Adverse metabolic profiles in a cohort of obese Irish children

Abstract

Abstract

Background

The metabolic characteristics of obese Irish children are not well defined. We prospectively examined the relationship between the degree of obesity and glucose metabolism, insulin sensitivity and suspected non-alcoholic steatohepatosis (NASH) in a pilot study of obese Irish children.

Methods

We measured height, weight, body mass index (BMI), blood pressure, waist and hip circumference in 18 participants (mean age 15.5 years). Fasting blood glucose, insulin, lipid profile and alanine aminotransferase (ALT) concentrations were also measured. A standard 75 g oral glucose tolerance test was performed and insulin sensitivity was derived from this using a mathematical model – oral glucose insulin sensitivity.

Results

There were significant associations between the degree of obesity, insulin sensitivity and markers of liver steatosis. For example, when adjusted for pubertal status, there were significant associations between standardized BMI and insulin sensitivity (regression coefficient, β = −70.1, P = 0.018) and ALT (β = 20.7, P = 0.007)

Conclusion

This study suggests that the degree of obesity is associated with lower insulin sensitivity and possible NASH in obese Irish children.

Background

With increasing childhood obesity, many co-morbidities associated with adult obesity are now manifesting at a younger age. These include dyslipidaemia, insulin resistance, type 2 diabetes and non-alcoholic steatohepatosis (NASH). Obesity prevalence rates in Irish adolescents are among the highest in the world. ¹ However, no analysis has been performed on the metabolic characteristics of this group.

While about one-third of obese children and adolescents have an insulin resistance syndrome, ² the prevalence of obesity-induced dysglycaemia varies between ethnic groups. ³ We sought to define the metabolic characteristics of a clinic-based cohort of obese Irish children, and to elucidate the relationship between the degree of obesity, possible fatty liver disease and insulin sensitivity. As the criteria for diagnosis of ‘metabolic syndrome’ in children are not universally accepted, ⁴ rather than attempting to define its prevalence, we examined individual components of the syndrome and their relationship to metabolic risk.

Methods

Children previously referred to the endocrine clinic for assessment of obesity were invited to participate. For those under 16 years of age, parents were approached. Participants with coexisting illnesses, e.g. endocrine disorders or underlying syndromes, were excluded. The study was approved by the local ethics committee. Written informed consent was obtained (assent for those <16 years) from each participant.

Participants attended the testing site in the morning, after an overnight fast. A full history and physical examination were performed, including Tanner staging to measure pubertal development. Weight was measured using a Seca^® scale (Medical Scales and Measuring Sytems, Birmingham, UK) and height with a Harpenden stadiometer (Holtain Ltd, Pembs, UK). Statistical software available on the Center for Disease Control website (http://www.cdc.gov/nccdphp/dnpa/growthcharts/sas.htm) was used to calculate standard deviation scores (SDS) for body mass index (BMI), based on age and gender. Waist and hip circumference were measured using a fibreglass D-loop tape. Blood pressure was measured using an oscillometric device (Omron^® 705CP, Omron Healthcare, Inc, IL, USA). Three readings were taken from the right arm and the lowest one recorded.

Fasting bloods were drawn for total, HDL- and LDL- cholesterol, triglycerides and alanine aminotransferase (ALT). Samples were also drawn for full blood count, renal, liver and bone profiles and thyroid function tests (OGTT), in order to exclude confounding conditions such as anaemia and hypothyroidism. An oral glucose tolerance test (OGTT) was performed (1.75 g/kg to a maximum of 75 g) with glucose, insulin and C-peptide concentrations measured at 30min intervals over 2 h. Insulin sensitivity was measured using a mathematical model of oral glucose insulin sensitivity (OGIS). ⁵ This model expresses the relationship between insulin secretion and changes in glucose concentration after oral glucose loading.

Laboratory analysis

Serum insulin and C-peptide were measured using fluoroimmunoassays (Autodelfia^®, Perkin Elmer, MA, USA). Plasma total cholesterol and triglycerides were measured using enzymatic methods (Human liquicolor kits/Hitachi Modular, Stanbio Laboratory, TX, USA). Plasma HDL- and LDL-cholesterol were measured directly with enzymatic methods (Randox direct kits/Hitachi Modular, Randox Laboratories Ltd, Crumlin, UK). Plasma glucose was measured using a glucose oxidase method (bioMerieux kit/Hitachi Modular, bioMerieux, Marcy l'Etoile, France). Serum ALT was measured using an enzymatic method on a Roche® Automated Hitachi Modular-P system (Roche, CA, USA).

Statistical analysis

All values are expressed as means ± standard deviation. Associations between variables were calculated using multiple linear regression analysis, adjusted for pubertal stage and where appropriate further adjusted for gender, age and height. 95% confidence intervals for each regression coefficient are also reported. Results were deemed to be statistically significant when P < 0.05. JMP^® statistical software, version 5.1 (SAS Institute, Cary, NC, USA) was used in all statistical calculations.

Results

Eighteen participants (6 boys, 12 girls) were recruited and their characteristics are summarized in Table 1. All of the participants were well at the time of assessment, with normal full blood count, renal, liver and bone profiles and thyroid function tests. Each participant had normal glucose tolerance confirmed during the OGTT. All but the youngest child had started puberty, while five participants had completed pubertal development. Of note, one 18-year-old male participant had marked hyperinsulinaemia (fasting insulin 1050 pmol/L, 2 h insulin 726 pmol/L), but had normal glucose tolerance. These two outlying values were excluded from statistical analyses.

Table 1

Participant characteristics and associations with insulin sensitivity (OGIS) and serum alanine aminotransferase (ALT)

	Mean	SD	Range	Association with OGIS†	Confidence interval	P value	Association with ALT†	Confidence interval	P value
Age (years)	15.5	2.8	8.6–18.9	−8.3	(−26.3, 9.7)	0.34	2.1	(−3, 7.1)	0.39
Waist (cm)	110.8	15.2	91–134	‡ −1.9	(−3.9, 0.1)	0.057	‡0.6	(0.1, 1)	0.025
Hip (cm)	117.4	13.4	96–148	‡ − 2	(−3.9, −0.03)	0.047	‡0.4	(−0.04, 0.9)	0.07
WHR	0.94	0.08	0.78–1.11	‡ − 39.3	(−585.5, 506.9)	0.88	‡25.2	(−90.4, 140.7)	0.64
Weight (kg)	105.1	22.7	61.2–141.8	‡ − 1.7	(−2.9, −0.5)	0.011	‡0.4	(0.2, 0.7)	0.003
Height (m)	1.69	0.09	1.47–1.86	‡ −168.8	(−616, 278.3)	0.43	‡111.3	(46.2, 176.4)	0.003
BMI (kg m−2)	36.5	6	25.7–49.1	‡ −5.1	(−9.3, −0.8)	0.024	‡0.8	(−0.4, 1.9)	0.16
BMI SDS	2.27	0.42	1.32–2.77	−70.1	(−126.3, −14)	0.018	20.7	(6.9, 34.6)	0.007
Systolic BP (mmHg)	110.7	17.2	76–141	¶ −0.4	(−2.3, 1.4)	0.64	¶0.02	(−0.3, 0.3)	0.86
Diastolic BP (mmHg)	66.4	10.4	41–82	¶ −1.5	(−4.4, 1.4)	0.29	¶0.3	(−0.2, 0.8)	0.2
Fasting glucose (mmol/L)	4.74	0.48	4.1–6.0	−60.5	(−113.7, −7.2)	0.029	13.1	(−1.8, 28)	0.08
2 h glucose (mmol/L)	5.5	0.95	3.9–7.4	−41.1	(−60.6, −21.6)	0.0005	5.5	(−2.2, 13.2)	0.15
Fasting insulin (pmol/L)	102	44	45–198	−0.4	(−1.1, 0.4)	0.33	0.2	(0.1, 0.4)	0.002
2 h insulin (pmol/L)	308	246	52–960	−0.1	(−0.2, 0.02)	0.09	0.02	(−0.02, 0.05)	0.28
OGIS (mL min−1 m−2)	428	48	332–515				−0.1	(−0.3, 0.02)	0.084
ALT (IU/L)	23.2	12.6	8–51	−1.8	(−3.9, 0.3)	0.084
Total cholesterol (mmol/L)	3.95	0.77	2.43–5.2	−35.5	(−66.6, −4.4)	0.028	−1.8	(−12.8, 9.1)	0.73
HDL-cholesterol (mmol/L)	1.06	0.22	0.78–1.45	−33.8	(−174.1, 106.4)	0.61	−22	(−60, 15.9)	0.23
LDL-cholesterol (mmol/L)	2.32	0.64	1.33–3.42	−38.2	(−80, 3.6)	0.07	−4.5	(−18.3, 9.4)	0.5
Triglycerides (mmol/L)	1.26	0.78	0.29–3.11	−19.6	(−56, 16.9)	0.27	4.2	(−5.5, 13.9)	0.37
Total: HDL-cholesterol ratio	3.82	0.82	2.11–5.05	−34.1	(−70.2, 2)	0.062	4.2	(−6.5, 14.9)	0.41
LDL: HDL-cholesterol ratio	2.24	0.59	0.94–3.06	−50.7	(−102.8, 1.5)	0.056	1.7	(−13.9, 17.3)	0.81

^†Adjusted for pubertal status; ^‡Adjusted for pubertal status and gender; ^¶Adjusted for pubertal status, gender, age and height

WHR, waist-hip ration; BMI, body mass index; SDS, standard deviation score; BP, blood pressure

Associations between various participant characteristics, insulin sensitivity and ALT are summarized in Table 1. Insulin sensitivity was associated with crude and standardized BMI, body weight and hip circumference, with a tendency towards significance for waist circumference. For example, an increase in BMI SDS of 1 would account for a reduction in OGIS of 70.1 mL m⁻² min⁻¹. There were no associations between insulin sensitivity and waist:hip ratio, height or blood pressure. There were also significant associations with insulin sensitivity, fasting and 2 h glucose concentrations and total cholesterol concentrations. Associations between insulin sensitivity, total:HDL and LDL:HDL cholesterol ratios approached significance.

There were significant associations between serum ALT concentrations and standardized BMI, waist circumference, body weight, height and fasting insulin levels. For example, a 1 pmol/L rise in fasting insulin accounted for a 0.22 IU/L rise in ALT.

Discussion

Though our study involved a small number of subjects, we have demonstrated significant associations between the degree of obesity and insulin resistance in obese Irish children. ALT correlates strongly with the degree of hepatic steatosis in children. ⁶ Its association with obesity in our study suggests hepatic fatty infiltration proportional to body fatness, even at an early age. While most of the participants we studied had ‘normal’ ALT concentrations, reference ranges for ALT are not well defined in children. It was recently shown that deterioration in glucose and lipid metabolism was associated with even modest ALT elevations, in a large multiethnic cohort of American adolescents. ⁷ Clearly our findings in relation to obesity and ALT concentrations are preliminary; further studies are required to examine the relationship between body fatness and liver steatosis, using more precise methods such as liver ultrasound and magnetic resonance spectroscopy.

The onset of puberty causes a significant reduction in insulin sensitivity, but its relationship to NASH has not been fully elucidated. This is a factor that needs to be examined when our cohort is significantly larger. Also, this clinic-based cohort might not be representative of the overall population of obese Irish children. Larger studies would allow an assessment of a more representative population. However, participants in this study were otherwise well and had no underlying medical problems.

Conclusion

These preliminary data suggest the degree of obesity in Irish children is associated with insulin resistance and possible liver steatosis. The effects of important confounders such as family history need to be assessed in larger studies. Childhood obesity is not a benign entity. Nor is it simply a predictor of adult obesity and co-morbidity. Rather, some of the pathological processes, which are known to lead to diabetes and cardiovascular disease later in life are already established in obese Irish children.