Abstract
Introduction: Some clinical resemblance may exist between obesity, particularly abdominal obesity, and Cushing’s syndrome. This has stimulated ongoing interest in the role of cortisol’s secretion pattern, control, and metabolism in obesity. Goals: The aim of the study was to investigate whether basal and stimulated levels of cortisol differ between healthy people with obesity and individuals with normal weight. Methods: Total, free, and salivary cortisol was tested at baseline state and after 1 μg ACTH stimulation in 60 healthy subjects with obesity and 54 healthy lean controls. Results: Baseline total cortisol was lower in subjects with obesity compared to lean controls (347 [265–452] nmol/L vs. 422 [328–493] nmol/L, respectively; p < 0.05). Similarly, basal salivary cortisol was significantly lower in subjects with obesity (7.5 [5.2–9.7] nmol/L vs. 10.7 [7.5–17.6] nmol/L; p < 0.05). Upon challenge with ACTH, total peak serum and salivary peak cortisol responses were significantly lower in people with obesity than in lean subjects (665.16 ± 151.8 vs. 728.64 ± 124.2 nmol/L; p < 0.05 and 31.66 [19–38.64] vs. 40.05 [31.46–46.64] nmol/L; p < 0.05, respectively). Additionally, baseline total cortisol and salivary cortisol were inversely related to BMI (r = −0.24, r = −0.27; p < 0.05 for both) and waist circumference (r = −0.27, r = −0.34; p < 0.05 for both). Conclusion: Baseline as well as peak stimulated total serum and salivary cortisol were significantly lower in subjects with obesity. It thus appears that obesity is not associated with enhanced basal or ACTH-stimulated cortisol.
Introduction
There is considerable confusion regarding the secretion, bioavailability, metabolism, and role of cortisol in human obesity. Owing to some obvious similarities in the clinical and biochemical features of true hypercortisolism and obesity, particularly central obesity, the latter has previously been referred to as a “pseudo-Cushing’s” state [1]. Multiple reports dealing with specific aspects of cortisol secretion, plasma levels, metabolism, and excretion in obesity seem to collectively, though not uniformly, suggest some ill-defined form of increased and/or abnormal cortisol availability or response in obesity [2]. Early reports indicated increased urinary excretion of cortisol metabolites, which was interpreted as evidence of increased secretion of cortisol coupled with accelerated metabolism and turnover in subjects with obesity. This was further supported by evidence for increased conversion of cortisol to cortisone as well as enhanced 5α reduction in people with obesity [3‒5]. These findings suggest that obesity may involve slightly higher cortisol levels, which are often undetected due to efficient metabolic clearance. This could indicate increased HPA axis activity and possibly greater tissue exposure to cortisol in obese individuals [6, 7]. Several cortisol stimulation procedures were found to elicit an accentuated cortisol response in people with obesity. This included some extensively studied dynamic tests such as the intravenous ACTH challenge as well as less commonly used procedures such as the CRH/AVP tests or psychological stress challenge [2, 8‒11]. Curiously, though, some individuals with obesity, such as those with a binge eating disorder, actually display diminished cortisol response to the Trier Social Stress Test [12]. The cortisol wakening response has showed a positive correlation between waist/hip ratio and cortisol response in both men and women [13‒15]. But there seemed to be a high variability of cortisol response [14, 15].
In view of this cumulative evidence favoring increased unprovoked secretion and easiness to trigger cortisol response in obesity, the findings in some studies that baseline morning total cortisol (TC) levels were inversely related to obesity and waist circumference (WC) both in men and women [16‒18] were somewhat unexpected. The observation that serum free cortisol (FC) actually increased in subjects with obesity only after profound weight loss due to bariatric surgery report was also not entirely in line with the proposed direct linkage of obesity to cortisol [19].
Here we set out to examine the possibility that otherwise healthy subjects with obesity differ from normal weight individuals in terms of circulating FC and salivary cortisol (SC). Because baseline cortisol may vary in the same individual [20], we also tested the FC and SC response to 1 µcg ACTH, a submaximal challenge for cortisol reserve, which, in our longstanding experience, outperforms the standard high dose test, in sensitivity and overall accuracy. This finding aligns with numerous discussions in the scientific literature on the topic [21, 22].
Methods
Study Design and Subjects
Sixty healthy men and women with obesity (mean BMI 40 ± 6 kg/m2), recruited from our obesity clinic, and fifty-four age- and sex-matched lean controls (mean BMI 23 ± 3 kg/m2) participated in the study (Table 1). Of note, the control group comprised 4 subjects with BMIs 27–28 and 5 subjects with BMIs in the 25–26 range. Data were collected prospectively, with subjects matched by age and gender. Exclusion criteria included diabetes (subjects with impaired fasting glucose were included), alcohol abuse, depression, current/recent corticosteroid use, and oral contraceptive use. To exclude diurnal interference with HPA axis, we excluded night shift workers from both groups. The study was conducted in compliance with the ethical principles based on the Declaration of Helsinki and the applicable International Conference on Harmonization (ICH) guidelines and was approved by the local Ethical Committee.
Anthropometric, clinical, and biochemical characteristics of the study groups
. | Subjects with obesity (n = 60) . | Lean controls (n = 54) . | p value for difference . |
---|---|---|---|
Mean age, years | 39±12 | 39±11 | ns |
Females:males, % | 72:28 | 70:30 | ns |
Smoking, n (%) | 8 (14) | 3 (5.6) | ns |
BMI, kg/m2 | 40±6 | 23±3 | <0.001 |
Waist circumference, cm | 116±18 | 81±10 | <0.001 |
Hip circumference, cm | 121 (114–134) | 94 (88–99) | <0.001 |
WHR | 0.99 (0.88–1.07) | 0.88 (0.79–0.97) | ns |
WHtR | 0.73 (0.66–0.77) (n = 24) | 0.48 (0.46–0.56) (n = 35) | <0.001 |
Systolic BP, mm Hg | 126±13 | 115±11 | <0.001 |
Diastolic BP, mm Hg | 79±10 | 70±10 | <0.001 |
Heart rate, beats/min | 75±11 | 68±12 | <0.001 |
Fasting plasma glucose, mg/dL | 90±13 | 75±8 | <0.001 |
Insulin, pmol/L | 174±97 | 86±51 | 0.001 |
HbA1c, % | 5.7±0.5 | 5.3±0.4 | <0.001 |
HOMA-IR, mg/dL × pmol/L | 6.4 | 2.7 | <0.01 |
Fasting triglycerides, mg/dL | 111 (80–144) | 67 (51–87) | <0.001 |
HDL, mg/dL | 49 (44–64) | 60 (52–74) | <0.01 |
WBC | 7.2 (6.2–8.6) | 6 (5.3–6.8) | <0.001 |
TSH, mIU/L | 2.3±0.9 | 1.8±0.8 | 0.01 |
Prolactin, µg/L | 8.55 (7.4–10.81) | 7.2 (5.6–13.7) | <0.001 |
. | Subjects with obesity (n = 60) . | Lean controls (n = 54) . | p value for difference . |
---|---|---|---|
Mean age, years | 39±12 | 39±11 | ns |
Females:males, % | 72:28 | 70:30 | ns |
Smoking, n (%) | 8 (14) | 3 (5.6) | ns |
BMI, kg/m2 | 40±6 | 23±3 | <0.001 |
Waist circumference, cm | 116±18 | 81±10 | <0.001 |
Hip circumference, cm | 121 (114–134) | 94 (88–99) | <0.001 |
WHR | 0.99 (0.88–1.07) | 0.88 (0.79–0.97) | ns |
WHtR | 0.73 (0.66–0.77) (n = 24) | 0.48 (0.46–0.56) (n = 35) | <0.001 |
Systolic BP, mm Hg | 126±13 | 115±11 | <0.001 |
Diastolic BP, mm Hg | 79±10 | 70±10 | <0.001 |
Heart rate, beats/min | 75±11 | 68±12 | <0.001 |
Fasting plasma glucose, mg/dL | 90±13 | 75±8 | <0.001 |
Insulin, pmol/L | 174±97 | 86±51 | 0.001 |
HbA1c, % | 5.7±0.5 | 5.3±0.4 | <0.001 |
HOMA-IR, mg/dL × pmol/L | 6.4 | 2.7 | <0.01 |
Fasting triglycerides, mg/dL | 111 (80–144) | 67 (51–87) | <0.001 |
HDL, mg/dL | 49 (44–64) | 60 (52–74) | <0.01 |
WBC | 7.2 (6.2–8.6) | 6 (5.3–6.8) | <0.001 |
TSH, mIU/L | 2.3±0.9 | 1.8±0.8 | 0.01 |
Prolactin, µg/L | 8.55 (7.4–10.81) | 7.2 (5.6–13.7) | <0.001 |
WHR, waist to hip ratio; WHtR, waist to height ratio.
Values are mean ± SD for normal distributed values, median (IQR) for other.
Testing Procedure
Morning ACTH, serum TC, FC, and SC levels were drawn using an indwelling venous catheter after 30 min of rest in the recumbent position, between 8:00 and 9:00 a.m. Low dose, 1 µcg of ACTH Synacthen (tetracosactrin, Novartis, France) was prepared and administered as an IV bolus as previously described [23]. Based on previous reports [24‒26], criteria for a normal response were a peak TC >500 nmol/L, peak free cortisol >25 nmol/L [19], and peak SC >27.6 nmol/L. Samples were collected 20, 30, and 40 min after ACTH injection, in accordance with the pharmacokinetics of the challenge [25, 27].
Assays
TC was measured using an electro-chemiluminescence immunoassay (Cobas e411, Roche, Mannheim, Germany). SC was analyzed using modification of radioimmunoassay coated tube, Count-A Count-Cortisol (Siemens, Malvern, PA, USA) as previously described by us [20, 25]. Serum free cortisol was measured using an in-house method, as previously described [26]. ACTH was analyzed using solid-phase, two-site chemiluminescent immunometric assay on Immulite 2000 XPi (Siemens, Malvern, PA, USA).
Statistics
The collected data were analyzed with SPSS, version 29.0.1.0. Data were inspected for normality and are presented as means and standard deviations (M) or as median and interquartile range, accordingly. Comparison between groups were done using t test for normally distributed variables and with the Mann-Whitney test for nonparametric variables. We calculated sample size required for a 20% difference between TC with a power of 80 and 2-tailed alpha error 0.05 as a minimum of at least 23 subjects in each group. The study cohort was expanded to ensure balance in age and gender distributions as both factors were identified as influencing cortisol levels [27, 28]. Linear regression was done using mixed-effects models, where the random effect is the participant and the fixed effect is the interaction term (the cross-product term). The area under the curve (AUC) for cortisol response was quantified using a previously described methodology [29].
The Mann-Whitney test was applied to assess the differences in baseline characteristics and blood test measures, and the Wilcoxon signed-rank test was applied to assess the change in the AUC of all the tested measures. A p value <0.05 was considered statistically significant. All reported p values are 2 tailed.
Results
Table 1 summarizes the baseline characteristics of the subjects in the study. As expected, subjects with obesity had statistically higher WC, waist to hip ratio, waist to height ratio, systolic and diastolic blood pressure, as well as higher heart rate. Markers of metabolic syndrome such as fasting plasma glucose, insulin, HOMA-IR, HbA1c, and triglyceride levels were higher in subjects with obesity as compared to controls. Levels of HDL were significantly lower.
Serum baseline ACTH concentrations were similar in individuals with obesity and lean controls (3.92 [2.99–5.24] pmol/L and 4.38 [2.73–5.9] pmol/L, respectively; p = ns). Median (± standard deviations) basal TC was within the accepted normal range in all subjects but ∼20% lower in individuals with obesity relative to lean subjects (347 [265–452] nmol/L vs. 422 [328–493] nmol/L; p < 005). Median basal SC was also significantly lower in subjects with obesity (7.5 [5.2–9.7] nmol/L; p < 0.05 vs. 10.7 [7.5–17.6] nmol/L).
Following stimulation with 1 μg ACTH, serum TC was significantly lower in subjects with obesity as compared to the lean subjects after 30 and 40 min (670 [552–758] nmol/L vs. 723 [662–767] nmol/L; p < 0.05; 607 [505–717] nmol/L vs. 717 [676–770]; p < 0.001) (Fig. 1). Stimulated SC was also found to be significantly lower in participants with obesity as compared to controls, at 30 and 40 min (30 [19–38] nmol/L vs. 33 [30–44] nmol/L; p < 0.05; 30 [19–38] nmol/L vs. 36 [27–41] nmol/L; p < 0.001) (Fig. 2).
Total serum cortisol as function of time in a low dose 1 μg ACTH challenge (*p < 0.05).
Total serum cortisol as function of time in a low dose 1 μg ACTH challenge (*p < 0.05).
SC as function of time in a low dose 1 μg ACTH challenge (*p < 0.05).
AUC for TC was significantly lower in subjects with obesity (805 ± 22 vs. 866 ± 23; p = 0.028). There was no difference between AUC for serum free and SC. The percentage of free cortisol (FC) relative to total cortisol (TC), as measured by the AUC, was lower in subjects with obesity, though this difference did not reach statistical significance (p = 0.077).
There was a negative correlation between BMI and basal TC (r = −0.24; p < 0.05). Of note, this correlation remained significant in women only (r = −0.25; p < 0.05), while basal SC was negatively correlated only in men (r = −0.56; p < 0.05). Similarly, a conventional measure of central fat accumulation, WC, was also negatively related to baseline TC (r = −0.26; p < 0.05). Additionally, triglycerides were negatively correlated to basal TC (r = −0.23; p < 0.05). Basal SC was also negatively correlated to both BMI and WC (r = −0.27, r = −0.33; p < 0.01 for both) (Fig. 3). Interestingly, basal SC was also negatively correlated with waist-to-height ratio (r = −0.39; p = 0.01). Correlation of basal FC and SC with BMI did not reach statistical significance (r = −0.19; p = 0.06; r = −0.27; p = 0.09, respectively).
a Correlation between basal SC and BMI (using Spearman correlation). b Correlation between basal SC and WC (using Pearson correlation).
a Correlation between basal SC and BMI (using Spearman correlation). b Correlation between basal SC and WC (using Pearson correlation).
Discussion
The key finding in our study is that baseline cortisol is actually lower in people with obesity than in lean, age- and sex-matched individuals. This was corroborated through two independent measurements: serum TC and SC, which were ∼20%, ∼30% lower than the respective levels seen in control subjects. In agreement with these basal measurements, the cortisol response to a low dose (1 μg) ACTH stimulation was also lower in the subjects with obesity as compared to the lean controls. A previous study has shown similar results for TC [30], but, to the best of our knowledge, this is the first study to show the decreased responsiveness of SC to the low dose ACTH stimulation in people with obesity compared to normal weight controls.
It might be argued that the 1 μg ACTH dose should have been adjusted according to weight, to avert the potential problem of increased distribution volume in the obese state, leading to lesser effective serum ACTH during this challenge in obesity but not in the lean subjects. While this point cannot be entirely discarded, we believe that it is unlikely to have significantly impacted our findings: first, Olkers et al. [31] measured serum ACTH levels following the 1 μg challenge and reported them to be risen to ∼1,900 pg/mL, much higher than ever measured in the course of a physiologically relevant stress. Pasquali et al. [10] using a considerably larger dose of ACTH (0.2 μg/kg) failed to detect a significant difference in the cortisol response in women with visceral obesity versus those with subcutaneous obesity. Concordant with the latter is a report that serum cortisol was not different between insulin-sensitive and insulin-resistant obese adolescents [32].
It is noteworthy that the 1 μg test has been now successfully used in many centers for more almost 3 decades and a fixed dose, rather than a body surface- or a BMI-related dose, is routinely applied. The observation in our study that baseline and peak TC levels are not only significantly higher in lean subjects, but are also negatively correlated with BMI and WC are clearly not suggestive, and in fact are actually inconsistent with a putative state of “irritable adrenal cortex” in obesity. If anything, cortisol secretion, when thus assessed, focusing on free cortisol measured in the SC, appears to be somewhat restrained in subjects with obesity, albeit still clearly within the defined normal range. Taken one step further, these findings add another piece of evidence refuting the common belief that HPA overactivity begets obesity. In general, these findings are concordant with other large cohorts in which SC and TC were negatively correlated with BMI and WC [33‒35]. An increased activity of 11beta-hydroxysteroid type 1 in metabolic tissues (adipose tissue, liver, skeletal muscle, etc.), suggested by some studies, may cause an increased level of cortisol in those tissues rather than the serum and explain the hypercotisolemic phenotype of obesity [36].
Interestingly, some of the studies have shown lower decline in cortisol levels in the evening, thus flattening diurnal cortisol rhythm in children [37] as well as adults [8, 38‒40] with obesity, which may suggest disruption of the cortisol circadian rhythms. This may also resolve the discrepancy seen between studies showing elevated total daily cortisol (using assays as UFC [3] or hair cortisol assays [41]) in subjects with obesity and the decreased morning and awakening levels and responsiveness as seen in our study.
In our study, we did not find any differences in FC in people with obesity as compared to controls. Although previous studies have found a good correlation between SC and FC [42], the discrepancy found between SC and FC may be explained by difference in variability between those assays (higher in SC) [42]. Interestingly, CBG levels were found to be negatively correlated with BMI; thus, lower levels in people with obesity may influence FC levels but not SC [43].
Our findings diverge from those reported by Lu et al. [44], as we observed no significant differences in ACTH levels between individuals with obesity and controls. This discrepancy might be attributed to demographic variations in the study populations, with our cohort comprising Caucasian participants, whereas Lu et al. [44] examined a Chinese population.
The limitation of our study is the small number of subjects. Also, the study’s single-center design may introduce bias, potentially limiting the generalizability of results to broader populations.
Conclusion
The observation of lower baseline and stimulated total and salivary cortisol levels, along with the negative correlations between cortisol and BMI/WC, clearly does not support the concept of HPA axis overactivity or heightened cortisol secretion in obesity. If anything, cortisol secretion appears somewhat restrained in otherwise healthy individuals with obesity, albeit still within the defined normal range. Whether this is due to increased cortisol metabolism in obesity [4] or other factors (disturbed HPA circadian rhythm?) remains to be discovered.
Statement of Ethics
This study protocol was reviewed and approved by the local Ethical Committee at Ichilov Sourasky Medical Center (Approval No. IRB# 0574-08 TLV). Written informed consent was obtained from all participants.
Conflict of Interest Statement
All authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Funding Sources
The authors received no funding for the research, authorship, or publication of this article.
Author Contributions
Y.S.: design of work and drafting of manuscript. E.O., Y.G., and K.T.: revision of manuscript. W.A.A.: statistical analysis. Y.M., S.S., M.Y., and M.S.: data collection. N.S.: substantial conception of work and drafting of manuscript.
Data Availability Statement
Due to IRB requirements, personal patient data are unidentified and not available to the general public. In specific cases, the data will be shared from Y.S. according to the guidelines of our local IRB.