Relationship between serum leptin and insulin resistance among obese Nigerian women

Abstract

Background:

Leptin is a hormone produced directly from adipocytes and has been associated with type 2 diabetes mellitus (T2DM) which is characterized by insulin resistance (IR). Due to the increasing prevalence of obesity in sub-Saharan Africa, serum leptin can be explored as a predictive risk factor for T2DM. Therefore, the aim of this study was to determine the relationship between serum leptin and IR among obese women.

Methods:

This was a cross-sectional study of obese, adult Nigerian females. Participants with body mass index (BMI) ≥30 kg/m² and nondiabetic were recruited as subjects. Fasting serum leptin, insulin, and plasma glucose were determined. IR was calculated using the formula: Homeostatic model assessment-IR (HOMA-IR) = (glucose × insulin)/22.5. Statistical analyses were performed using SPSS and P < 0.05 was considered to be significant.

Results:

Eighty obese females with mean ± standard deviation BMI 39.1 ± 7.2 kg/m² and serum leptin level 48.4 ± 24.4 ng/ml participated in study. Prevalence of hyperleptinemia was 92.5% (confidence interval: 87.3–97.7%). The relationship between leptin and HOMA-IR among the subjects was: BMI 30–34.9 kg/m²: n = 27, r = 0.18, P = 0.42; BMI 35–39.9 kg/m²: n = 24, r = 0.36, P = 0.11; BMI ≥ 40 kg/m²: n = 29, r = 0.52, P = 0.004*; and after controlling for BMI (n = 29, r = 0.46, P = 0.014*). Multiple linear regression showed that leptin did not predict for IR (P = 0.837).

Conclusion:

Serum leptin levels were positively correlated with IR, which was significant among the Class III (morbid) obesity class. However, leptin was not a predictive factor for IR in obese Nigerian women.

Introduction

Obesity, an increase in body fat mass, is one of the leading preventable causes of death worldwide.[1] It is well-documented in the developed world with increasing prevalence in developing countries. In Nigeria, adult obesity prevalence rates of 12.5%, 13.1%, and 22.2% have been reported.[2,3,4]

Obesity has been associated with insulin resistance (IR) which precedes type 2 diabetes mellitus (T2DM).[5] Although the mechanism how this occurs is not fully understood, many cellular and molecular factors interplay, including inflammatory processes.[6] Some studies have demonstrated the prevalence of IR/T2DM as a complication of obesity,[7] while others have shown an increased prevalence of obesity in T2DM patients.[8,9]

A direct product of adipose cells is leptin, a satiety hormone, which regulates food intake and energy utilization.[10,11] Leptin was identified and cloned from the obesity gene discovered in ob/ob obese mice, and has also been observed in humans.[12] It has been independently associated with T2DM, hypercholesterolemia and metabolic syndrome in both men and women in developed countries.[13] Both leptin and insulin act at the same key hypothalamic areas to decrease food intake and increase energy expenditure thereby regulating long-term energy. Insulin stimulates production of leptin when adipocytes are exposed to glucose to encourage satiety; while leptin, via a negative feedback, decreases the secretion of insulin and enhances tissue sensitivity to it, leading to glucose uptake for energy utilization or storage.[14,15,16]

Contrary to the leptin deficiency observed in obese ob/ob mice, high serum leptin levels probably reflecting leptin resistance are seen in obese humans,[17] which can predict an increased risk for T2DM.[18] This may be because deregulation of the adipocyte-insulin axis in pancreatic beta cells in a state of hyperleptinemia results in hyperinsulinemia; of which the anabolic action of insulin stimulates adipogenesis thereby leading to a further increase in insulin secretion, and consequently, IR with the development of T2DM.[19,20]

Although anthropometric measurements such as; body mass index (BMI) and waist circumference (WC), are well-known variables used to describe general and central obesity, respectively,[21] a biomarker like human serum leptin concentration, which is closely linked with BMI and fat mass, and increases steadily with adiposity[22] may be proposed to designate obesity levels. Consequently, the aim of this study was to determine the relationship between leptin and IR among obese Nigerian women.

Methods

This was a cross-sectional study of obese, adult Nigerian females attending the medical outpatient clinic of the Lagos University Teaching Hospital (LUTH), which is an 800-bed tertiary health institution located in Lagos, South-West Nigeria. This study was reviewed and approved by the Research and Ethics Committee of LUTH as being in compliance with the Declaration of Helsinki (HREC number: ADM/DCST/HREC/433). Informed consent was obtained from each participant.

Females aged 21–60 years were examined to obtain their anthropometric measurements. WC was measured with a tape rule to the nearest 0.1 cm at the end of a normal expiration at the midpoint between the subcostal plane and the iliac crest of an exposed abdomen. Hip circumference (HC) was also measured at the largest standing horizontal circumference of the buttock to determine the waist-to-hip ratio (WHR = WC/HC). Body weight in kilogram (kg), was measured in the upright position using a portable, digital weighing scale that had been validated with calibrated weighing masses; and their height in meters (m), for the calculation of BMI = weight in kg/(height in m²). Those with BMI ≥ 30 kg/m² were voluntarily recruited as subjects while those with BMI 18.5–24.9 kg/m² were recruited as normal weight controls. Adequately tested questionnaires were administered, and patients being managed for T2DM, polycystic ovarian syndrome, hypothyroidism, and Cushing’s syndrome were excluded.

The subjects were instructed to return on an agreed morning within the week after an overnight 10–12 h fast. After antiseptic preparation of the antecubital vein was effected with methylated spirit swabs, venous blood was collected into a plain specimen tube (5 ml) for serum leptin and insulin; and fluoride oxalate tube (3 ml) for fasting plasma glucose determination.

The specimens were taken to the laboratory for processing, where the samples in the plain tube were allowed to clot and retract for 30 min, and then all samples were centrifuged at 4000 rpm for 10 min at room temperature. Plasma specimens were analyzed for glucose daily, but the serum was aliquoted into two cryogenic storage tubes (one for serum leptin and insulin each) and stored at − 20°C in a well-monitored freezer (NuAire, USA), for 1-month until the analyses were performed. Hemolyzed, icteric, and lipemic samples were excluded.

Glucose oxidase method was used to estimate fasting plasma glucose concentration on a Roche Hitachi 902 analyzer (Roche diagnostics, Germany). A sandwich, solid-phase enzyme-linked immunosorbent assay (ELISA) was used to measure serum leptin (Diagnostic Automation Inc., USA); and serum insulin (Diagnostic Automation Inc., USA), both with intra- and inter-assay coefficients of variation of <10%; which were read using a BioRad microwell reader (BioRad, USA).

Insulin resistance was calculated with the homeostatic model assessment (HOMA) method which reflects the degree of IR as: HOMA-IR = (glucose × insulin)/22.5 for glucose in mmol/L.[23] Quality control was performed for plasma glucose using bovine precision multi-sera (levels 2 and 3) (RANDOX laboratories, UK); while the ELISA kit quality control material (low and high) (Diagnostic Automation Inc., USA) with reference intervals of leptin: 1.7–4.5 ng/ml, 22.1–40.9 ng/ml and insulin: 8.7–17.3 µIU/ml, 27.1–53.7 µIU/ml, respectively, were used.

Statistical analysis of the data was performed using SPSS software version 20 (IBM Corp., Armonk, New York) statistical package. The data were tested for normality using Kolmogorov–Smirnov test and data that were not normally distributed were log-transformed. Student’s t-test was used to compare the means of leptin, insulin, glucose, and HOMA-IR between the subjects and controls. Pearson’s correlation coefficient was performed to determine relationships while multiple linear regression was performed to determine prediction using HOMA-IR as the outcome, and age, BMI, WC, WHR, leptin, glucose as predictors. A P < 0.05 was considered to be statistically significant.

Results

Eighty female subjects with mean ± standard deviation (SD) age 44.9 ± 9.8 years; and BMI 39.1 ± 7.2 kg/m² were recruited for this study and grouped into Class I, Class II, and Class III obesity (morbid obesity) using BMI groups of 30.0–34.9, 35.0–35.9, ≥ 40 kg/m², respectively, while 76 age-matched, normal weight (BMI = 18.5–24.9 kg/m²) females were recruited as controls.

The mean ± SD serum leptin in the subjects was 48.5 ± 24.7 ng/ml, which was significantly higher than that of the controls 19.8 ± 17.2 ng/ml (P < 0.001); but not the fasting serum insulin, plasma glucose, and HOMA-IR (P > 0.05) [Table 1].

Table 1.

Mean (SD) values in obese subjects versus normal weight controls

	Mean (SD)		P
	Subjects (BMI ≥30 kg/m2) n=80	Controls (BMI 18.5-24.9 kg/m2) n=76
Serum leptin (ng/ml)	48.4 (24.4)	19.8 (17.2)	<0.001*
Plasma glucose (mmol/L)	4.9 (1.4)	5.4 (1.5)	0.06
Serum insulin (μIU/ml)	13.6 (10.4)	15.7 (15.5)	0.41
HOMA-IR	2.85 (2.8)	2.97 (3.5)	0.85

*means p<0.05 is statistically significant. SD=Standard deviation, HOMA-IR=Homeostasis model assessment insulin resistance, BMI=Body mass index

The prevalence of hyperleptinemia observed among the obese subjects was 74 (92.5%, confidence interval: 87.3–97.7%), using a cut-off value of >5.5 ng/ml for females.[13]

The correlation between serum leptin and insulin levels among all the obese subjects showed a positive relationship (r = 0.31, P = 0.008⋆), as well as with glucose (r = 0.25, P = 0.036⋆) and HOMA-IR (r = 0.422, P < 0.001⋆). On subdivision into obesity classes, the positive correlation with HOMA-IR was replicated in the morbid obesity group only (r = 0.52, P = 0.004⋆) even after controlling for BMI (r = 0.46, P = 0.014⋆) [Table 2]. Multiple linear regression of data obtained from all obese subjects (BMI ≥ 30 kg/m²) showed only WC (P = 0.026⋆) was a significant predictor of HOMA-IR; but when subdivided, the morbid obesity group (BMI ≥ 40 kg/m²) showed both WC (P = 0.005⋆) and WHR (P = 0.017⋆) were significant predictors of HOMA-IR. Leptin did not show significant prediction for HOMA-IR (all subjects with BMI ≥ 30 kg/m² P = 0.837; BMI ≥ 40 kg/m² P = 0.840).

Table 2.

Relationship between serum leptin, BMI, and insulin resistance

Obesity level	Subjects n	Serum leptin	BMI			HOMA-IR
		Mean (SD)	Mean (SD)	r	P	Mean (SD)	r	P
Class I	27	33.9 (2.1)	32.8 (1.3)	0.15	0.50	2.6 (1.7)	0.18	0.42
Class II	24	38.8 (1.6)	37.4 (1.3)	0.31	0.16	2.8 (1.4)	0.36	0.11
Class III (morbid obesity)	29	50.1 (1.8)	46.3 (7.2)	0.27	0.008*	3.5 (2.6)	0.52	0.004*
Class III after controlling for BMI	29	50.1 (1.8)				3.5 (2.6)	0.46	0.014*

*means p<0.05 is statistically significant. SD=Standard deviation, HOMA-IR=Homeostasis model assessment insulin resistance, BMI=Body mass index

Discussion

A high prevalence of hyperleptinemia (92.5%) was observed in this study which deliberately comprised of only females, because serum leptin concentration has been described to exhibit significant sexual dimorphism, with females having higher concentrations at every level of relative and absolute adiposity than males.[22,24] This is because hyperleptinemia is more closely associated with adipose cell hypertrophy, which is seen more in females, than adipose tissue hyperplasia which is seen more in males.[25]

As expected, serum leptin levels were higher among the subjects than controls in this study because they have increased adiposity which directly produces leptin. However, the controls still had a mean level of 19.8 ng/ml which well above the reference interval of 0.07–0.13 ng/ml for healthy, normal weight females given in a Caucasian population study;[26] as well as >5.51 ng/ml defined as hyperleptinemia among Taiwanese adults.[13] This supports the findings of a study in South Africa that observed that lean and overweight black females had higher serum leptin concentrations than their age and obesity-matched white females.[27]

The positive, linear relationship shown between leptin and BMI in our study is consistent with other studies[24,28] suggesting that obese females may be insensitive to endogenous leptin production, and therefore try to compensate by producing more; supporting the theory of leptin resistance and hyperleptinemia seen in humans which differs from animal models where leptin deficiency predominates in obesity.[29]

A study demonstrated higher leptin levels in normal weight diabetics than nondiabetic controls.[30] This corroborates the findings of Fischer et al. that subjects with T2DM have higher serum leptin levels independent of body fat mass,[31] but differs from an Iranian study which discovered that serum leptin levels were reduced in nonobese subjects with T2DM.[32] Based on these conflicting reports, known diabetics were excluded from this study.

Despite the similar HOMA-IR observed between the subjects and controls, serum leptin levels showed a linear relationship with insulin levels as well as HOMA-IR, corroborating a study which described a significant positive correlation between serum leptin levels and parameters of IR in a nondiabetic, healthy population,[33] and another which showed a strong relationship between fasting insulin level and leptin concentration.[34] Karatela and Sainani reported an association between leptin and IR/T2DM in obese patients,[35] which may be explained by the close interaction of the adipocyte-insulin axis and leptin regulatory pathways.

After obesity class subdivision, the significant positive relationship between serum leptin and HOMA-IR initially seen among all the subjects was attributed to the Class III obese (morbid obesity) group, even after correcting for BMI. This may suggest that Nigerian women may need to attain very high BMI percentiles, and consequently high leptin levels before frank IR or T2DM develops, but more research will be needed to confirm this. The findings can be welcoming in African societies where obesity is socially acceptable.[36]

Our study could not demonstrate leptin as a predictive factor for IR as similarly reported by Ezenwaka et al.[37] This may be due the limited sample size both studies used. They sampled 43 nondiabetics while we sampled 80 nondiabetics. But the popular anthropometric WC measurement is still useful in predicting IR,[30,38] which our study corroborates.

Further studies can be done to determine the obesity level at which IR commences. The cross-sectional design of this study limits our ability to determine this, and it is beyond the scope of this research work.

Conclusion

We demonstrated a positive relationship between leptin and IR, which was particularly significant at the morbid obesity level. Further molecular studies may be needed to explain this. Furthermore, we did not observe leptin as a predictor for IR. Given the high prevalence of obesity in developing countries, as well as its associated risk for T2DM, large population cohort studies on serum leptin levels can be performed to predict risk for this condition.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.