very low consumption of meat (less than 2 days per week) (Biederlack
and Rivers, 2009).
Between the three patterns however, the broad group showed a more varied and
balanced diet and was better associated with fruits and vegetables intake than
the others. The convenient pattern seemed the worse although more participants
followed this pattern. The convenience cluster was characterized
by very low consumption of whole grain foods such as wheat bread, pasta and
rice, fruits and vegetables, fish and meat. Majority of persons (96%) in this
cluster also reported to skip breakfast frequently. Consumption of fruits and
vegetables and green tea, known to increase the antioxidant status of the body
thereby reducing the level of oxidized fats in the blood (Tribble,
1999; Cabrera et al., 2006) was more unlikely
among consumers of this pattern.
Many studies have found variations in lipid disorders among different dietary
patterns (Berg et al., 2008; Wirfalt
et al., 2001). Healthy food patterns are normally associated with
low lipid disorders as compared to less healthy dietary patterns (Olinto
et al., 2012). In the present study, mean triglycerides level was
significantly higher in the convenience than the ordinary
cluster (p = 0.021). The broad cluster did not differ significantly,
although the level of triglycerides was lower than the convenience
cluster. Furthermore, when stratified by gender, total cholesterol and triglyceride
levels among males who followed the convenience cluster was significantly
higher (p = 0.028 and p = 0.005 respectively) than the other two clusters, indicating
that the convenience dietary pattern (37.9%) was associated with
increased metabolic risks for CVD among males. It seems the broad cluster was
not different from the ordinary patterns with regards to association with MRF
whereas the convenient pattern was different from both patterns.
Frequent consumption of diets rich in fat, cholesterol and refined carbohydrates
are strongly linked to central obesity (Zhao et al.,
2008). In this study although prevalence of obesity was low, it did seem
some of these young university students had lifestyles that put them on the
road to development of obesity and its consequences in the future such as low
intake of fruits and vegetables, low physical activity and breakfast skipping.
In conclusion, this study identified siginifant prevalence of MRF for CVDs
in this university population. The three unique patterns of dietary intake identified
were all associated with inadequate fruits and vegetables intake but the broad
category was most optimal compared to the other two. Our study also showed possible
association between patterns of dietary intake and MRF, which seem to be apparent
in this young population. Further research among other young Ghanaian populations
are recommended in order to fully understand what is happening and to intervene
Abanilla, P.K.A., K.Y. Huang, D. Shinners, A. Levy, K. Ayernor, A.D.G. Aikins and O. Ogedegbe, 2011. Cardiovascular disease prevention in Ghana: Feasibility of a faith-based organizational approach. Bull. World Health Organiz., 89: 621-700.
Afoakwah, A.N. and W.B. Owusu, 2011. The relationship between dietary intake, body composition and blood pressure in male adult miners in Ghana. Asian J. Clin. Nutr., 3: 1-13.
CrossRef | Direct Link |
Aikins, A.D.G., 2007. Ghana's neglected chronic disease epidemic: A developmental challenge. Ghana Medical J., 41: 154-159.
Anderson, J.W., 2003. Whole grains protect against atherosclerotic cardiovascular disease. Proc. Nutr. Soc., 62: 135-142.
CrossRef | PubMed | Direct Link |
Berg, C.M., G. Lappas, E. Strandhagen, A. Wolk and K. Toren et al., 2008. Food patterns and cardiovascular disease risk factors: The Swedish INTERGENE research program. Am. J. Clin. Nutr., 88: 289-297.
Biederlack, L. and J. Rivers, 2009. Comprehensive food security and vulnerability analysis (CFSVA) Ghana. Republic of Ghana. United Nations World Food Programme, Rome, May 2009. http://www.unglobalpulse.org/sites/default/files/reports/GhanaGov%20Foodsecurityandvulnerability%20analysis%202009.pdf.
Bourne, L.T., E.V. Lambert and K. Steyn, 2002. Where does the black population of South Africa stand on the nutrition transition? Public Health Nutr., 5: 157-162.
CrossRef | PubMed | Direct Link |
Cabrera, C., R. Artacho and R. Gimenez, 2006. Beneficial effects of green tea-A review. J. Am. Coll. Nutr., 25: 79-99.
PubMed | Direct Link |
Danquah, A.O., Y. Odjoji, S. Graham-Acquaah and M. Steiner-Asiedu, 2010. A pilot study of the dietary and physical activity behaviours of international students: Implications for health promotion. Afr. J. Food Sci., 4: 86-92.
Direct Link |
Government of Ghana, 2012. New policy on non-communicable diseases. Ghana Government Portal, Accra.
Hong, M.Y., 2010. Atherosclerotic cardiovascular disease beginning in childhood. Korean Circ. J., 40: 1-9.
CrossRef | PubMed |
Hu, F.B., E.B. Rimm, M.J. Stampfer, A. Ascherio, D. Spiegelman and W.C. Willett, 2000. Prospective study of major dietary patterns and risk of coronary heart disease in men. Am. J. Clinical Nutr., 72: 912-921.
Direct Link |
Kelly, C.N. and S.A. Stanner, 2003. Diet and cardiovascular disease in the UK: Are the messages getting across? Proc. Nutr. Soc., 62: 583-589.
Keyhani, D., M. Kargarfard, N. Sarrafzadegan and M. Sadeghi, 2013. Autonomic function change following a supervised exercise program in patients with congestive heart failure. ARYA Atheroscler., 9: 150-156.
Lorenzo, C., K. Williams, K.J. Hunt and S.M. Haffner, 2007. The national cholesterol education program-adult treatment panel III, international diabetes federation and World Health Organization definitions of the metabolic syndrome as predictors of incident cardiovascular disease and diabetes. Diabetes Care, 30: 8-13.
CrossRef | Direct Link |
Ministry of Health, 2010. Dietary and physical activity guidelines for Ghana. Ministry of Health, Accra. April 2010. http://alwag.org/dahl/dpag.pdf.
Nkum, B.C. and F.B. Micah, 2012. Lipid disorders in hospital attendants in Kumasi, Ghana. Ghana Med. J., 46: 14-21.
Direct Link |
Olinto, M.T., D.P. Gigante, B. Horta, V. Silveira, I. Oliveira and W. Willett, 2012. Major dietary patterns and cardiovascular risk factors among young Brazilian adults. Eur. J. Nutr., 51: 281-291.
CrossRef | PubMed |
Onwuanyi, A., D. Hodges, A. Avancha, L. Weiss, D. Rabinowitz, S. Shea and C.K. Francis, 1998. Hypertensive vascular disease as a cause of death in blacks versus whites: autopsy findings in 587 adults. Hypertension, 31: 1070-1076.
CrossRef | PubMed |
Plange-Rhule, J., R. Phillips, J.W. Acheampong, A.K. Saggar-Malik, F.P. Cappuccio and J.B. Eastwood, 1999. Hypertension and renal failure in Kumasi, Ghana. J. Hum. Hypertens, 13: 37-40.
Ruxton, C.H. and T.R. Kirk, 1997. Breakfast: A review of associations with measures of dietary intake, physiology and biochemistry. Br. J. Nutr., 18: 197-213.
Direct Link |
Slattery, M.L., K.M. Boucher and B.J. Caan, 1998. Eating patterns and risk of colon cancer. Am. J. Epidemiol., 148: 4-16.
Tribble, D.L., 1999. Antioxidant consumption and risk of coronary heart disease: Emphasis on vitamin C, vitamin E and β-carotene. Circulation, 99: 591-599.
Direct Link |
WHO, 2012. Cardiovascular diseases. WHO, Geneva. http://www.who.int/topics/cardiovascular_diseases/en/.
Wirfalt, E., B. Hedblad, B. Gullberg, I. Mattisson and C. Andren et al., 2001. Food patterns and components of the metabolic syndrome in men and women: A cross-sectional study within the Malmo Diet and Cancer cohort. Am. J. Epidemiol., 154: 1150-1159.
Direct Link |
Zhao, S., Y. Chu, C. Zhang, Y. Lin and K. Xu et al., 2008. Diet-induced central obesity and insulin resistance in rabbits. J. Anim. Physiol. Anim. Nutr., 92: 105-111.
CrossRef | PubMed |