Hypothalamic-pituitary-adrenal axis dysregulation and cortisol activity in obesity: A systematic review

Highlights

• HPA activity and obesity are related, but the literature is highly inconsistent.

• Much of this inconsistency arises from divergent methodologies across studies.

• This systematic review finds obesity may be related to HPA hyperresponsivity.

• Dysregulation in adipocyte cortisol metabolism may affect HPA activity in obesity.

Abstract

Background

Although there is substantial evidence of differential hypothalamic-pituitary-adrenal (HPA) axis activity in both generalized and abdominal obesity, consistent trends in obesity-related HPA axis perturbations have yet to be identified.

Objectives

To systematically review the existing literature on HPA activity in obesity, identify possible explanations for inconsistencies in the literature, and suggest methodological improvements for future study.

Data sources

Included papers used Pubmed, Google Scholar, and the University of California Library search engines with search terms body mass index (BMI), waist-to-hip ratio (WHR), waist circumference, sagittal diameter, abdominal versus peripheral body fat distribution, body fat percentage, DEXA, abdominal obesity, and cortisol with terms awakening response, slope, total daily output, reactivity, feedback sensitivity, long-term output, and 11β-HSD expression.

Study eligibility criteria

Empirical research papers were eligible provided that they included at least one type of obesity (general or abdominal), measured at least one relevant cortisol parameter, and a priori tested for a relationship between obesity and cortisol.

Results

A general pattern of findings emerged where greater abdominal fat is associated with greater responsivity of the HPA axis, reflected in morning awakening and acute stress reactivity, but some studies did show underresponsiveness. When examined in adipocytes, there is a clear upregulation of cortisol output (due to greater expression of 11β-HSD1), but in hepatic tissue this cortisol is downregulated. Overall obesity (BMI) appears to also be related to a hyperresponsive HPA axis in many but not all studies, such as when acute reactivity is examined.

Limitations

The reviewed literature contains numerous inconsistencies and contradictions in research methodologies, sample characteristics, and results, which partially precluded the development of clear and reliable patterns of dysregulation in each investigated cortisol parameter.

Conclusions and implications

The literature to date is inconclusive, which may well arise from differential effects of generalized obesity vs. abdominal obesity or from modulators such as sex, sex hormones, and chronic stress. While the relationship between obesity and adipocyte cortisol seems to be clear, further research is warranted to understand how adipocyte cortisol metabolism influences circulating cortisol levels and to establish consistent patterns of perturbations in adrenal cortisol activity in both generalized and abdominal obesity.

Introduction

In an age where obesity has reached epidemic levels and is implicated in several of the leading causes of death in the United States (Ogden et al., 2014), it is essential to understand the physiological correlates, predictors, and consequences of obesity. This systematic review examines the relationship between obesity and perturbations of the hypothalamic-pituitary-adrenal (HPA) axis. Understanding these perturbations in obesity is particularly important given that dysregulation of the HPA axis is a risk factor for physical health conditions such as cardiovascular disease, insulin resistance and type 2 diabetes, stroke, and Cushing’s syndrome (Pasquali et al., 2006, Rosmond and Bjorntorp, 2001), as well as mental health conditions such as depression and cognitive impairment (Hinkelmann et al., 2009, Reppermund et al., 2007). Additionally, in both human and animal models, cortisol has been causally demonstrated to promote the accumulation of fat cells and weight gain (Björntorp and Rosmond, 2000, Bjorntorp, 2001), implicating HPA axis functioning in the etiology of obesity. The literature, however, is inconsistent at best in terms of how the HPA axis is dysregulated in obesity. This gap precludes a comprehensive understanding of the pathophysiology of obesity. Furthermore, because it is unclear whether obesity contributes to HPA dysregulation or vice versa, it is difficult to identify appropriate intervention targets, thus hindering the development of effective treatments.

To aid in the reading of this review, we first briefly summarize the broader relationships among stress, cortisol, adipocyte biology, and weight gain (see Fig. 1). Fluctuations in cortisol concentrations occur according to a natural diurnal pattern as well as in response to both physiological and psychological stressors. Stress-related HPA activation begins in the hypothalamus, with the release of corticotrophin-releasing hormone. This stimulates the release of adrenocorticotropic hormone (ACTH) from the anterior pituitary gland, which in turn circulates through the bloodstream to the adrenal cortex, signaling the adrenal glands to secrete the hormone cortisol (Kyrou and Tsigos, 2009). Circulating cortisol can then be measured through saliva, blood (serum or plasma), or urine. Although plasma has been the “gold standard” for measuring cortisol in biological research, each of these three methods have been broadly and reliably used to measure cortisol concentrations (Kirschbaum and Hellhammer, 1994, Vining et al., 1983, Zumoff et al., 1974). When considering total cortisol output, salivary or urinary measurements may in fact be the better indices as they index free cortisol rather than bound cortisol (Yehuda et al., 2003). However, salivary and urinary cortisol are different measures and do not always correlate. This may be because urinary cortisol reflects total cortisol secreted whereas salivary cortisol reflects free cortisol already filtered from plasma, thus already accounting for clearance. Finally, while these methods index patterns of daily cortisol concentrations, hair can be used to measure long-term systemic cortisol output and has been shown to correlate with 24-h urinary free cortisol (UFC; Sauvé et al., 2007).

As psychological and physiological stressors can trigger cortisol secretion, both have implications for weight gain and obesity. Chronic psychological life stress, for example, has been linked to weight gain, especially in men (Torres and Nowson, 2007). In a meta-analysis of 13 studies including over 160,000 workers, job strain was found to be positively related to body mass index (BMI) both cross-sectionally and longitudinally (Nyberg et al., 2012). This is particularly relevant to obesity, as recent findings show that that obese individuals tended to report higher perceived stress (Abraham et al., 2013). Physiologically, increased cortisol concentrations have been causally linked to fat accumulation and weight gain, as glucocorticoids promote conversion of preadipocytes to mature adipocytes (for a review, see Peckett et al., 2011). In rodent models, male Sprague–Dawley rats that were stressed for 28 days showed significantly larger adipocytes than controls and had a tendency to display a heavier abdominal fat pad (Rebuffé-Scrive et al., 1992). This relationship is also echoed in human literature, as a large subset of findings have evidenced the relationship between stress and weight gain via elevated cortisol levels (e.g., Björntorp and Rosmond, 2000, Björntorp et al., 2000, Bjorntorp, 2001, Peeke and Chrousos, 1995, Wallerius et al., 2003). We do note, however, that cortisol’s role in fat physiology is very complex, and it does have other functions such as regulating lipolytic enzymes (Ottosson et al., 2000) and enhancing lipolysis and triglyceride uptake (Björntorp, 1996), thereby potentially causing weight loss in some instances. Cortisol also drives insulin resistance (Andrews and Walker, 1999) via proliferation of adipokines and the secretion of proinflammatory cytokines (Antuna-Puente et al., 2008). Effects may in fact be depot-specific. Cushing’s syndrome patients typically show truncal and visceral adipose depot gain, but peripheral (limb) adipose tissue depletion. Overall, though, it is clear that stress-related cortisol concentrations play a significant role in adipocyte biology and weight gain, potentially implicating it as a key component in the development of obesity. This review, therefore, endeavors to provide a more comprehensive understanding of cortisol activity and obesity in terms of perturbations in HPA axis activity.

This review has three overarching aims: (1) To provide a systematic review of the existing literature regarding HPA dysregulation and obesity. (2) To identify explanations for inconsistencies in the existing literature. (3) To suggest potential methodological improvements for future research directions to resolve discrepancies and achieve a definitive mechanistic understanding in this area.

In systematically reviewing the existing research on obesity and HPA axis activity, we make two important distinctions in our categorization of the literature:

We review cortisol activity and metabolism stemming both from the adrenal glands and from adipocyte metabolic sources, such as abdominal and subcutaneous adipose tissue and liver tissue (Fig. 1). Empirical evidence suggests that adipocyte cortisol metabolism may be a special case of cortisol dysregulation in obesity (Livingstone et al., 2000, Ljung et al., 2002, Ljung et al., 1996), and we examine the possibility that dysregulated adipocyte cortisol metabolism may relate to HPA dysregulation. General cortisol activity measurements include salivary, blood, urinary, and hair cortisol. These are used to assess cortisol parameters including the cortisol awakening response (CAR), total daily output, diurnal slope, reactivity, feedback sensitivity, and long-term output. Adipocyte cortisol metabolism is best indexed through measures of 11-β-hydroxysteroid dehydrogenases (11β-HSD1 and 2) from biopsied fat tissue.

The literature is highly varied regarding the anthropometric composition of study samples. For example, although obesity and adiposity are often linked, it is possible to concurrently have an obese body mass index (BMI; weight [kg]/height²[m]) and low abdominal obesity, or vice versa, a normal BMI and high abdominal obesity (Epel et al., 2001). Therefore, we consider two anthropometric types of obesity: generalized obesity, indexed crudely by BMI, and abdominal obesity and adiposity, indexed by measures such as waist-to-hip ratio (WHR; the ratio of the minimum circumference value between the iliac crest and the lateral costal margin or the circumference at the umbilicus to the maximum circumference value over the buttocks), waist circumference, and sagittal diameter (an index of visceral obesity measured as the distance from the small of the back to the front of the body). Larger WHR, waist circumference, and sagittal diameter measurements indicate more abdominal obesity; a WHR > 1 is usually indicative of abdominal obesity. Unfortunately, much of the literature includes only one type of obesity or includes both types but does not analyze them independently. Rather, these two types of obesity are often used interchangeably. Accordingly, we discuss which types of obesity were considered in the literature we review here and how they were measured and analyzed. Hereafter, when describing sample characteristics, we use the terms “obese” and “generalized obesity” to refer to BMI > 30, and we use measure-specific terminology to indicate abdominal obesity (i.e. WHR, waist circumference, etc.).