The Activity of the Hypothalamic-Pituitary-Adrenal Axis and the Sympathetic Nervous System in Relation to Waist/Hip Circumference Ratio in Men


Department of Heart and Lung Diseases, Sahlgrenska University Hospital, SE-413 45 Göteborg, Sweden. E-mail:


Objective: To investigate possible differences, between generally and abdominally obese men, in activity and regulation of the hypothalamic-pituitary-adrenal (HPA) axis and the sympathetic nervous system.

Research Methods and Procedures: Fifty non-diabetic, middle-aged men were selected to obtain two groups with similar body mass index (BMI) but different waist/hip circumference ratio (WHR). Measurements were performed of the activity of the HPA axis and the sympathetic nervous system, as well as metabolic and endocrine variables.

Results: Men with a high WHR, in comparisons with men with a low WHR, had higher insulin, glucose, and triglyceride values in the basal state and higher glucose and insulin after an oral glucose tolerance test. Men with high WHR had elevated diurnal adrenocorticotropic hormone (ACTH) values but similar cortisol values, except lower cortisol values in the morning. Diurnal growth hormone concentrations showed reduced peak size. Stimulation of the HPA axis with corticotropin-releasing hormone (CRH) and laboratory stress showed no difference in ACTH values between groups, but cortisol values were lower in men with high WHR. In comparison with men with a low WHR, men with a high WHR had elevated pulse pressure and heart rate in the basal state and after challenges by CRH and laboratory stress. They also had increased urinary excretion of catecholamine metabolites.

Discussion: These results suggest a mild dysregulation of the HPA axis, occurring with elevated WHR independent of the BMI. The results also indicate a central activation of the sympathetic nervous system, such as in the early phases of hypertension, correlating with insulin resistance.


A prevailing hypothesis regarding the pathogenesis of essential (primary) hypertension is that, in early phases, central sympathetic activity is elevated. This is followed by an increased reactivity of hemodynamic functions such as heart rate, cardiac output, and pulse pressure, particularly after challenges of the regulatory system (1).

Insulin resistance is a common occurrence in essential hypertension (2). Insulin has been suggested to be involved in the development of hypertension. This might be due to facilitation of the central sympathetic nervous system (SNS) activation by insulin (3).

Obesity is a condition characterized by frequent occurrence of insulin resistance and hypertension (2), particularly so in the subgroup of central abdominal obesity (4). Consequently, abdominal obesity is of interest from the standpoint of examining the relationships between insulin resistance, hypertension, and the activity of the SNS. Furthermore, abdominal obesity is frequently associated with perturbations of the hypothalamic-pituitary-adrenal (HPA) axis (5). Because the HPA axis and the SNS are tightly coupled with each other at several levels (6), we examined the HPA axis and the SNS simultaneously in men with abdominal obesity. To control for the potential influence of obesity per se, the male subjects were selected to have widely different distribution of body fat but similar degree of obesity in terms of body mass index (BMI, kg/m2). Activities of the central SNS and of the HPA axis were measured in parallel under steady-state conditions and standardized stimulations. The objective of the study was thus to examine the function of the HPA axis and the central SNS in groups of men with different waist/hip circumference ratio (WHR).

Research Methods and Procedures

Fifty men were selected for the study after recruitment by advertisements. Selection criteria were age 45 to 60 years, absence of diabetes, and a BMI between 25 and 33. A further selection was performed to obtain two groups with different WHR with a division line of WHR = 1.0. Twenty-seven men had a WHR < 1.0, and 23 men had a WHR > 1.0. Four men in each of these subgroups were habitual smokers, and all men were apparently healthy and reported moderate alcohol consumption according to The Alcohol Use Disorders Identification Test (AUDIT).

Study Design

The men underwent an oral glucose tolerance test (OGTT), stimulation with corticotropin-releasing hormone (CRH) and an arithmetic stress test. Furthermore, diurnal concentrations of adrenocorticotropic hormone (ACTH), cortisol, and growth hormone (GH) were measured. All investigations, except 24-hour hormone profiles, were performed in the overnight-fasting state.

Body weight was measured to the nearest 0.1 kg with the men in underwear, and the height was measured to the nearest 0.5 cm. The waist circumference was measured horizontally midway between the lowest rib and iliac crest, and hip circumference as the widest circumference in the gluteal region. The WHR was then calculated. Sagittal abdominal diameter was determined as the distance between the examination table and the highest point of the abdomen (7). Blood pressures were measured after a 5-minute rest in the supine position with a mercury sphygmomanometer. The average of two measurements with a 1-minute interval was used. Glucose was determined by a commercially available enzymatic method, and serum lipids were measured as described previously (8). Insulin was determined by radioimmunoassay (Phadebas, Amersham Pharmacia Biotech, Uppsala, Sweden).

Profiles (24-hour) of circulating ACTH, cortisol, and GH were obtained from samples collected at 30-minute intervals during 26 hours. The subjects were admitted to the ward between 0900 and 1000 hours. Food was given ad libitum as breakfast (0800 hours), lunch (1200 hours), dinner (1700 hours), and an evening snack (2000 hours). A goal was to have minimal disturbance of sleep, and the men were usually sleeping during the night samplings. Blood samples for analyses of ACTH, cortisol, and GH were collected through an intravenous nonthrombogenic catheter inserted into an antecubital vein. The catheter was connected to a small portable withdrawal pump. Blood was withdrawn at a constant rate and collected in EDTA tubes.

OGTT was performed at 0900 hours with 100 g of glucose dissolved in water ingested within 5 minutes. Blood glucose and plasma insulin were measured via a patent intravenous catheter before and 30, 60, 90, and 120 minutes after glucose administration.

On another morning at 0900 hours, a pituitary stimulation with CRH was performed. An intravenous catheter was inserted, and then the subjects rested for 30 minutes in a recumbent position. At 0900 hours, 100 μg of CRH (Corticorelin [human] as trifluoroacetate, Ferring, Kiel, Germany) was administered intravenously, with the participants remaining comfortably in a recumbent position. Blood samples for ACTH and cortisol were collected 15 minutes before CRH was injected, immediately before and then 15, 30, 60, 90, and 120 minutes after the CRH injection.

An arithmetic stress test was performed as described previously (9). In short, an intravenous catheter was inserted and then the subjects rested for 30 minutes in a recumbent position. During the stress test, which involved forced calculation for 10 minutes, and thereafter, the men were sitting comfortably. Blood pressure and pulse rate were recorded 15 minutes before commencing the stress test, immediately before, and then 10, 20, 30, 40, and 60 minutes after start. Blood samples for ACTH, cortisol, adrenaline, and noradrenaline were drawn at the same points of time (except at 60 minutes).

A dexamethasone suppression test was performed using Salivette (Sarstedt, Rommelsdorf, Germany), which is a sampling device that consists of a small cotton swab inside a centrifugation tube, used to collect saliva (10). The participants were given five Salivettes and one tablet of dexamethasone (Decadron, MSD, Sweden) of 0.5 mg (11). They were asked to chew on the cotton swab during 60 seconds in the morning of each of four days. At 2200 hours on the final day, the dexamethasone tablet was taken. The following morning the salivary sampling was repeated. The decrease in salivary cortisol level after dexamethasone administration was calculated as the mean of the four noninhibited morning values minus the cortisol value the morning after dexamethasone intake.

Total serum testosterone was determined by a nonextraction method where testosterone bound to bovine serum albumin was used as the antigen (testosterone radioimmunoassay; ICN Biomedicals, Costa Mesa, CA). Insulin-like growth factor I (IGF-I) was measured with a non-extraction radioimmunoassay (Nichols, Institute Diagnostics, San Juan Capistrano, CA). ACTH was determined by an immunoassay method (Nichols Institute Diagnostics), cortisol by radioimmunoassay (Orion Diagnostica, Turku, Finland), and GH by a radioimmunochemical method (Amersham Pharmacia Biotech, hGH radioimmunoassay kit, Kabi-Pharmacia, Uppsala, Sweden). Adrenaline and noradrenaline were measured by high-pressure liquid chromatography, and methoxycatecholamines in urine were collected during 24 hours (12).

Statistical Methods

Comparisons between groups were analyzed with the Mann–Whitney U-test and repeated-measures ANOVA. In Table 2, the Spearman rank correlation coefficient p < 0.05 was considered significant. Data analyses were performed using StatView for Macintosh.

Table 2.  Correlations between pulse pressure during stress test and selected variables
 Pulse pressure during stress
  • Spearman rank correlation coefficient.

  • *

    p < 0.10,

  • p < 0.05,

  • p < 0.01,

  • §

    p < 0.001.

  • Abbreviations: Pulse pressure during stress = the individual mean difference between systolic and diastolic blood pressure at −15, 0, 10, 20, 30, 40 and 60 minutes after commencing stress test, Sum glucose OGTT = the individual sum of glucose values during oral glucose tolerance test, HDL = high density lipoproteins, Morning cortisol = the individual mean morning value from four measurements between 0845 and 0900 hours at different days.

BMI (kg/m2)0.253*
Sagittal abdominal diameter (cm)0.433
Fasting blood glucose (mM)0.318
Fasting plasma insulin (mU/liter)0.433
Sum glucose OGTT0.433§
Triglycerides (mM)0.279
Cholesterol (mM)0.176NS
HDL (mM)−0.419
Morning cortisol−0.330
Methoxycatecholamines (μmol/24-hour urine)0.255*
Methoxycatecholamines (mmol/mol creatinine)0.366


Table 1 shows the descriptive characteristics of the men in both subgroups of WHR. Age, body weight, and BMI were not different between the groups. The WHR did not show any overlap between groups. Sagittal abdominal diameter also differed markedly.

Table 1.  Basal data for the 50 men subdivided into two groups with WHR below and above 1.0
 WHR < 1.0 (n = 27)WHR > 1.0 (n = 23)p
  • Data are means (SD). Comparisons between groups were analyzed with the Mann–Whitney U-test.

  • p < 0.05,

  • p < 0.01,

  • §

    p < 0.001; NS, not significant.

Age (years)51.3 (4.6)53.7 (4.2)NS
Body weight (kg)92.5 (9.9)95.7 (10.4)NS
BMI (kg/m2)28.7 (2.0)29.6 (1.7)NS
WHR0.94 (0.03)1.06 (0.04)§
Sagittal abdominal diameter (cm)23.5 (2.0)26.0 (1.3)§
Systolic blood pressure (mm Hg)123.1 (14.7)132.8 (19.2)NS
Diastolic blood pressure (mm Hg)74.8 (9.7)77.2 (8.7)NS
Heart rate (beats/min)60.8 (11.8)66.9 (8.3)NS
Fasting blood glucose (mM)4.51 (0.33)4.95 (0.43)§
Fasting plasma insulin (mU/liter)9.8 (3.7)16.4 (8.5)
OGTT, sum of glucose (mmol/liter)28.7 (5.0)38.3 (7.4)§
OGTT, sum of insulin (mU/liter)243.4 (129.4)503.5 (265.5)§
Triglycerides (mM)1.52 (0.88)2.21 (1.19)
Cholesterol (mM)5.58 (1.11)6.00 (1.05)NS
High density lipoproteins (mM)1.27 (0.40)1.05 (0.21)
Testosterone (nM)19.6 (4.9)20.6 (6.0)NS
IGF-I (μg/liter)188.7 (42.4)190.6 (45.9)NS
Basal morning ACTH (ng/liter)29.4 (12.6)32.4 (22.3)NS
Basal morning cortisol (nM)318 (80)265 (58)
Cortisol (nmol/24-hour urine)266 (123)254 (91)NS
Cortisol (μmol/mol creatinine)17.7 (6.4)16.9 (5.7)NS
Methoxycatecholamines (μmol/24-hour urine)2.11 (0.48)3.13 (1.25)
Methoxycatecholamines (mmol/mol creatinine)0.135 (0.051)0.205 (0.069)

Blood glucose and insulin, basal and OGTT values, and triglycerides and methoxycatecholamines in 24-hour urine were significantly higher in the group of men with a WHR > 1.0 in comparison with the men with a WHR < 1.0, whereas high density lipoprotein and morning cortisol were lower. Resting blood pressures, heart rate, cholesterol, testosterone, IGF-I, basal morning ACTH, and cortisol excretion in 24-hour urine were not different.

Profiles (24-hour) of circulating ACTH, cortisol, and GH are shown in Figures 1 and 2 as means for the groups at each time point.

Figure 1.

(A) ACTH concentrations measured in samples collected over 24 hours in the groups of men with a WHR above or below 1.0. Mean values are shown. * = p < 0.10, † = p < 0.05, ‡ = p < 0.01 for the following times: 1400†, 1430‡, 1500*, 1600*, 2000†, 2030†, 2230*, 2300*, 0130*, 0430*, and 0500* hours. (B) Cortisol concentrations measured in samples collected over 24 hours in the groups of men with a WHR above or below 1.0. Mean values for the groups, respectively, are shown. P < 0.05 at 0200 hours.

Figure 2.

GH concentrations measured in samples collected over 24 hours in the groups of men with a WHR above or below 1.0. Mean values are shown. * = p < 0.10, † = p < 0.05, ‡ = p < 0.01 for the following times: 1100†, 1330‡, 1630†, 1700†, 2200*, 0100*, 0200*, and 0230† hours.

Figure 1A shows the ACTH values. On several occasions men with a WHR > 1.0 had higher ACTH values than the men with a WHR < 1.0.

The cortisol values (Figure 1B) showed no certain differences except at one point at 0200 hours with higher values in the WHR > 1.0 group (p < 0.05). However, the morning cortisol values, from four measurements between 0845 and 0900 hours at different days, before breakfast, showed lower values in the men with a WHR > 1.0 than men with a WHR < 1.0 (see Table 1; Basal morning cortisol).

Figure 2 shows the GH concentration over a day. Men with a WHR > 1.0 clearly showed lower GH levels on several occasions (significant at 1100, 1330, 1630, 1700, 0230; and p < 0.10 > 0.05 at 2200, 0100, and 0200). This seemed to be mainly due to smaller secretion peaks. In general, the secretion pattern was similar in both groups.

Measurements of blood pressures and pulse pressure (systolic minus diastolic blood pressures) before, during, and after CRH injections are depicted in Figure 3 A. Systolic blood pressure (borderline significance) and pulse pressure were significantly higher in men with a WHR > 1.0 in comparison with men with a WHR < 1.0. Diastolic blood pressures were not different between groups and the CRH test was followed by reduced diastolic blood pressures in both groups (p < 0.05).

Figure 3.

(A) Blood pressures before and during pituitary stimulation with CRH in the groups of men with a WHR above or below 1.0. Pulse pressure was calculated as the difference between systolic blood pressure (SBP) and diastolic blood pressure (DBP). (B) Heart rate before and during pituitary stimulation with CRH in the groups of men with a WHR above or below 1.0. (C) ACTH and cortisol values before and during pituitary stimulation with CRH in the groups of men with a WHR above or below 1.0. Mean values for the groups, respectively, and p values, claculated using ANOVA repeated measures, are shown.

Heart rate (Figure 3B) was consistently higher in men with a WHR > 1.0, with a peak after CRH injection.

ACTH values were not different, although cortisol was lower (p < 0.05) in the group of men with a WHR > 1.0 during pituitary stimulation with CRH (Figure 3C).

Blood pressures before, during, and after the arithmetic stress test are found in Figure 4 A. The test was followed by elevated systolic and diastolic blood pressures in both groups (p < 0.05). Men with WHR > 1.0 had higher systolic and pulse pressures before, during, and after the stress test.

Figure 4.

(A) Blood pressures before and during stress test in the groups of men with a WHR above or below 1.0. Pulse pressure was calculated as the difference between systolic blood pressure (SBP) and diastolic blood pressure (DBP). (B) Heart rate before and during stress test in the groups of men with a WHR above or below 1.0. Mean values and ANOVA repeated measures are shown.

Heart rate recordings are seen in Figure 4B and were consistently higher in the group of men with a WHR > 1.0 before, during, and after the stress.

The difference in ACTH and cortisol values during the stress test did not reach significance in comparisons between the groups (not shown).

Dexamethasone suppression showed no difference (not shown).

Serum adrenaline and noradrenaline before, during, and after the arithmetic stress test were not different between groups. The increase in noradrenaline during stress was significantly smaller in the group of men with WHR > 1.0 (not shown).

Table 2 shows correlations between pulse pressure during the stress test and other variables, significant for several anthropometric and metabolic values as well as excretion of catecholamine metabolites. High density lipoprotein and morning cortisol values showed negative correlations. Heart rate during stress exhibited similar correlations to the selected variables as pulse pressure did (not shown).


To study the separate effect of centralization of body fat, men were recruited for a study in which two groups were selected with non-overlapping values of WHR. Furthermore, to minimize the effects of total body fat mass, men in both groups were matched for BMI as equally as possible.

As reported repeatedly previously (for review, see Ref. (5)), men with a high WHR showed significantly higher glucose and insulin during fasting and after an oral glucose tolerance test as well as higher triglycerides and lower high density lipoprotein cholesterol.

The HPA Axis

Diurnal curves of ACTH concentration showed frequently elevated values in the men with higher WHR. Cortisols were, however, not different except for a lower morning cortisol value in men with higher WHR, as reported previously (11, 13, 14). The ACTH/cortisol ratio curves consequently showed generally higher values in men with elevated WHR. This has been observed previously in Cushing's disease (15), in severe obesity (16), and in men with abdominal obesity (17). Explanations for this have included insensitive adrenals or a decreased 21-hydroxylase activity, which would direct cortisol precursors mainly into synthesis of adrenal androgens (17). Other possibilities might be increased activity of 11β-hydroxysteroid dehydrogenase, which converts cortisol to inactive cortisone (18).

Previous studies have demonstrated an increased ACTH and cortisol response to CRH injection (19) and combined CRH/arginine-vasopressin administration (20) in abdominally obese women. This is less pronounced in obese men (21).

Diurnal GH secretion was clearly lower in the men with an elevated WHR, confirming previous results of both IGF-I measurements (22) and diurnal GH determinations (23).

Taken together, the results indicate that an elevated WHR, without the influence of total body fat or obesity, is associated with a mild malfunction of the HPA axis, visible as elevated diurnal ACTH values, low morning cortisol values, and, probably secondarily, inhibition of GH secretion (6). Suppression of cortisol secretion by low dose dexamethasone showed no differences between the groups with high or low WHR, suggesting that the HPA axis abnormalities found were subtle and apparently did not affect the feedback regulatory function, i.e., the glucocorticoid receptors (GR) (11). Tentatively, the differences in cortisol concentrations seen among these relatively healthy men might be due to a desensitization of the adrenals for ACTH in the high WHR group, and the GR down-regulation might be a phenomenon that comes later in the development of a complete metabolic syndrome. Based on other studies (14, 24) we believe that morning cortisol decrease is an early sign of a perturbed function of the HPA axis. This may or may not be associated with elevated urinary cortisol excretion (13, 14) depending on the “stage” in the HPA perturbations and might also involve gender differences.

In comparison with previous work (13, 14, 25) there are both similarities and differences in results. Previous reports (14, 25) involved a large group of men, recruited at random. In these men, salivary cortisol was analyzed during an ordinary working day. The results obtained in these studies show that men with elevated WHR have an abnormality of the HPA axis of varying severity (14, 25, 26). In comparison with these men, the men examined in the present study apparently show milder HPA abnormalities.

There are, however, several potentially important differences between these studies. BMI was excluded from influence on the examined variables by matching them in this study, which was not done in previous work. It cannot be excluded that there is a connection between elevated HPA axis activity and obesity. Therefore, it is possible that exclusion of the influence of obesity in the present investigation diminished the display of HPA axis involvement. In the population studies of the same problem, HPA axis perturbations are usually associated with both measurements of body fat centralization (WHR and sagittal abdominal diameter) and obesity (BMI) (14, 25, 27).

Another difference is the recruitment. In the present investigation the men responded to an advertisement, whereas in the previous studies an attempt was made to examine an entire population using a randomization procedure. In the present study this probably caused a selection of men in terms of, for example, socioeconomic and psychosocial factors, alcohol intake, and with less tendencies to depressive and anxiety traits, which all have been shown previously to affect HPA axis function (for review, see Ref. (5)). Such selection mechanisms may have been followed by less affected HPA axis regulation in the men examined here.

The Sympathetic Nervous System

Systolic and diastolic blood pressures were not different, but pulse pressure and heart rate were significantly higher in the men with WHR > 1.0. Such characteristics have been described in the earliest stages of development of hypertension, and are probably due to elevated cardiac output, which are signs of activation of the central SNS (1).

The interpretation of an elevated central activation of the SNS in men with higher WHR is consistent with higher excretion of methoxycatecholamines, i.e., metabolites of catecholamines. This finding then suggests an elevated total secretion of catecholamines.

The results of measurements of adrenaline and noradrenaline in blood did not correlate with urinary output of catecholamine metabolites (not shown). These observations suggest that measurements of circulating catecholamines reflect other phenomena besides circulatory events and urinary output of catecholamine metabolites. Blood for the measurements of catecholamines was drawn from a peripheral vein, and local contributions, either to production or to breakdown of catecholamines, may have yielded results not representative for the whole body and, particularly, not representative for hemodynamic events. The latter are regulated by the activation of specific branches of the SNS, which are not necessarily followed by activation of other parts of this nervous system (1). Therefore, the events are not followed by elevated circulating catecholamines in peripheral venous blood.

Taken together, we have interpreted these results to mean that the central SNS shows an elevated activity, at least in branches regulating central hemodynamics, in men with elevated WHR.

Relationship of Central SNS Activity to Metabolism

The elevated SNS activity and the metabolic abnormalities seem to be connected. This is suggested by the correlation analyses in Table 2, which show that pulse pressure as a surrogate measurement of the activity of the SNS correlated with BMI, WHR and sagittal diameter, glucose, insulin, and an unfavorable lipid profile. These results indicate that the SNS activation probably influences metabolic variables. Lipid mobilization is highly dependent on catecholamines in humans (28). The products, circulating free fatty acids, interfere with peripheral metabolism of both lipids and carbohydrate (29, 30), but were not measured in this study. Previous studies have shown that free fatty acids are elevated in abdominal obesity (31).

Morning cortisol values correlated negatively with measurements of pulse pressure during stress (Table 2). This is in agreement with previous population-based studies (32), which have shown that an abnormal regulation of the HPA axis, with low morning cortisol values as a characteristic feature, is strongly associated with elevated blood pressure and heart rate in men with abdominal obesity (32). These observations suggest a combined abnormality of regulation of the HPA axis and the SNS in abdominal obesity. Partly contrary to these findings is a recently published study in which BMI and WHR were inversely correlated with both sympathetic and parasympathetic tone assessed by use of heart rate variability (33).

Insulin has been described to facilitate activation of the central SNS (3), helping to explain the frequent statistical associations between insulin resistance and hypertension (2, 31). A central interaction between insulin and the SNS has, however, been questioned, and several other possibilities remain in the complex statistical relationships between hyperinsulinemia and hypertension (34, 35). Previous studies have indicated that correlations between blood pressure and insulin might be due to a parallel activation of the SNS, resulting in elevated blood pressure, and the HPA axis, resulting in insulin resistance and elevated insulin secretion (32). The results presented here show a correlation between insulin and pulse pressure (Table 2). This observation seems to be compatible with both concomitant central regulatory abnormalities (32) and insulin facilitation of the SNS (3).

In summary, the results of this study indicate that men with localization of a disproportionately large fraction of body fat to central fat depots are characterized by an elevated activity of the SNS, responsible for the regulation of central hemodynamics. Findings of multiple metabolic perturbations are well known from previous observations, and this study suggests that activation of the SNS influences these abnormalities. Men with abdominal obesity have previously been shown to exhibit perturbations of the central regulation of the HPA axis, which also seems to be the case in the men studied here. The centers in the hypothalamus, which regulate the sympathetic nervous activity and the HPA axis, are closely interconnected (6). It is therefore hypothesized that the signs of combined elevated and perturbed activity along the HPA axis and the central SNS are due to a hypothalamic arousal, responsible for hemodynamic and metabolic abnormalities associated with abdominal obesity.

The HPA and the SNS are both involved in responses to stressful events (1, 6, 36, 37). It is therefore of interest that subjects characterized by centralization of body fat are exposed to several psychosocial and socioeconomic handicaps, which would be expected to enhance stress reactions and tend to expose such subjects to chronic stress (5). This in turn might be the basis for hypothalamic arousal reactions.


We thank Marie-Louise Norberg, Carola Gustafsson, Raija Saikkonen, and Birgitta Odén for their dedicated clinical work. This study was supported by grants from the John D. and Catherine T. MacArthur Foundation Research Network on Socioeconomic Status and Health and the Swedish Medical Research Council (project no. B96-19X-00251-34B).