Resistin Levels In Healthy Non-Diabetes Nigerian-Africans: Effect Of Obesity The ABU Adiporesistin Survey

Background Conflicting findings exist on the mechanisms of resistin in obesity and insulin resistance among white populations, as data on these in black Africans are scarce, on account of genetic/geographical differences. Hence, the study aimed to determine plasma resistin levels in Nigerian-Africans and explore its relationship with obesity and selected cardiometabolic risks. Materials and Methods A cross-sectional study on 87 randomly-selected non-diabetes Nigerians of both genders allocated into three study groups: 24 normal weight; 23 overweight and 40 obese groups by the WHO criteria. Fasting Insulin (FI), Homeostasis Model Assessment-Insulin resistance (HOMA-IR) and obesity indices were determined. Plasma resistin was measured via enzyme linked immunosorbent assay. One-way Kruskal-Wallis test determined cardiometabolic parameters across the groups with Spearman’s correlation assessing relationships. Results Resistin levels were higher in overweight and obese than normal weight subjects {6.6±3.8, 6.7±4.2 versus 5.6±2.9 µIU/mL, ( p<0.001 )}, with highest concentrations in severely obese than mildly and non-obese subjects {(7.4±5.6 versus 6.7±4.2 versus 6.05±3.0 µIU/mL ( p<0.001 , One-way Kruskal-Wallis). However, resistin showed no significant ( p>0.05 ) correlations to HOMA-IR, FI and obesity indices. The higher HOMA-IR found in overweight and obese than normal weight subjects (3.7±3.5, 2.9±2.4 vs. 2.1±0.4, p<0.001 ), was positively correlated to obesity and FI, which also increased from mild to severe BMI categories. Conclusion Higher resistin concentrations are found in severe obesity among non-diabetes black Africans, but its lack of correlation with insulin resistance and obesity indices may suggests possible interplay of other pro-inflammatory cytokines or hormones which may be evaluated in further studies.


Introduction
Resistin otherwise known as Found in the Inflammatory Zone three (FIZZ3) or adipose tissue-specific secretory factor (ADSF) belongs to a family of cysteine-rich C-terminal protein called resistin-like molecules (RELMS) which has its RETN gene encoded in chromosome 19 in humans. 1 First discovered in the white adipose tissue of mice, its expression as a 114-amino-acid peptide produced predominantly in murine adipocytes 2 is structurally and functionally different from its human expression as a 108 amino-acid polypeptide in adipocytes, placenta, 3 gut, 4 pancreatic cells, 1,5 as well as its abundant expression in mononuclear cells & macrophages. 5 Higher levels of expression of both gene and protein in central adipose tissue depots, supported a note for resistin in linking central obesity to diabetes with recent meta-analysis showing the positive correlation of resistin with insulin resistance in type 2 DM and obese subjects with hyperresistinaemia but not in those with normal circulating resistin levels. 1,[5][6][7][8] On the contrary, conflicting data exist regarding resistin and its linkage to obesity and insulin sensitivity and or resistance, as some studies have reported low resistin messenger ribonucleic acid (mRNA) expression in isolated human adipocytes and associated lack of correlation with obesity or insulin resistance, thereby making its role in insulin resistance rather unclear. 1 Furthermore, some other study found no difference in serum resistin levels between lean healthy and obese insulin resistant non-diabetes as well as type 2 diabetes adolescence. 9 More so, although resistin levels increased in the second trimester of pregnancy, there were paradoxically lower values found in gestational diabetes mellitus. 3 Resistin is also expressed in cardiomyocytes; its over-expression alters cardiac contractility and via activation of the insulin receptor substrate-1/mitogen-activated protein kinases (IRS-1/MAPK) pathway, promotes cardiac hypertrophy. 10,11 Endothelial cell function and homeostasis may also be directly affected by adipocyte-derived hormones: Reilly and coworkers showed that in both nondiabetes and diabetes subjects, plasma resistin levels were not only associated with metabolic and inflammatory markers viz.: {tumour necrosis factor alpha (TNF-alpha); interleukin (IL-6), lipoprotein associated phospholipase A} but correlated with coronary artery calcification which is a quantitative measure of atherosclerosis. 11,12 Resistin stimulates endothelin-1 (ET-1), vascular cell adhesion molecule-1 (VCAM-1) and macrophage chemoattractant protein-1 (MCP-1), contributing to endothelial cell dysregulation and consequent cardiovascular disease. 13 It has also been shown to upregulate the expression of adhesion molecules inclusive of VCAM-1, CAM-1 and long pentraxin 3 (PTX 3) in human aortic endothelial cells. 14 Obesity is a strong risk factor for later onset of type 2 diabetes and may often occur concurrently. 5 It has been documented that 60-90% of all type 2 DM patients are or have been obese, with the World Health Organization (WHO) coining the duet, a 21 st century pandemic. 5 A parallel rise in the prevalence of obesity and diabetes has attracted the term "diabesity" and studies have reported underlying mechanisms in which central obesity leads to insulin resistance and consequent diabetes. 1,15 The fat cells release pro-inflammatory adipocytokines such as leptin, visfatin, apelin, ghrelin, angiotensinogen, adiponectin, TNF-alpha, plasminogen activator inhibitor-1 and resistin which underlies the inflammation and IR linking obesity to metabolic syndrome. [11][12][16][17] Significant correlations between resistin and IL-6, intercellular adhesion molecule one (ICAM-1) have been documented in obese subjects while lipopolysaccharide (LPS), IL-1, TNF-alpha and IL-6 strongly stimulate resistin expression in human mononuclear cells. 1,[11][12][16][17] Furthermore, while several studies regarding biomarkers and mechanisms underlying obesity and IR have been elucidated in western populations; 18 studies regarding this in black African populations especially Nigerians are lacking, bearing in mind the current rise in obesity, type -2 diabetes and cardiovascular diseases. 19 Hence, there is a need for more data from different ethnic groups, bearing in mind the impact of genetic and geogaphical differences on resistin expression which can be subject to ethnic variations. 1,16,20 Hence, the study aimed to determine baseline plasma resistin levels in apparently healthy non-diabetes Nigerian-Africans and explore its relationship with obesity and selected cardiometabolic risk factors.

Study Design & Setting
The study was a cross-sectional study carried out at the Ahmadu Bello University Teaching Hospital

Inclusion Criteria
Adults greater than 20 years with no family history of hypertension or diabetes; no clinical history of cardiorespiratory, renal or liver disease as well as no neurological disease (stroke/heart attack) in the previous 36 months. 16 Subjects with stable weight over 3 months prior to the study and no history of diabetes with FBG < 7.0 mmol/L, were also included. Fasted state (last meal was 10-12 hours) prior to blood sampling. All subjects with human immunodeficiency virus (HIV), Hepatitis Bsurface antigen (HBsAg) and Hepatitis C viral antibody (HCV Ab) negativity and normal serum electrolyte, urea and creatinine (Creatinine <1.5 mg/dL) were also included.

Exclusion Criteria
Subjects were excluded on account of the following: Clinical and or laboratory evidence of any illness; chronic pro-inflammatory diseases like arthritis or rhinitis; 17 smoking >20 cigarettes per day in the past one year/ one pack year smoking; clinical evidence of endocrine diseases (hypo or hyperthyroidism); 9 concurrent use of any form of medications such as psychoactive drugs, steroids, metformin or insulin injectable for obesity therapy in the preceding 7 days; 22 engagement in competitive sports or activities and such factors which interfere with insulin secretion and action as well as pregnant or breastfeeding women. 9 Study Procedure

Sample Collection and Anthropometric Measurements
Socio-demographic information was collected by interviewer administered questionnaire by four trained medical doctors. Body weight in kilogram (Kg) was measured with light clothing and no shoes, to the nearest 0.5 kg. Height in meters (m) was taken with a stadiometer to the nearest 0.5 cm with subjects standing erect without shoes or head gear. 23 Body mass index (BMI) was derived by dividing the weight by the height (Ht 2 ) in accordance with the WHO recommendation. 23 Waist circumference (WC) was measured to the nearest 1 cm, at a level parallel to the floor, midpoint between the top of the iliac crest and the lower margin of the last palpable rib in the mid-axillary line while the subjects expired gently. 24 This was done with a non-stretch one cm wide measuring tape wrapped snugly around the subjects without any constricting effect and with the tape level and parallel to the floor at the measurement point. The subjects were standing upright with their arms relaxed at their side, feet spread apart and body weight evenly distributed during the procedure. 24 BMI was classified based on the WHO criteria viz.: 25 Normal (BMI 18.5-24.9 kg/m 2 ); overweight (BMI 25.0-29.9 kg/m 2 ) and obese (BMI ≥ 30.0 kg/m2). General obesity was further graded into mild obesity (BMI 30.0-34.9 kg/m 2 ); moderate (BMI 35.0-39.9 kg/m 2 ) and severe obesity (BMI ≥ 40.0 kg/m 2 ). 25 Blood pressure was measured twice after a 10 minutes rest in the recumbent position, standing and sitting with a random zero mercury sphygmomanometer and averaged. Deviation (Mean ± SD) and range (Minimum -Maximum). The Kolmogorov-Smirnov test was used to determine the normality of data distribution. Independent Student's t test determined the sex differences in numerical variables. Three or more Independent numerical variables with normality and equal variances were analysed via One-way Analysis of Variance (ANOVA) with Post-hoc Bonferroni test. When data was not normally distributed with equal variances observed, the Kruskal-Wallis test with Pairwise comparison was used. Two Independent Non-parametric numerical variables were analysed via the Man Whitney U test. Spearman Correlation analysis was used for correlations relationships between plasma resistin and other cardio-metabolic parameters. Differences between groups were considered significant at p ≤ 0.05 at 95% confidence interval.

Subject Participation
A total of 100 healthy volunteers were screened by simple random sampling at the MOPD of ABUTH, Zaria, of which 13 were excluded on account of the following: 2 were incidentally found to be HIV

Characteristics of the Study Population by Gender
The Mean age of the study population was 46.0 ± 10.7 years with no significant (p > 0.05) gender differences. Likewise, the mean waist circumference (WC), body mass index (BMI) and fasting blood glucose showed no significant (p > 0.05) difference between the males and females ( Table 1). The fasting insulin and plasma resistin levels appeared higher in males than females but this was statistically (p > 0.05) insignificant. Homeostasis model assessment insulin resistance (HOMA-IR) also showed no significant (p > 0.05) difference between both sexes (Table 1). HOMA-IR, p = 0.02) than normal weight subjects ( Table 2). There was no significant (p>0.05) difference in fasting insulin and HOMA-IR levels between the overweight and obese subjects. Likewise, plasma resistin levels showed significant difference across the three BMI categories by One-way Kruskall Wallis test, with highest levels found in the obese subjects, which was significantly higher than normal weight subjects but showed no difference with overweight subjects (

Resistin Levels across WHO classification of Obesity in Comparison to Non-Obese Subjects
(p>0.05) in resistin levels between the moderately obese and mildly obese subjects as well as moderately obese and non-obese subjects respectively (Figure 2).

Reisistin in Relation to Selected Cardiometabolic Parameters
The study showed no significant correlation of plasma resistin with age, central obesity by waist circumference, general obesity by BMI, FBG, fasting insulin, HOMA-IR, systolic and diastolic blood pressures in all subjects by Spearman Correlation Analysis ( Table 3). The same trend was found for males and female respectively (Data not shown).
When obese subjects were considered independently, there was no significant (r = 0.06; p = 0.73) correlation of diastolic blood pressure with resistin levels as well as with systolic blood pressure (r = 0.09; p = 0.84) (Data not shown).

Relationship of Metabolic Parameters with Obesity Indices
HOMA-IR showed no significant positive correlation with obesity indices, WC and BMI in all subjects (Table 4) HOMA-IR also showed significant (p<0.001) positive correlation with fasting insulin (Table 4).
FBG was significantly (p< 0.001) negatively correlated with fasting insulin in all subjects (Table 4) with a marginal (p = 0.06) correlation in females (r = -0.05) and no significant correlation in males (Data not shown).

Discussion
To the author's best knowledge, this is the first Nigerian study to report on resistin levels in Nondiabetes Nigerian-African adults. Preliminary animal studies showed that resistin is upregulated in rodent models of obesity and is associated with insulin resistance with its down-regulation by insulin sensitizer, rosiglitazone. 1 Later studies however, documented contrary findings attributing the lower levels of resistin in obese rodents to its suppression by free fatty acids. 28 The present study showed that in Nigerian-Africans, plasma resistin levels rose with increasing BMI categories with highest levels in subjects with severe obesity, however, with no significant positive correlation with age, insulin resistance, fasting insulin, fasting blood glucose, body mass index and central obesity. This is consistent with some previous study, which showed significantly higher resistin concentrations in overweight and obese subjects but without correlation with insulin resistance, blood glucose, lipids, insulin and changes in body mass index & visceral fat except for positive correlations with age. 29 Contrary reports in humans however exist, in which resistin showed positive correlation with insulin resistance in type 2 diabetes, non-diabetes, 30 obese as well as healthy individuals. 1,8,9 Zaidi et al. showed that Pakistani individuals with severe IR had higher resistin levels than those with normal insulin action and also demonstrated a positive correlation of resistin with IR in obese non-diabetes subjects similar to diabetes subjects. 8 Some other study showed positive relations of resistin to metabolic markers such as fat mass, high density lipoprotein cholesterol (HDL-c), triglycerides, Creactive protein and blood pressure in morbidly obese Spanish subjects without correlation to IR. 16 On a further note, HOMA-IR were higher in overweight and obese than normal weight subjects and showed significantly (p<0.001) positive correlation to fasting insulin in all subjects, males and females respectively as well as, marginal correlation to central obesity majorly in females thereby, supporting the myriads of existing evidence implicating insulin resistance as an underlying mechanism in visceral obesity. 1,[6][7][8][9]16 However, despite this, IR was not shown to correlate with resistin concentrations in this study similar to some other study. 31 The possible explanation for this lack of correlation may be the indirect effect of many other inflammatory markers released from the adipocytes which affect insulin resistance, 16 as well as, the fact that resistin may not be the primary determinant of IR; even though this study might be underpowered to make such an inference, as these inflammatory markers were not objectively assayed. 8,[11][12][13]16 Recent meta-analysis has shown mild correlation of resistin with insulin resistance (IR, r = 0.21) however, with significant heterogeneity as well as increased resistin levels correlating with inflammatory markers. 11,12 Underlying mechanisms adduced are not exactly clear in humans but may be attributed to the release of pro-inflammatory cytokines inclusive of TNF-alpha and IL-6 through the nuclear factor-kB (NF-kB, p50/p65) signalling pathway, 1 on account of the effect of resistin on endothelial cells at all levels and its high expression in mononuclear cells. 5 This has been supported by both in vitro and human studies. [1][2][3][4][5][7][8][9] Resistin has several functional receptors inclusive of Toll-like receptor four (TLR-4), an isoform of decorin, receptor tyrosine kinase like orphan receptor one (ROR-1) and adenyl cyclase associated protein one (CAP-1) in both animals and humans. 1,11,17 Its binding to these sites is the first step in its activation and subsequent cascade of events that result in low grade inflammation. 16,17 Some other reason for the lack of correlation with IR may be the lower mean resistin levels across the different BMI categories which fell within the normal range of values for the United States and European countries. 16  Consistent with a previous report, fasting insulin levels were higher in overweight and obese subjects than normal non-diabetes subjects with similar trend in resistin levels. 9 Reports have shown that resistin concentrations rise in response to supra-physiological doses of insulin in obese subjects with consequent acute regulation of resistin by insulin, 30 suggesting that subjects with higher serum fasting insulin will have higher resistin concentrations. 9 However, resistin showed no correlation with fasting insulin in this study possibly due to the influence of genetic variation 9,32 which will need to be determined in further studies.
Furthermore, age 29,30 and gender 9,32 are factors shown to influence resistin levels; however, age in this study could not be implicated as a confounder as there was no significant age difference between the three BMI categories and age showed no correlation with resistin levels. There were also no significant differences in resistin levels between the male and female subjects although it trended towards higher levels in males in all BMI categories similar to previous reports. 9,33 Another human study, on the other hand, has documented higher resistin levels in females than males, 30,[34][35] while some other in rats, reported higher levels in males than females. 1,16 Limitations Cross-sectional studies such as this may not elucidate mechanisms or determine the cause-effect

Conclusion
Higher resistin levels were found with increasing severity of obesity in healthy non-diabetes Nigerian-Africans. HOMA-IR and fasting insulin were also higher in overweight and obese non-diabetes than normal weight subjects. However, plasma resistin showed no correlation with IR and obesity indices amongst the subjects as well as gender, suggesting possible interplay of other underlying proinflammatory cytokines or hormones.
Declaration Figure 1 Study Participation of Healthy Non-Diabetes Subjects in the AdipoResistin Survey 1