J Endocrinol Invest (2014) 37:51–56 DOI 10.1007/s40618-013-0014-0
ORIGINAL ARTICLE
Correlation between cortisol and components of the metabolic syndrome in obese children and adolescents C. Guzzetti • S. Pilia • A. Ibba • S. Loche
Received: 26 July 2013 / Accepted: 17 November 2013 / Published online: 8 January 2014 Italian Society of Endocrinology (SIE) 2013
Abstract Background In obese subjects it has been shown that cortisol (F) contributes to the reduction in insulin sensitivity, suggesting a role in the development of the metabolic syndrome (MS). Aim The aim of this retrospective study was to evaluate the relationship between F and components of MS in 1,027 obese children and adolescents. Subjects and methods Waist circumference, systolic and diastolic blood pressure (SP, DP), F, serum glucose (Glyc), cholesterol HDL, triglycerides and homeostatic model assessment (HOMA index) were evaluated in all subjects. MS was defined according to the International Diabetes Federation criteria. Accordingly, patients were subdivided into three age groups: 6–10, 10–16 and[16 years. Results In univariate regression analysis, F was correlated with Glyc, SP and HOMA in groups 1 and 2, with DP in Group 2. In multivariate regression analysis including age, sex, puberty, BMI-SDS and F as independent variables and one of the component of the MS as the dependent variable, F was a weak predictor of the variability when DP and Glyc were introduced as dependent variables in Group 2 and when SP was introduced as dependent variable both in groups 1 and 2. When patients were subdivided into subgroups according to the IDF criteria, in Group 2 patients with one or more components of the MS had higher F concentrations. Conclusions In this cohort of obese children and adolescents, F was weakly associated with components of the MS. These findings do not support a major role for F in the development of MS. C. Guzzetti S. Pilia A. Ibba S. Loche (&) Pediatric Endocrinology Unit, Microcitemico Hospital, ASL Cagliari, Via Jenner, 09121 Cagliari, Italy e-mail: [email protected]
Keywords Cortisol Metabolic syndrome Obesity Children Adolescents Abbreviations BMI Body mass index DP Diastolic blood pressure F Cortisol Glyc Serum glucose HOMA Homeostatic Model Assessment IDF International Diabetes Federation MS Metabolic syndrome SP Systolic blood pressure TG Triglycerides WC Waist circumference WCp Waist circumference percentile
Introduction In recent decades the prevalence of childhood obesity has increased worldwide. Sedentary lifestyle and high calorie diets contributed to the increase in the prevalence of obesity and related complications. In 2012 the World Health Organization estimated that, globally, 200 million school-aged children are either overweight, and 40–50 million are classified as obese. Particularly, in the European region 22.1 % of boys and 20.3 % of girls are overweight, 5.3 % of boys and 4.4 % of girls are obese [1]. In Italy, in 2010, 26.2 % of children and adolescent between 6 and 17 years were overweight or obese [2]. Obesity is associated with dyslipidemia (elevated levels of triglycerides (TG) and low levels of HDL), hyperglycemia, insulin resistance, type 2 diabetes and hypertension.
123
52
J Endocrinol Invest (2014) 37:51–56
All these factors are components of the metabolic syndrome (MS). Metabolic syndrome has many similarities with Cushing’s syndrome. It has been hypothesized that subtle increases in cortisol levels could have an important role in the development of insulin resistance and MS. In this regard, many studies confirmed a correlation between MS and cortisol in adults, demonstrating that increased cortisol levels and obesity, especially visceral fat, cooperate in determining glucose intolerance, raised blood pressure and dyslipidemia [3–8]. The identification of children at increased risk for developing MS, and thus predisposed to type 2 diabetes and cardiovascular disease in later life [9, 10], is of great clinical importance. Few studies demonstrated a relationship between cortisol and insulin resistance or MS in children and adolescents [11–15]. The aim of this study was to evaluate the relationship between morning serum cortisol concentrations and component of MS in a large number of obese children and adolescents. We then investigated the predictive value of cortisol concentrations and whether they would be of help in identifying children at risk of developing MS.
Patients and methods
In all children and adolescents height, weight, waist circumference (WC), systolic and diastolic blood pressure (SP and DP, respectively), morning serum concentrations of cortisol, glycemia, insulin, cholesterol HDL, TG were evaluated at entry, between 08.00 and 09.00 h, after an overnight fast, in supine position. Waist circumference was measured with an anthropometric tape midway between the lower rip margin and the iliac crest at the end of a gentle expiration. Homeostatic model assessment (HOMA index, Glucose (mmol/L) 9 Insulin (mU/L)/22.5) was used to evaluate insulin resistance. Body mass index (BMI) was calculated as weight divided by height squared. BMI-SDS was derived from the Italian reference data [16]. WC percentile (WCp) was defined for all patients according to Cook et al. [17]. MS was defined according to the criteria provided by the International Diabetes Federation (IDF) [10]. Accordingly, patients were subdivided into three age groups: in Group 1, patients aged 6–10 years; in Group 2, patients aged 10–16 years; and in Group 3, patients older than 16 years. In Group 1 MS cannot be diagnosed, but children with WCp C 90 and family history of MS, diabetes or cardiovascular disease require further evaluation. In Groups 2 and 3, MS is defined by the presence of central obesity plus any two of the other four factors altered (TG, HDL, blood pressure and glycemia) (Table 2).
Patients Assays Data were collected from 1,027 obese children and adolescents referred to our Endocrine Unit between 2002 and 2010. All children included in the study were seen in our Unit for obesity and were otherwise healthy. Their main clinical characteristics are summarized in Table 1.
Serum glucose was determined by the glucose oxidases method (Autoanalyser, Beckman Coulter, USA). Serum insulin and cortisol concentrations were measured using reagents provided by Siemens (Lianberis, UK) and an
Table 1 Clinical characteristics of the patients studied All patients (n = 1,027)
Group 1 (n = 408)
Group 2 (n = 390)
Group 3 (n = 89)
Sex (M/F)
475/552
175/233
210/180
24/65
Age (years)
10.51 ± 3.29
7.79 ± 1.03
12.62 ± 1.34
16.74 ± 0.89
Puberty (Prep/pub)
593/429
367/39
117/272
1/86
BMI-SDS
2.62 ± 0.51
2.65 ± 0.42
2.52 ± 0.42
2.76 ± 0.64
WC (cm) WCp
81.96 ± 11.24 92.37 ± 8.47
75.45 ± 6.89 94.07 ± 5.20
89.14 ± 9.38 90.77 ± 9.96
93.83 ± 13.72 85.75 ± 14.96 118 ± 17
SP (mmHg)
106 ± 15
98 ± 12
112 ± 14
DP (mmHg)
62 ± 9
58 ± 8
65 ± 8
69 ± 10
Glycemia (mmol/L)
5 ± 0.4
4.9 ± 0.4
5 ± 0.4
5 ± 0.4
HOMA
3.67 ± 2.46
2.77 ± 1.77
4.42 ± 2.7
4.69 ± 3
Insulin (mU/L)
16.98 ± 12.19
13.31 ± 10.63
19.82 ± 12.12
20.56 ± 12.59
HDL (mmol/L)
1.3 ± 0.3
1.4 ± 0.3
1.3 ± 0.3
1.2 ± 0.3
TG (mmol/L)
0.7 ± 0.4
0.7 ± 0.4
0.8 ± 0.4
0.7 ± 0.4
Cortisol (nmol/L)
290 ± 159
270 ± 136
306 ± 172
324 ± 157
123
J Endocrinol Invest (2014) 37:51–56
53
Table 2 Definition of MS according to the International Diabetes Federation [10] Age group
Obesity (WC)
6–9 years
C90th percentile
In this age group MS cannot be diagnosed.
10–15 years
C90th percentile
C150 mg/dl
C16 years
C94 cm in M
C150 mg/dl
C80 cm in F
Triglycerides
HDL
Blood pressure
Glucose
\40 mg/dl
SP C 130 or DP C 85 mmHg
C100 mg/dl
\40 mg/dl in M
SP C 130 or DP C 85 mmHg
C100 mg/dl
\50 mg/dl in F
Immulite system (IMMULITE 2000). TG and HDL were measured by standard routine methods.
Table 3 Univariate correlations between cortisol and component of MS Partial correlationsa
Correlations
Statistical analysis Distribution of the data was evaluated using Kolmogorov– Smirnov test. Glycemia in Group 2 and cortisol, glycemia, HDL and WC in Group 3 were the only normally distributed variables. Univariate analyses were performed using Pearson correlation coefficient for normally distributed variables, and using Spearman rank order correlation for not normally distributed variables. All correlations were also adjusted for age, gender, puberty and BMI-SDS. Comparisons between two groups were performed using the Mann–Whitney U test. Comparisons between multiple groups were performed using ANOVA and ANCOVA, with age, gender, puberty and BMI-SDS as covariates, followed by the Bonferroni’s multiple comparison test. Multivariate regression analysis was performed to determine independent predictors of the MS components. The covariates entered into the model as potential predictors included sex, age, BMI-SDS, pubertal status and cortisol. All components of the MS were used as dependent variables, one at a time. All values are reported as mean ± SD (continuous variables) or as counts and percentages (categorical variables). P \ 0.05 (two sided) was considered significant. All statistical calculations were performed using STATISTICA version 9.1 software (StatSoft Inc., Tulsa, OK, USA).
Results
r
P
r
P
WCp
0.01
0.74
0.02
0.47
SP
0.17
1027
0.15
1024
DP Glycemia
0.13 0.16
8 3 1025 1027
0.09 0.10
0.007 0.001
HDL
0.007
0.81
0.02
0.46
TG
0.02
0.45
-0.03
0.35
HOMA
0.15
2 3 1026
0.13
1024
Significant values in bold letters a
Adjusted for age, gender, pubertal status and BMI-SDS
patients (3.33 %) had MS. In Group 3 (n = 89), 6 patients (6.74 %) had MS. Univariate regression analysis The results of all correlation analyses are reported in Table 3. Univariate regression analysis in the entire cohort showed that cortisol was weakly associated with SP, DP, glycemia and HOMA, also when adjusted for age, gender, puberty and BMI-SDS. In Group 1, cortisol was correlated with glycemia, and with SP and HOMA when adjusted for age, gender, puberty and BMI-SDS. In Group 2, cortisol was correlated with SP, DP, glycemia and HOMA. The correlation remained significant also when adjusted for age, gender, puberty and BMI-SDS. In Group 3, cortisol was not correlated with components of the MS.
Patients’ characteristics Multivariate regression analysis Clinical characteristics of the patients (552 girls and 472 boys) are summarized in Table 1. According to the IDF criteria [10], patients were subdivided into three age groups. In Group 1 (n = 408), 280 patients had a WCp [ 90th, and thus 68.63 % were at risk to developing MS (according to the IDF criteria, WCp in this group is the only factor which suggests MS). In Group 2 (n = 390), 13
In Group 1, results of multivariate regression analysis including age, sex, puberty, BMI-SDS and cortisol as independent variables and one of the component of the MS as the dependent variable, indicated that cortisol was a weak predictor of the variability only in the model with SP as dependent variable (P = 0.003) (Table 4).
123
54
J Endocrinol Invest (2014) 37:51–56
Table 4 Adjusted R2 in models of multivariate analysis with sex, age, puberty, BMI-SDS and cortisol among independent variables Depend variable
Group 1
Group 2
Group 3
WCp (WC in group 3)
0.220
0.139
0.740
SP
0.079
0.276
0.249
DP
0.049
0.134
0.120
-0.006
0.049
0.073
HDL
0.004
0.006
0.021
TG
0.024
0.017
-0.010
Glycemia
Table 5 ANOVA and ANCOVA for groups 2 and 3. Subgroups analyzed: 2A: WCp \ 90; 2B: WCp [ 90 and the other components in the normal range; 2C: WCp [ 90 and 1 of the other components of the MS abnormal; 2D: WCp [ 90 and 2 of the other components of the MS abnormal; 3A: WC \ 94 cm for M and WC \ 80 cm for F; 3B: WC C 94 cm for M and WC C 80 cm for F and the other components in the normal range; 3C: WC C 94 cm for M and WC C 80 cm for F and 1 of the other components of the MS abnormal; 3D: WC C 94 cm for M and WC C 80 cm for F and 2 of the other components of the MS abnormal Group 2 F
Group 3 P
F
P
ANOVA
7.12
0.0001
0.59
0.63
ANCOVAa
4.53
0.004
0.67
0.58
a
Covariates: age, gender, puberty and BMI-SDS
In Group 2, cortisol was a predictor of the variability in the models with SP, DP and glycemia as dependent variables (P = 4 9 10-4; P = 0.002; P = 7 9 10-4, respectively) (Table 4). In Group 3, cortisol was never a significant predictor of the variability (Table 4). In order to further investigate the effect of cortisol on the variability of the component of MS, multivariate regression analysis was repeated excluding cortisol among the independent variables. The results obtained confirmed that the effect of cortisol as predictor of MS was very weak, if any (data not shown). ANOVA and ANCOVA Patients of Group 1 were subdivided into two subgroups according to the IDF criteria [10]: subgroup 1A with WCp \ 90; subgroup 1B with WCp [ 90. Mean ± SD cortisol concentration was not different in the two subgroups (P = 0.87). Patients of Groups 2 and 3 were subdivided into four subgroups according to the IDF criteria [10] and ANOVA and ANCOVA between the four subgroups were performed (Table 5). In Group 2, mean ± SD cortisol concentration
123
Fig. 1 Mean ± SD serum cortisol levels in the 4 subgroups of Group 2. 2A patients with WCp \ 90; 2B patients with WCp [ 90 and the other components in the normal range; 2C patients with WCp [ 90 and 1 of the other components of the MS abnormal; 2D patients with WCp [ 90 and 2 of the other components of the MS abnormal
was significantly different among the four subgroups. Patients with one or more components of the MS had higher F concentration (Fig. 1). In Group 3, mean ± SD cortisol concentration was not different between the four subgroups (Table 5).
Discussion MS in adults is defined as a group of cardiovascular and diabetes risk factors which include abdominal obesity, dyslipidemia, glucose intolerance, and hypertension [18]. At the moment, there is no consensus on the definition of pediatric MS. This is mostly due to physiological developmental changes of the metabolic traits used to define MS during childhood and puberty [19]. Among the many definitions of MS in children, each with advantages and disadvantages, the IDF definition of MS [10], based on accessible diagnostic tools, best suits the pediatric and adolescent population. The IDF definition, because of the age-related developmental changes, considers abdominal obesity (WC), lipid profile (TG and HDL), blood pressure and serum glucose levels in different age groups: 6 to \10, 10 to \16 and C16 years. To our knowledge, data about MS in children and adolescents are limited, only few studies evaluated the role of cortisol in MS in childhood [11–15], and none based on IDF criteria. According to the IDF definition, 68.6 % of our patients aged 6–10 years had high risk to developing MS, 3.3 % of our patient aged 10–16 years had MS and 6.7 % of our patients aged [16 years had MS. In our large cohort we found no correlation between cortisol and WC, in accordance with previous studies showing no correlation or a weak correlation between
J Endocrinol Invest (2014) 37:51–56
cortisol and indices of visceral fat accumulation [11–13, 20]. Conversely, other studies have shown a significant correlation between serum cortisol levels and WC in adult women [6, 21], in men [22] or in both sexes [23]. Glucocorticoids are involved in differentiation, proliferation and redistribution of human adipocytes from peripheral to central depots. Glucocorticoids also increase the size and number of adipocytes. Overall, there is an increase of the HPA axis responsiveness to different stimuli in patients with abdominal obesity [7, 24]. Furthermore the negative feedback on HPA axis is attenuated in obesity, with reduced diurnal variation in cortisol levels. Adam et al. [11] hypothesized that the relationship between cortisol and visceral fat accumulation could not be evident in children and adolescent because it is a consequence of prolonged glucocorticoid exposure. We found a significant correlation between cortisol and both SP and DP in all patients considered together and in Group 2. Cortisol was correlated with SP also in Group 1. Results of multivariate correlation analysis showed that cortisol was a significant predictor of the variability of SP in Group 1 and 2, and of DP in Group 2. However, albeit significant, the effect of cortisol on the variability of blood pressure was as little as 2 %. A relationship between cortisol and blood pressure has been observed also in other studies on adult subjects [3–5, 8, 12, 20, 21]. In one study cortisol was found correlated with SP also in children and adolescents [11]. Among the possible mechanisms to explain the correlation between cortisol and blood pressure are the activation of the HPA axis and sympathetic nervous system caused by stress [7, 12]. An increased responsiveness to vasoconstrictors with a decreased vasodilator production, like nitric oxide [7], an increase in cardiac output or peripheral resistance and/or mineralocorticoid-induced salt and water retention could also play a role [5]. We also found a significant correlation between cortisol and fasting morning glycemia in all patients considered together, as well as separately in Group 1 and in Group 2. In multivariate analysis cortisol was a significant weak predictor (3 %) of the variability of glycemia only in Group 2. These results are consistent with other studies showing a significant correlation between cortisol and glycemia in children and adolescents [11, 12], as well as in adult subjects [3, 4, 8, 20, 22, 25]. We found a significant correlation between cortisol and HOMA index, as indicator of insulin resistance, in all patients considered together, as well as in Group 1 and 2. Many studies have reported a significant correlation between serum cortisol and fasting plasma insulin, both in adult men and women [3, 15, 21]. Adam et al. [11] showed a correlation between cortisol and Disposition Index (an indicator of b-cell function) and acute insulin release to a glucose challenge in children and adolescents, suggesting
55
that cortisol may be associated with the decrease in insulin sensitivity. Also there is a significant correlation between cortisol and insulin resistance in both obese adults and children [4, 5, 7]. No correlation was found between serum cortisol levels and lipid serum profile, confirming the results of Weigensberg et al. [12]. Conversely, many other studies performed in adult obese subjects showed a significant correlation between cortisol and TG [4, 6, 21], between cortisol and total cholesterol, HDL, LDL [2, 8, 20] or both [5, 7]. In Group 2, we found that patients with one or more components of the MS had higher serum morning cortisol concentrations. Nevertheless, we found no relationship between serum cortisol levels and the component of MS in Group 3. Taken together, our results do not confirm an important effect of cortisol in the development of MS. Some authors [5, 8, 15, 21] have suggested that in obesity, although cortisol excretion is increased, also owing to an increased production of ACTH, circulating cortisol may be normal or low because of an increased peripheral clearance. They hypothesized that peripheral clearance of cortisol is enhanced in obesity and that central control of the HPA axis is altered, with reduced diurnal variation in cortisol levels (lower in the morning but higher in the evening). This might explain the fact that we found only weak correlations between morning serum cortisol and the component of MS. Other authors [12] hypothesized that this relationship would be mediated by insulin sensitivity since it does exist only with parameters of MS which are influenced by insulin sensitivity. This study presents some limitations. This is a retrospective study performed in a very large number of obese children and adolescent, and we used all the available information. Therefore, we have correlated morning serum cortisol values, which could not reflect 24 h cortisol production, although it has been previously shown that even a single morning measurement of serum cortisol reflects HPA activity [11]. Furthermore, cortisol serum levels are related to many factors, some of which could be controlled for their impact on the analysis (morning samples, adjusting analysis by sex, age, puberty, …), while others, like cortisol response to physiological or psychological stressors, could not be controlled and may have caused bias in data analysis. Moreover, we did not perform the same study in a control group, and we do not know whether cortisol correlates with components of the MS also in normal weight children. However, the large number of patients and the homogeneity of our cohort provide strength to the results. In conclusion, albeit we have shown that patients with MS have slightly higher cortisol concentrations, we found
123
56
only weak or no correlations between cortisol and component of MS, suggesting that cortisol does not play an important role in the development of MS in obese children and adolescents. Acknowledgments Supported in part by a Grant from Regione Autonoma della Sardegna to SL. We are gratefully indebted with our nurse staff (Donatella Arghittu and Patrizia Sanna), and our Lab technicians (Maria Grazia Contini, Danilo Mosino, Teresa Trogu) for their continuous and skilful help in patients’ care and sample processing. Conflict of interest The authors C. Guzzetti, S. Pilia, A. Ibba, and S. Loche declare that they have no conflict of interest.
References 1. http://www.iaso.org/iotf/obesity/obesitytheglobalepidemic/. Accessed Aug 2012 2. http://www.istat.it/it/archivio/43508. Accessed 25 Oct 2010 3. Rosmond R, Dallman MF, Bjo¨rntorp P (1998) Stress-related cortisol secretion in men: relationships with abdominal obesity and endocrine, metabolic and hemodynamic abnormalities. J Clin Endocrinol Metab 83:1853–1859 4. Ward AM, Fall CH, Stein CE et al (2003) Cortisol and the metabolic syndrome in South Asians. Clin Endocrinol 58: 500–505 5. Whitworth JA, Williamson PM, Mangos G, Kelly JJ (2005) Cardiovascular consequences of cortisol excess. Vasc Health Risk Manag 1:291–299 6. Esteghamati A, Morteza A, Khalilzadeh O, Noshad S, Novin L, Nakhjavani M (2011) Association of serum cortisol levels with parameters of metabolic syndrome in men and women. Clin Invest Med 34:E131–E137 7. Anagnostis P, Athyros VG, Tziomalos K, Karagiannis A, Mikhailidis DP (2009) Clinical review: the pathogenetic role of cortisol in the metabolic syndrome: a hypothesis. J Clin Endocrinol Metab 94:2692–2701 8. Reynolds RM, Syddall HE, Walker BR, Wood PJ, Phillips DI (2003) Predicting cardiovascular risk factors from plasma cortisol measured during oral glucose tolerance tests. Metabolism 52: 524–527 9. Capizzi M, Leto G, Petrone A et al (2011) Wrist circumference is a clinical marker of insulin resistance in overweight and obese children and adolescents. Circulation 123:1757–1762 10. Zimmet P, Alberti KG, Kaufman F et al (2007) The metabolic syndrome in children and adolescents—an IDF consensus report. Pediat Diabetes 8:299–306
123
J Endocrinol Invest (2014) 37:51–56 11. Adam TC, Hasson RE, Ventura EE et al (2010) Cortisol is negatively associated with insulin sensitivity in overweight latino youth. J Clin Endocrinol Metab 95:4729–4735 12. Weigensberg MJ, Toledo-Corral CM, Goran MI (2008) Association between the metabolic syndrome and serum cortisol in overweight latino youth. J Clin Endocrinol Metab 93:1372–1378 13. Misra M, Bredella MA, Tsai P, Mendes N, Miller KK, Klibanski A (2008) Lower growth hormone and higher cortisol are associated with greater visceral adiposity, intramyocellular lipids, and insulin resistance in overweight girls. Am J Physiol Endocrinol Metab 295:E385–E392 14. Ten S, Maclaren N (2004) Insulin resistance syndrome in children. J Clin Endocrinol Metab 89:2526–2539 15. Reinehr T, Andler W (2004) Cortisol and its relation to insulin resistance before and after weight loss in obese children. Horm Res 62:107–112 16. Cacciari E, Milani S, Balsamo A, Spada E, Bona G, Cavallo L (2006) Italian cross-sectional growth charts for height, weight and BMI (2 to 20 yr). J Endocrinol Invest 29:581–593 17. Cook S, Auinger P, Huang TT (2009) Growth curves for cardiometabolic risk factors in children and adolescents. J Pediatr 155:S6.e15–26 18. Alberti KG, Zimmet P, Shaw J (2006) Metabolic syndrome—a new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med 23:469–480 19. Goodman E, Daniels SR, Meigs JB, Dolan LM (2007) Instability in the diagnosis of metabolic syndrome in adolescents. Circulation 115:2316–2322 20. Park SB, Blumenthal JA, Lee SY, Georgiades A (2011) Association of cortisol and the metabolic syndrome in korean men and women. J Korean Med Sci 26:914–918 21. Walker BR, Soderberg S, Lindahl B, Olsson T (2000) Independent effects of obesity and cortisol in predicting cardiovascular risk factors in men and women. J Intern Med 247:198–204 22. Wallerius S, Rosmond R, Ljung T, Holm G, Bjo¨rntorp P (2003) Rise in morning saliva cortisol is associated with abdominal obesity in men: a preliminary report. J Endocrinol Invest 26:616–619 23. Fraser R, Ingram MC, Anderson NH, Morrison C, Davies E, Connell JM (1999) Cortisol effects on body mass, blood pressure, and cholesterol in the general population. Hypertension 33:1364–1368 24. Bjo¨rntorp P, Rosmond R (2000) Obesity and cortisol. Nutrition 16:924–936 25. Phillips DI, Barker DJ, Fall CH (1998) Elevated plasma cortisol concentrations: a link between low birth weight and the insulin resistance syndrome? J Clin Endocrinol Metab 83:757–760
ORIGINAL ARTICLE
Correlation between cortisol and components of the metabolic syndrome in obese children and adolescents C. Guzzetti • S. Pilia • A. Ibba • S. Loche
Received: 26 July 2013 / Accepted: 17 November 2013 / Published online: 8 January 2014 Italian Society of Endocrinology (SIE) 2013
Abstract Background In obese subjects it has been shown that cortisol (F) contributes to the reduction in insulin sensitivity, suggesting a role in the development of the metabolic syndrome (MS). Aim The aim of this retrospective study was to evaluate the relationship between F and components of MS in 1,027 obese children and adolescents. Subjects and methods Waist circumference, systolic and diastolic blood pressure (SP, DP), F, serum glucose (Glyc), cholesterol HDL, triglycerides and homeostatic model assessment (HOMA index) were evaluated in all subjects. MS was defined according to the International Diabetes Federation criteria. Accordingly, patients were subdivided into three age groups: 6–10, 10–16 and[16 years. Results In univariate regression analysis, F was correlated with Glyc, SP and HOMA in groups 1 and 2, with DP in Group 2. In multivariate regression analysis including age, sex, puberty, BMI-SDS and F as independent variables and one of the component of the MS as the dependent variable, F was a weak predictor of the variability when DP and Glyc were introduced as dependent variables in Group 2 and when SP was introduced as dependent variable both in groups 1 and 2. When patients were subdivided into subgroups according to the IDF criteria, in Group 2 patients with one or more components of the MS had higher F concentrations. Conclusions In this cohort of obese children and adolescents, F was weakly associated with components of the MS. These findings do not support a major role for F in the development of MS. C. Guzzetti S. Pilia A. Ibba S. Loche (&) Pediatric Endocrinology Unit, Microcitemico Hospital, ASL Cagliari, Via Jenner, 09121 Cagliari, Italy e-mail: [email protected]
Keywords Cortisol Metabolic syndrome Obesity Children Adolescents Abbreviations BMI Body mass index DP Diastolic blood pressure F Cortisol Glyc Serum glucose HOMA Homeostatic Model Assessment IDF International Diabetes Federation MS Metabolic syndrome SP Systolic blood pressure TG Triglycerides WC Waist circumference WCp Waist circumference percentile
Introduction In recent decades the prevalence of childhood obesity has increased worldwide. Sedentary lifestyle and high calorie diets contributed to the increase in the prevalence of obesity and related complications. In 2012 the World Health Organization estimated that, globally, 200 million school-aged children are either overweight, and 40–50 million are classified as obese. Particularly, in the European region 22.1 % of boys and 20.3 % of girls are overweight, 5.3 % of boys and 4.4 % of girls are obese [1]. In Italy, in 2010, 26.2 % of children and adolescent between 6 and 17 years were overweight or obese [2]. Obesity is associated with dyslipidemia (elevated levels of triglycerides (TG) and low levels of HDL), hyperglycemia, insulin resistance, type 2 diabetes and hypertension.
123
52
J Endocrinol Invest (2014) 37:51–56
All these factors are components of the metabolic syndrome (MS). Metabolic syndrome has many similarities with Cushing’s syndrome. It has been hypothesized that subtle increases in cortisol levels could have an important role in the development of insulin resistance and MS. In this regard, many studies confirmed a correlation between MS and cortisol in adults, demonstrating that increased cortisol levels and obesity, especially visceral fat, cooperate in determining glucose intolerance, raised blood pressure and dyslipidemia [3–8]. The identification of children at increased risk for developing MS, and thus predisposed to type 2 diabetes and cardiovascular disease in later life [9, 10], is of great clinical importance. Few studies demonstrated a relationship between cortisol and insulin resistance or MS in children and adolescents [11–15]. The aim of this study was to evaluate the relationship between morning serum cortisol concentrations and component of MS in a large number of obese children and adolescents. We then investigated the predictive value of cortisol concentrations and whether they would be of help in identifying children at risk of developing MS.
Patients and methods
In all children and adolescents height, weight, waist circumference (WC), systolic and diastolic blood pressure (SP and DP, respectively), morning serum concentrations of cortisol, glycemia, insulin, cholesterol HDL, TG were evaluated at entry, between 08.00 and 09.00 h, after an overnight fast, in supine position. Waist circumference was measured with an anthropometric tape midway between the lower rip margin and the iliac crest at the end of a gentle expiration. Homeostatic model assessment (HOMA index, Glucose (mmol/L) 9 Insulin (mU/L)/22.5) was used to evaluate insulin resistance. Body mass index (BMI) was calculated as weight divided by height squared. BMI-SDS was derived from the Italian reference data [16]. WC percentile (WCp) was defined for all patients according to Cook et al. [17]. MS was defined according to the criteria provided by the International Diabetes Federation (IDF) [10]. Accordingly, patients were subdivided into three age groups: in Group 1, patients aged 6–10 years; in Group 2, patients aged 10–16 years; and in Group 3, patients older than 16 years. In Group 1 MS cannot be diagnosed, but children with WCp C 90 and family history of MS, diabetes or cardiovascular disease require further evaluation. In Groups 2 and 3, MS is defined by the presence of central obesity plus any two of the other four factors altered (TG, HDL, blood pressure and glycemia) (Table 2).
Patients Assays Data were collected from 1,027 obese children and adolescents referred to our Endocrine Unit between 2002 and 2010. All children included in the study were seen in our Unit for obesity and were otherwise healthy. Their main clinical characteristics are summarized in Table 1.
Serum glucose was determined by the glucose oxidases method (Autoanalyser, Beckman Coulter, USA). Serum insulin and cortisol concentrations were measured using reagents provided by Siemens (Lianberis, UK) and an
Table 1 Clinical characteristics of the patients studied All patients (n = 1,027)
Group 1 (n = 408)
Group 2 (n = 390)
Group 3 (n = 89)
Sex (M/F)
475/552
175/233
210/180
24/65
Age (years)
10.51 ± 3.29
7.79 ± 1.03
12.62 ± 1.34
16.74 ± 0.89
Puberty (Prep/pub)
593/429
367/39
117/272
1/86
BMI-SDS
2.62 ± 0.51
2.65 ± 0.42
2.52 ± 0.42
2.76 ± 0.64
WC (cm) WCp
81.96 ± 11.24 92.37 ± 8.47
75.45 ± 6.89 94.07 ± 5.20
89.14 ± 9.38 90.77 ± 9.96
93.83 ± 13.72 85.75 ± 14.96 118 ± 17
SP (mmHg)
106 ± 15
98 ± 12
112 ± 14
DP (mmHg)
62 ± 9
58 ± 8
65 ± 8
69 ± 10
Glycemia (mmol/L)
5 ± 0.4
4.9 ± 0.4
5 ± 0.4
5 ± 0.4
HOMA
3.67 ± 2.46
2.77 ± 1.77
4.42 ± 2.7
4.69 ± 3
Insulin (mU/L)
16.98 ± 12.19
13.31 ± 10.63
19.82 ± 12.12
20.56 ± 12.59
HDL (mmol/L)
1.3 ± 0.3
1.4 ± 0.3
1.3 ± 0.3
1.2 ± 0.3
TG (mmol/L)
0.7 ± 0.4
0.7 ± 0.4
0.8 ± 0.4
0.7 ± 0.4
Cortisol (nmol/L)
290 ± 159
270 ± 136
306 ± 172
324 ± 157
123
J Endocrinol Invest (2014) 37:51–56
53
Table 2 Definition of MS according to the International Diabetes Federation [10] Age group
Obesity (WC)
6–9 years
C90th percentile
In this age group MS cannot be diagnosed.
10–15 years
C90th percentile
C150 mg/dl
C16 years
C94 cm in M
C150 mg/dl
C80 cm in F
Triglycerides
HDL
Blood pressure
Glucose
\40 mg/dl
SP C 130 or DP C 85 mmHg
C100 mg/dl
\40 mg/dl in M
SP C 130 or DP C 85 mmHg
C100 mg/dl
\50 mg/dl in F
Immulite system (IMMULITE 2000). TG and HDL were measured by standard routine methods.
Table 3 Univariate correlations between cortisol and component of MS Partial correlationsa
Correlations
Statistical analysis Distribution of the data was evaluated using Kolmogorov– Smirnov test. Glycemia in Group 2 and cortisol, glycemia, HDL and WC in Group 3 were the only normally distributed variables. Univariate analyses were performed using Pearson correlation coefficient for normally distributed variables, and using Spearman rank order correlation for not normally distributed variables. All correlations were also adjusted for age, gender, puberty and BMI-SDS. Comparisons between two groups were performed using the Mann–Whitney U test. Comparisons between multiple groups were performed using ANOVA and ANCOVA, with age, gender, puberty and BMI-SDS as covariates, followed by the Bonferroni’s multiple comparison test. Multivariate regression analysis was performed to determine independent predictors of the MS components. The covariates entered into the model as potential predictors included sex, age, BMI-SDS, pubertal status and cortisol. All components of the MS were used as dependent variables, one at a time. All values are reported as mean ± SD (continuous variables) or as counts and percentages (categorical variables). P \ 0.05 (two sided) was considered significant. All statistical calculations were performed using STATISTICA version 9.1 software (StatSoft Inc., Tulsa, OK, USA).
Results
r
P
r
P
WCp
0.01
0.74
0.02
0.47
SP
0.17
1027
0.15
1024
DP Glycemia
0.13 0.16
8 3 1025 1027
0.09 0.10
0.007 0.001
HDL
0.007
0.81
0.02
0.46
TG
0.02
0.45
-0.03
0.35
HOMA
0.15
2 3 1026
0.13
1024
Significant values in bold letters a
Adjusted for age, gender, pubertal status and BMI-SDS
patients (3.33 %) had MS. In Group 3 (n = 89), 6 patients (6.74 %) had MS. Univariate regression analysis The results of all correlation analyses are reported in Table 3. Univariate regression analysis in the entire cohort showed that cortisol was weakly associated with SP, DP, glycemia and HOMA, also when adjusted for age, gender, puberty and BMI-SDS. In Group 1, cortisol was correlated with glycemia, and with SP and HOMA when adjusted for age, gender, puberty and BMI-SDS. In Group 2, cortisol was correlated with SP, DP, glycemia and HOMA. The correlation remained significant also when adjusted for age, gender, puberty and BMI-SDS. In Group 3, cortisol was not correlated with components of the MS.
Patients’ characteristics Multivariate regression analysis Clinical characteristics of the patients (552 girls and 472 boys) are summarized in Table 1. According to the IDF criteria [10], patients were subdivided into three age groups. In Group 1 (n = 408), 280 patients had a WCp [ 90th, and thus 68.63 % were at risk to developing MS (according to the IDF criteria, WCp in this group is the only factor which suggests MS). In Group 2 (n = 390), 13
In Group 1, results of multivariate regression analysis including age, sex, puberty, BMI-SDS and cortisol as independent variables and one of the component of the MS as the dependent variable, indicated that cortisol was a weak predictor of the variability only in the model with SP as dependent variable (P = 0.003) (Table 4).
123
54
J Endocrinol Invest (2014) 37:51–56
Table 4 Adjusted R2 in models of multivariate analysis with sex, age, puberty, BMI-SDS and cortisol among independent variables Depend variable
Group 1
Group 2
Group 3
WCp (WC in group 3)
0.220
0.139
0.740
SP
0.079
0.276
0.249
DP
0.049
0.134
0.120
-0.006
0.049
0.073
HDL
0.004
0.006
0.021
TG
0.024
0.017
-0.010
Glycemia
Table 5 ANOVA and ANCOVA for groups 2 and 3. Subgroups analyzed: 2A: WCp \ 90; 2B: WCp [ 90 and the other components in the normal range; 2C: WCp [ 90 and 1 of the other components of the MS abnormal; 2D: WCp [ 90 and 2 of the other components of the MS abnormal; 3A: WC \ 94 cm for M and WC \ 80 cm for F; 3B: WC C 94 cm for M and WC C 80 cm for F and the other components in the normal range; 3C: WC C 94 cm for M and WC C 80 cm for F and 1 of the other components of the MS abnormal; 3D: WC C 94 cm for M and WC C 80 cm for F and 2 of the other components of the MS abnormal Group 2 F
Group 3 P
F
P
ANOVA
7.12
0.0001
0.59
0.63
ANCOVAa
4.53
0.004
0.67
0.58
a
Covariates: age, gender, puberty and BMI-SDS
In Group 2, cortisol was a predictor of the variability in the models with SP, DP and glycemia as dependent variables (P = 4 9 10-4; P = 0.002; P = 7 9 10-4, respectively) (Table 4). In Group 3, cortisol was never a significant predictor of the variability (Table 4). In order to further investigate the effect of cortisol on the variability of the component of MS, multivariate regression analysis was repeated excluding cortisol among the independent variables. The results obtained confirmed that the effect of cortisol as predictor of MS was very weak, if any (data not shown). ANOVA and ANCOVA Patients of Group 1 were subdivided into two subgroups according to the IDF criteria [10]: subgroup 1A with WCp \ 90; subgroup 1B with WCp [ 90. Mean ± SD cortisol concentration was not different in the two subgroups (P = 0.87). Patients of Groups 2 and 3 were subdivided into four subgroups according to the IDF criteria [10] and ANOVA and ANCOVA between the four subgroups were performed (Table 5). In Group 2, mean ± SD cortisol concentration
123
Fig. 1 Mean ± SD serum cortisol levels in the 4 subgroups of Group 2. 2A patients with WCp \ 90; 2B patients with WCp [ 90 and the other components in the normal range; 2C patients with WCp [ 90 and 1 of the other components of the MS abnormal; 2D patients with WCp [ 90 and 2 of the other components of the MS abnormal
was significantly different among the four subgroups. Patients with one or more components of the MS had higher F concentration (Fig. 1). In Group 3, mean ± SD cortisol concentration was not different between the four subgroups (Table 5).
Discussion MS in adults is defined as a group of cardiovascular and diabetes risk factors which include abdominal obesity, dyslipidemia, glucose intolerance, and hypertension [18]. At the moment, there is no consensus on the definition of pediatric MS. This is mostly due to physiological developmental changes of the metabolic traits used to define MS during childhood and puberty [19]. Among the many definitions of MS in children, each with advantages and disadvantages, the IDF definition of MS [10], based on accessible diagnostic tools, best suits the pediatric and adolescent population. The IDF definition, because of the age-related developmental changes, considers abdominal obesity (WC), lipid profile (TG and HDL), blood pressure and serum glucose levels in different age groups: 6 to \10, 10 to \16 and C16 years. To our knowledge, data about MS in children and adolescents are limited, only few studies evaluated the role of cortisol in MS in childhood [11–15], and none based on IDF criteria. According to the IDF definition, 68.6 % of our patients aged 6–10 years had high risk to developing MS, 3.3 % of our patient aged 10–16 years had MS and 6.7 % of our patients aged [16 years had MS. In our large cohort we found no correlation between cortisol and WC, in accordance with previous studies showing no correlation or a weak correlation between
J Endocrinol Invest (2014) 37:51–56
cortisol and indices of visceral fat accumulation [11–13, 20]. Conversely, other studies have shown a significant correlation between serum cortisol levels and WC in adult women [6, 21], in men [22] or in both sexes [23]. Glucocorticoids are involved in differentiation, proliferation and redistribution of human adipocytes from peripheral to central depots. Glucocorticoids also increase the size and number of adipocytes. Overall, there is an increase of the HPA axis responsiveness to different stimuli in patients with abdominal obesity [7, 24]. Furthermore the negative feedback on HPA axis is attenuated in obesity, with reduced diurnal variation in cortisol levels. Adam et al. [11] hypothesized that the relationship between cortisol and visceral fat accumulation could not be evident in children and adolescent because it is a consequence of prolonged glucocorticoid exposure. We found a significant correlation between cortisol and both SP and DP in all patients considered together and in Group 2. Cortisol was correlated with SP also in Group 1. Results of multivariate correlation analysis showed that cortisol was a significant predictor of the variability of SP in Group 1 and 2, and of DP in Group 2. However, albeit significant, the effect of cortisol on the variability of blood pressure was as little as 2 %. A relationship between cortisol and blood pressure has been observed also in other studies on adult subjects [3–5, 8, 12, 20, 21]. In one study cortisol was found correlated with SP also in children and adolescents [11]. Among the possible mechanisms to explain the correlation between cortisol and blood pressure are the activation of the HPA axis and sympathetic nervous system caused by stress [7, 12]. An increased responsiveness to vasoconstrictors with a decreased vasodilator production, like nitric oxide [7], an increase in cardiac output or peripheral resistance and/or mineralocorticoid-induced salt and water retention could also play a role [5]. We also found a significant correlation between cortisol and fasting morning glycemia in all patients considered together, as well as separately in Group 1 and in Group 2. In multivariate analysis cortisol was a significant weak predictor (3 %) of the variability of glycemia only in Group 2. These results are consistent with other studies showing a significant correlation between cortisol and glycemia in children and adolescents [11, 12], as well as in adult subjects [3, 4, 8, 20, 22, 25]. We found a significant correlation between cortisol and HOMA index, as indicator of insulin resistance, in all patients considered together, as well as in Group 1 and 2. Many studies have reported a significant correlation between serum cortisol and fasting plasma insulin, both in adult men and women [3, 15, 21]. Adam et al. [11] showed a correlation between cortisol and Disposition Index (an indicator of b-cell function) and acute insulin release to a glucose challenge in children and adolescents, suggesting
55
that cortisol may be associated with the decrease in insulin sensitivity. Also there is a significant correlation between cortisol and insulin resistance in both obese adults and children [4, 5, 7]. No correlation was found between serum cortisol levels and lipid serum profile, confirming the results of Weigensberg et al. [12]. Conversely, many other studies performed in adult obese subjects showed a significant correlation between cortisol and TG [4, 6, 21], between cortisol and total cholesterol, HDL, LDL [2, 8, 20] or both [5, 7]. In Group 2, we found that patients with one or more components of the MS had higher serum morning cortisol concentrations. Nevertheless, we found no relationship between serum cortisol levels and the component of MS in Group 3. Taken together, our results do not confirm an important effect of cortisol in the development of MS. Some authors [5, 8, 15, 21] have suggested that in obesity, although cortisol excretion is increased, also owing to an increased production of ACTH, circulating cortisol may be normal or low because of an increased peripheral clearance. They hypothesized that peripheral clearance of cortisol is enhanced in obesity and that central control of the HPA axis is altered, with reduced diurnal variation in cortisol levels (lower in the morning but higher in the evening). This might explain the fact that we found only weak correlations between morning serum cortisol and the component of MS. Other authors [12] hypothesized that this relationship would be mediated by insulin sensitivity since it does exist only with parameters of MS which are influenced by insulin sensitivity. This study presents some limitations. This is a retrospective study performed in a very large number of obese children and adolescent, and we used all the available information. Therefore, we have correlated morning serum cortisol values, which could not reflect 24 h cortisol production, although it has been previously shown that even a single morning measurement of serum cortisol reflects HPA activity [11]. Furthermore, cortisol serum levels are related to many factors, some of which could be controlled for their impact on the analysis (morning samples, adjusting analysis by sex, age, puberty, …), while others, like cortisol response to physiological or psychological stressors, could not be controlled and may have caused bias in data analysis. Moreover, we did not perform the same study in a control group, and we do not know whether cortisol correlates with components of the MS also in normal weight children. However, the large number of patients and the homogeneity of our cohort provide strength to the results. In conclusion, albeit we have shown that patients with MS have slightly higher cortisol concentrations, we found
123
56
only weak or no correlations between cortisol and component of MS, suggesting that cortisol does not play an important role in the development of MS in obese children and adolescents. Acknowledgments Supported in part by a Grant from Regione Autonoma della Sardegna to SL. We are gratefully indebted with our nurse staff (Donatella Arghittu and Patrizia Sanna), and our Lab technicians (Maria Grazia Contini, Danilo Mosino, Teresa Trogu) for their continuous and skilful help in patients’ care and sample processing. Conflict of interest The authors C. Guzzetti, S. Pilia, A. Ibba, and S. Loche declare that they have no conflict of interest.
References 1. http://www.iaso.org/iotf/obesity/obesitytheglobalepidemic/. Accessed Aug 2012 2. http://www.istat.it/it/archivio/43508. Accessed 25 Oct 2010 3. Rosmond R, Dallman MF, Bjo¨rntorp P (1998) Stress-related cortisol secretion in men: relationships with abdominal obesity and endocrine, metabolic and hemodynamic abnormalities. J Clin Endocrinol Metab 83:1853–1859 4. Ward AM, Fall CH, Stein CE et al (2003) Cortisol and the metabolic syndrome in South Asians. Clin Endocrinol 58: 500–505 5. Whitworth JA, Williamson PM, Mangos G, Kelly JJ (2005) Cardiovascular consequences of cortisol excess. Vasc Health Risk Manag 1:291–299 6. Esteghamati A, Morteza A, Khalilzadeh O, Noshad S, Novin L, Nakhjavani M (2011) Association of serum cortisol levels with parameters of metabolic syndrome in men and women. Clin Invest Med 34:E131–E137 7. Anagnostis P, Athyros VG, Tziomalos K, Karagiannis A, Mikhailidis DP (2009) Clinical review: the pathogenetic role of cortisol in the metabolic syndrome: a hypothesis. J Clin Endocrinol Metab 94:2692–2701 8. Reynolds RM, Syddall HE, Walker BR, Wood PJ, Phillips DI (2003) Predicting cardiovascular risk factors from plasma cortisol measured during oral glucose tolerance tests. Metabolism 52: 524–527 9. Capizzi M, Leto G, Petrone A et al (2011) Wrist circumference is a clinical marker of insulin resistance in overweight and obese children and adolescents. Circulation 123:1757–1762 10. Zimmet P, Alberti KG, Kaufman F et al (2007) The metabolic syndrome in children and adolescents—an IDF consensus report. Pediat Diabetes 8:299–306
123
J Endocrinol Invest (2014) 37:51–56 11. Adam TC, Hasson RE, Ventura EE et al (2010) Cortisol is negatively associated with insulin sensitivity in overweight latino youth. J Clin Endocrinol Metab 95:4729–4735 12. Weigensberg MJ, Toledo-Corral CM, Goran MI (2008) Association between the metabolic syndrome and serum cortisol in overweight latino youth. J Clin Endocrinol Metab 93:1372–1378 13. Misra M, Bredella MA, Tsai P, Mendes N, Miller KK, Klibanski A (2008) Lower growth hormone and higher cortisol are associated with greater visceral adiposity, intramyocellular lipids, and insulin resistance in overweight girls. Am J Physiol Endocrinol Metab 295:E385–E392 14. Ten S, Maclaren N (2004) Insulin resistance syndrome in children. J Clin Endocrinol Metab 89:2526–2539 15. Reinehr T, Andler W (2004) Cortisol and its relation to insulin resistance before and after weight loss in obese children. Horm Res 62:107–112 16. Cacciari E, Milani S, Balsamo A, Spada E, Bona G, Cavallo L (2006) Italian cross-sectional growth charts for height, weight and BMI (2 to 20 yr). J Endocrinol Invest 29:581–593 17. Cook S, Auinger P, Huang TT (2009) Growth curves for cardiometabolic risk factors in children and adolescents. J Pediatr 155:S6.e15–26 18. Alberti KG, Zimmet P, Shaw J (2006) Metabolic syndrome—a new world-wide definition. A Consensus Statement from the International Diabetes Federation. Diabet Med 23:469–480 19. Goodman E, Daniels SR, Meigs JB, Dolan LM (2007) Instability in the diagnosis of metabolic syndrome in adolescents. Circulation 115:2316–2322 20. Park SB, Blumenthal JA, Lee SY, Georgiades A (2011) Association of cortisol and the metabolic syndrome in korean men and women. J Korean Med Sci 26:914–918 21. Walker BR, Soderberg S, Lindahl B, Olsson T (2000) Independent effects of obesity and cortisol in predicting cardiovascular risk factors in men and women. J Intern Med 247:198–204 22. Wallerius S, Rosmond R, Ljung T, Holm G, Bjo¨rntorp P (2003) Rise in morning saliva cortisol is associated with abdominal obesity in men: a preliminary report. J Endocrinol Invest 26:616–619 23. Fraser R, Ingram MC, Anderson NH, Morrison C, Davies E, Connell JM (1999) Cortisol effects on body mass, blood pressure, and cholesterol in the general population. Hypertension 33:1364–1368 24. Bjo¨rntorp P, Rosmond R (2000) Obesity and cortisol. Nutrition 16:924–936 25. Phillips DI, Barker DJ, Fall CH (1998) Elevated plasma cortisol concentrations: a link between low birth weight and the insulin resistance syndrome? J Clin Endocrinol Metab 83:757–760
