Introduction
The World Health Organization (WHO) qualifies obesity as a global epidemic and considers obesity to be a public health problem worldwide. Notably, obesity is the second leading cause of preventable death after smoking [1]. In Africa, many countries have concentrated their efforts on malnutrition, food security, and infectious diseases; hence, the phenomenon of obesity has not been well studied in some countries or populations [2]. However, according to the World Health Organization (WHO), some areas of Sub-Saharan Africa, such as South Africa, experience a prevalence of obesity similar to those of European countries [2-4]. Recent studies have revealed that the burden of obesity is increasing in African countries, wherein the prevalence of obesity rapidly increased during the last two decades of the 20th century, and the prevalence continues to increase in the 21st century [2,4]. Many authors [2,5] indicated that the prevalence of obesity in West Africa between 2000 and 2004 was around 10%. In Senegal, Neupane, et al. [4] reported recently (2016) that 5.1% of general population 5% was obese while 13.3% was overweight; furthermore, obesity seems to affect more women than men [2].
Obesity is a risk factor for diseases such as cardiovascular and metabolic diseases [6-8]. Accumulation of lipids in insulin targets tissues (liver, skeletal muscles) could play an important role in the pathophysiological mechanisms of health status alterations linked to obesity, especially insulin resistance [9-11]. Complications related to excess body weight are numerous and include dyslipidemia, cardiovascular disease, and cancer [8,11]. However, few studies in Sub- Saharan Africa have evaluated the occurrence, mechanisms, morbidity, and mortality related to obesity. Our objective was to study the effect of fat mass on the occurrence of high blood pressure and diabetes among African sedentary adults.
Methodology
Setting
Our study was conducted at the Laboratory of Physiology and Functional Explorations of the Faculty of Medicine, Pharmacy, and Odontology at the University Cheikh Anta Diop of Dakar (UCAD), Senegal, West Africa, from December 2014 to November 2015. The protocol was designed in accordance with the guidelines set by the Declaration of Helsinki and was approved by the Ethics Committee of UCAD.
Sample
Our study population included 330 sedentary adults; was comprised of 96 women and 234 men; and was from the general population of Dakar City (Senegal). Subjects were recruited at Pharmacy Islam (located in City Center of Dakar), which is visited frequently by the Dakar general population to buy medicine or seek advice. Participants were informed of the procedures and purposes of the study and gave written informed consent to participate to the present study. A week later recruitment subjects were called to the Laboratory for measurements.
Measures
A full clinical examination was employed to measure anthropometric parameters (age, height, and body weight); systolic and diastolic arterial pressures (SAP and DAP respectively); and plasma level glucose in our subjects. Mean arterial pressure (MAP) was evaluated by MAP = (SAP + 2PAD)/3, and body mass index (BMI) of the subjects was calculated using Quetelet equation, BMI (kg / m2).
We measured body composition parameters such as fat mass (FM), lean mass (LM), and fat index (FI) and lean index (LI) using Bioelectric Impedance meter (Sokey, No. 8589). Measurement was conducted while the subject stood upright. After the manual introduction of anthropometric parameters (age and sex) of the subject in the apparatus, the measurement then took place as a simple weighing on a usual scale.
All measurements (anthropometrics, biochemical, cardiovascular, and body composition parameters) were performed after eight hours of fasting during the morning of the experimental day (i.e., 8:00 h) at the Laboratory of Physiology and Functional Explorations of the Faculty of Medicine, Pharmacy and Odontology at the University Cheikh Anta DIOP of Dakar (UCAD).
Statistical Analyses
Data were expressed as mean +/- standard deviation, percentages, and relative values. Our study population was divided to three groups according to their BMI: normal subjects (BMI 18 to 24.9 kg / m2) made up Group 1 (G1), the overweight subjects (BMI = 25-29.9 kg / m2) comprised the second group (G2), and the obese subjects (BMI ≥30 Kg/m2) were in the third group (G3).
Chi-square tests were employed to assess the relationship between BMI status and our outcomes of interest. For the comparison of frequencies of arterial hypertension (defined by the standards of the American Heart Association i.e. systolic blood pressure [SBP] ≥ 140 mmHg and / or diastolic blood pressure [DBP] ≥ 90 mm Hg [12]) and diabetes (according to WHO criteria in 2006 [fasting blood glucose ≥ 7.0 mmol/L] )between the different groups, we used the same statistical test (Ch-2 test). For cardiovascular and metabolic variables, we calculated the Odds Ratio (OR) with 95% confidence interval (95%IC) to evaluate the risk of occurrence of these complications in each group. Body composition parameters (FM, LM, FI and LI) were compared between the three groups using a one-way ANOVA with Newman- Keuls for multiple comparisons. We also used a Pearson test for correlations between fat mass and cardiovascular and metabolic parameters such as mean arterial pressure (MAP) and glycemia. State of the outcomes and the independent variables were examined with the linear regression. We then constructed multivariable regression models for FM, FMI, and BMI adjusted for age and sex, with glycaemia and MAP as the dependent variables. The significance level was defined as p < 0.05. Analyses were conducted using Statistical (version 5.5; Stat soft, Tulsa, OK).
Results
Anthropometric characteristics of the study population
Subject characteristics were as follows: mean age: 42.2 +/- 12.4 years, height: 174 +/- 0 9 cm body weight: 86.3 +/- 18.0 kg. Mean BMI was 28.6 +/- 6.2 kg/m2.
Population distribution according to BMI
Normal weight subjects (BMI: 18 - 24.9 kg/m2) represented 28.6% of the study population, and overweight (BMI: 25- 29.9 kg/m2) and obese (≥ 30kg/m2) subjects were 36.3% and 35.1% respectively. No significant difference was observed between the different groups. Women made up 53.91% of the obese group, while men made up 46.19% of the group. No difference in the rate of obesity was found between men and women. Proportion of men to women was 5.46 and 5.21 in normal weight and overweight groups respectively. In obese population the sex ratio (male/female) was 0.85.
Cardiovascular and metabolic risk in the different groups
Comparison of the biochemical data between different groups showed that 34.78% of obese subjects were diabetic compared to 22.03% in overweight subjects (P=0.0596; OR: 0523 95% CI: 0.279 - 0980) and 9.28% in subjects with normal BMI (P<0.0001; OR: 0.183; 95% CI: 0.08- 0.408). Prevalence of diabetic subjects was statistically higher in G2 than in G1 (P=0.018; OR: 0.35; 95% CI: 0.152 to .806) (Table 1). Arterial hypertension was found in 38.26% of obese subjects and in 7.22% of normal weight subjects, and a significant difference was observed between the two values (P<0.0001; OR: 0.122; 95% CI 0.051-0.2926). Proportion of hypertensive subjects in group 2 (7.63%) was not statistically different than G1 (7.22%); however, the prevalence of hypertension in obese subjects was significantly higher than that of overweight subjects (P<0.0001; OR: 0.14; 95% CI: 0.062-0.324) (Table 1). In group 1, all hypertensive individuals were men (p=0.006), and there were no significant difference between women and men with respect to diabetes (Table 2). However, among overweight individuals, high blood pressure was more frequent in women than men (Table 2); 21% of subjects who were hypertensive were women vs 5% for men (p=0.001; OR: 5.051; 95% CI: 1.821-14.01);while, diabetes was less frequently found in women compared to men (10% of subjects were women against 22% of men; p=0.0327; OR: 0.3939; 95%CI: 0.1758-0.8828). In the obese population, no significant difference was observed between women and men for prevalence of diabetes and arterial hypertension.
Body composition in the different groups
Fat mass and fat index were significantly (p<0.001) high in obese subjects (42.6 +/- 13.82% and 40.9 +/- 8.71 respectively) compared to overweight (FM: 27. 7 +/- 4.88% and FI: 26. 8+/- 5.00) or normal subjects (13.1 +/- 4.45% and 19.0 +/- 5.89 % for FM and FI respectively). Nevertheless lean mas and lean index were significantly (p<0.001) most elevated in obese and overweight populations than normal subjects. No significant difference was observed between obese and overweight subjects for LM or LI (Table 3).
Correlations between cardiovascular and metabolic parameters and body composition parameters
Univariates analyses showed that glycaemia was positively correlated with BMI (P<0.0001; r=0.3318); FM (P<0.0001; r=0.3363), FI (P<0.0001; r=0.3268) and systolic (p<0.0001; r=0.2626) and diastolic (p<0.0001; r=0.2862) arterial pressure. However, the linear regression suggested that BMI, FM, and FI were still the only factors independently associated with plasma glucose level (p<0.0001; r2=0.1101; p<0.0001; r2=0.1131 and p<0.0001; r2 = 0.1068 respectively). The same body composition parameters were independently correlated with MAP (BMI: p<0.0001; r2=0.3289; FM: p < 0.0001; r2=0.2866; FI p < 0.0001; r2 = 0. 2494). Besides age was also correlated with MAP (p < 0.0001; r2=0. 06) (Table 4).
Finally, the age and gender adjusted relationship between glycaemia and MAP in relation to BMI, FI and FM was reported in (Table 5) and showed a statistically significant (p<0.001) positive relationship between 1) MAP and BMI, FI & FM and 2) Glycaemia and BMI, FI & FM.
Discussion
Our results showed that 34.35% of study population was obese (Body Mass Index: BMI ≥ 30 kg/m2) with a slight female predominance (53.91%). Higher prevalence of diabetes and arterial hypertension was found in overweight or obese subjects, and the linear regression model analyses showed that BMI, fat mass (FM), and fat index (FI) were the determinants of vascular/metabolic complication occurrence. According to our results, the obesity phenomenon could be a real health problem in the African society, especially Senegalese. Our data confirm those of the literature that showed high prevalence of obesity in the West African region [4,13]. For example, in 2016, Neupane, et al. [4] showed high proportions of overweight or obese people among Sub-Saharan African subjects. Globalization and the current alimentation would be factors favoring the occurrence of obesity in the western part of the African continent. In fact, phenomena of globalization and the improvement of living standards promote an occidentalization of African food that is becoming increasingly rich in saturated fats and refined products [13,14]. A change in lifestyle is also observed in the African population, and this change might be a result of a particular high propensity for physical inactivity, which could contribute to the onset of obesity and overweight in African subjects [4,13,14]. According to our results, the phenomenon of overweight or obesity was slightly more common in women than in men, but not at a significant difference. This observation has been described in the literature [2,15] Indeed, authors [15] have shown that African women are more at risk of becoming obese compared to males. In African populations, the frequency of obesity (BMI ≥ 30 kg / m2) is greater in women essentially due to socio-cultural reasons. In many Sub-Saharan countries such as Senegal, overweight and/or obesity are considered criteria for beauty and well-being in women.
Our study showed other links between increased BMI and the risk of developing complications such as diabetes and/or arterial hypertension. In subjects who are overweight or obese, the prevalence of high blood pressure and diabetes was statistically higher compared to subjects with normal BMI. For example, 38.26% of obese subjects were hypertensive against 7.22% in normal BMI (p<0.0001; Odds ratio: 0.12; 95% CI: 0.051-0.2926) or overweight subjects (p <0.0001; Odds ratio: 0.14; 95% CI: 0.062 - 0.324). Biochemical data showed that 34.78% of obese subjects had diabetic profile compared to 22.03% in overweight population (p=0.0596; Odds ratio: 0.523; 95% CI: 0.279-0.980). In addition, correlations between BMI and blood glucose (p <0.0001; r = 0.3318) on one hand and between BMI and MAP (p<0.0001; r=0.5735) on the other hand have been observed in our study population. Although the relationship between increased BMI and the occurrence of cardiovascular and metabolic diseases has been well described by some authors [16,17]. Our study is the first to assess the links between the risk of occurrence of these related-obesity complications and the components of body composition in Sub-Saharan African subjects. Our results showed higher values of fat mass and fat index in obese or overweight populations compared to normal weight subjects. Moreover, univariate analyses revealed positive correlations between blood pressures and glycaemia with 1) BMI (p<0.0001; r=0.6112 and p<0.0001; r=0.3318 respectively) with 2) FM (p<0.0001; r=0.5523 and p<0.0001; r=0.3363) with FI (p<0.001; r=0.41 and p<0.001; r=0.32). Our results corroborate those found in other areas that have reported links between the accumulation of adipose tissue and occurrence of metabolic and cardiovascular complications related to increased fat mass [18-22]. Fat index (FI) defined by the distribution of fat in body composition compared to the muscular component is also correlated in our study with the risk of occurrence of metabolic and/or cardiovascular complications related to fat mass.
Our study has several limitations. First, our sample is cross sectional so we can infer causality between any of our measures of interest. Second, while the Pharmacy Islam is frequented by a large and diverse segment of Dakar's population, those who do not need or cannot afford over the counter or prescription medications would not be represented in our sample. In addition, we found that BMI, FM and FI are significantly related to MAP and glycaemia after adjusting for age and gender. While parameter estimates relating MAP to BMI, FM and FI are potentially clinically significant, parameter estimates that relate glycaemia to BMI, FM and FI are statistically significant however the magnitude is not indicative of clinical significance. Future studies should examine this relationship further.
Lastly, the study sample size was not based on a power calculation. The current study is a pilot study intended to generate hypotheses for future studies.
Our results suggest that BMI, FM, and FI could be the most important factors playing a role in the occurrence of metabolic and cardiovascular complications related to obesity regardless of gender or age in the Sub-Saharan African population. In conclusion, our study demonstrated that the obesity phenomenon could be a real health problem in sub-Saharan Africa and these phenomena are more common in women than men. Future researches should be necessary to confirm eventually this phenomenon and to focus on dietary and physical activity related strategies to reduce fat mass and encourage weight management.
Competing Interest
The authors declare that they have no competing interests.
Author Contributions
Conceived and designed the experiments: MD AS RA AKS VO AS RB.
Performed the experiments: MD AS RA AKS VO.
Recruited the patients: MD AS RA AKS VO.
Analyzed the data: MD AS RA AKS VO RB.
Wrote the paper: MD AS RA AKS VO RB.
Carefully revised the draft: MD AS RA AKS VO MT SHD AJFT FBS AM ABD FB AB AS RB.
Approved the submitted final version: MD AS RA AKS VO MT SHD AJFT FBS AM ABD FB AB AS RB
Acknowledgements
The authors would like to that all the participants of study, as well as the Islam pharmacy staff for the management of subjects recruitment.