CHAPTER II

LITERATURE REVIEW

In this chapter, the processes of fat deposition and mobilization will be reviewed

and the health risks associated with abdominal obesity will be revisited. Existing animal

and human research on Bjorntorp’s stress- induced abdominal fat deposition theory will

be reviewed along with the physiological impact of elevated cortisol levels, the key

hormonal aberration linked to this hypothesis. The rationale for selecting Black females

as study participants will also be presented. This chapter ends with a summary of the

available techniques for assessing central adiposity and measuring physiological cortisol

levels.

Abdominal Obesity

Prevalence and Distribution of Obesity

Data drawn from the third National Health and Nutrition Examination Survey

(NHANES III), conducted from 1988-1991, indicates that 33% of the United States

population is overweight (Kuczmarski, Flegal, Campbell, & Johnson, 1994). Relative to

similar data collected in the NHANES II survey conducted from 1976 to 1980, this

represents an 8% increase in the prevalence of overweight or obese adults aged 20-74

years of age. The prevalence of obesity is disproportionately high in minority groups and

is particularly noticeable among Black women, who exhibit a prevalence rate of 50%

(DiPietro, 1995). Moreover, Black women are two to four times more likely to

experience obesity-related health conditions such as NIDDM and cardiovascular disease

than White women (National Heart, Lung, and Blood Institute Growth and Health Study

Research Group. 1992; Otten, Teuch, Williamson, & Mark, 1984). Viewed in concert,

these findings highlight the pervasive nature of obesity and emphasize the need to target

research and prevention efforts towards minority groups such as Black females.

The first step in the process of understanding obesity is to examine how the

human body stores and mobilizes body fat. To address these issues, the process of

adipose tissue deposition and mobilization will be covered in the following section.

Fat Deposition and Mobilization

The primary enzyme involved in adipose accumulation is lipoprotein lipase

(LPL), which is the rate- limiting enzyme for the uptake of fat into cells. Lipid

mobilization occurs primarily through lipolysis, wherein lipid is released as free fatty

acids and glycerol in the basal state in response to noradrenaline stimulation (Rodin,

1992). The enzyme responsible for stimulating the breakdown of fat tissue is hormone-

sensitive lipase (HSL). The balance between accumulation and mobilization of adipose

tissue is dependent upon a series of hormonal interactions. For instance, at the level of

the adipocyte, insulin activates LPL and inhibits HSL to promote fat accumulation. Fat

mobilization, on the other hand, is supported by testosterone, the estrogens, and

catecholamines. Varying sex hormone levels in males and females result in distinct sex-

specific fat patterning.

The role of sex hormones in fat deposition and mobilization is well illustrated by

differences in fat patterning between males who tend to carry fat in the abdominal region

(android obesity) and females, who more typically accumulate fat in the gluteo-femoral

region (gynoid obesity). Progesterone regulates the accumulation of subcutaneous fat in

the femoral region by activating LPL activity and lowering the lipolytic response

(Bjorntorp, 1991a; Rodin, 1992). These actions become readily apparent as females

move through menopause. Post-menopausal women who are not on hormone

replacement therapy tend to deposit and retain fat in the abdominal region. When the

ovarian production of sex hormones has considerably decreased or ceased, the increased

LPL activity in the femoral region disappears and there is no longer any regional

difference in LPL activity (Rodin, 1992). When post-menopausal women are placed on

combined estrogen and progesterone hormone replacement therapy, femoral LPL activity

increases and the body’s preferential fat depot once again switches from the abdomen to

the gluteo-femoral region (Rebuffe-Scrive, Eldh, Hafstrom, & Bjorntorp, 1986). Xu and

Bjorntorp (1990) have also demonstrated that in addition to increasing femoral adipose

tissue accumulation, progesterone may influence the development of available fat cells by

increasing adipose precursor cell differentiation.

In men, testosterone acts to inhibit fat accumulation by preferentially inhibiting

LPL activity in the abdominal region (Rebuffe-Scrive, Walsh, McEwen, & Rodin, 1992).

Therefore, men with excessive intra-abdominal fat stores tend to have inadequate

concentrations of testosterone (Seidell, Bjorntorp, Sjostrom, Kvist, & Sannerstedt, 1990).

Testosterone treatments help to mobilize the high levels of visceral fat found in

hypogonadal men with a specificity for the intra-abdominal adipose tissue region (Marin,

Holmang, Jonsson, et al., 1992; Rebuffe-Scrive, Marin, & Bjorntorp 1991). Acting to

further decrease fat stores, testosterone enhances lipolysis by increasing the expression of

lipolytic-adrenergic receptors (Bjorntorp, 1992b; Marin & Bjorntorp, 1993). Indirect

evidence, drawn from aging men who generally have more visceral fat mass and lower

levels circulating testosterone levels than younger men (Bjorntorp, 1992b), suggests that

the visceral fat depot might be specifically sensitive to androgens.

Curiously, testosterone plays a varied role in female obesity. Women with upper

body obesity have elevated levels of androgens and free testosterone (Evans, Hoffman,

Kalkhoff, & Kissebah, 1983). Additionally, like men, android-obese women have low

femoral LPL activity and similar lipolytic responses to noradrenaline in both the

abdominal and gluteo-femoral regions (Seidell et al., 1990). The role of elevated

testosterone in abdominally-obese females is not yet known, although it may act to either

inhibit femoral LPL or abdominal lipolysis (Seidell et al., 1990).

The primary lipid-accumulating hormone is insulin. Insulin enhances lipogenesis

by stimulating the synthesis and release of LPL and stimulating the conversion of glucose

to free fatty acids (FFA) in adipose cells (Marks, Marks, & Smith, 1993). Insulin also

decreases the mobilization of FFA from adipose tissue by inhibiting HSL, and thereby

lowering circulating FFA concentrations (Murray, Granner, Moyes, & Rodwell, 1993).

Adipose tissue is more sensitive to insulin than other tissues in the body and, as such, is a

major site of insulin action.

Another hormone, cortisol, has effects on both lipid mobilization and

accumulation. Cortisol has a permissive effect on lipid mobilization which is stimulated

by catecholamines (Bjorntorp, 1991). In addition, cortisol inhibits the antilipolytic effect

of insulin on adipocytes. This effect may be more prominent in visceral adipose due to

the high density of glucocorticoid receptors found in this tissue (Bjorntorp, 1991).

Although the mechanism is not clear, the net result of an elevated cortisol level in the

presence of hyperinsulinemia is, surprisingly, an accumulation of visceral adipose tissue

in the truncal region (Bjorntorp, 1997). This phenomenon is best illustrated by the

metabolic disorder known as Cushing’s syndrome, which is characterized by

hypercortisolemia caused by an adrenocortical tumor that increases the release of ACTH

(Mark et al., 1993). A hallmark of this syndrome is a marked accumulation of visceral

adipose tissue. Upon treatment of the hypercortisolemia in Cushing’s syndrome, there is

a dramatic decrease in the level of visceral adipose tissue (VAT) (Rebuffe-Scrive,

Krotkiewski, Elfversson, & Bjorntorp, 1988). Although it is not known why

hypercortisolemia produces this effect, it is speculated that by interacting with a

glucocorticoid receptor-cortisol complex on adipose tissue cells, elevated cortisol levels

enhance LPL activity, which is the key regulator of lipid uptake (Marin & Bjorntorp,

1993). High concentrations of cortisol also appear to decrease the net capacity of

lipolysis to release FFA which, in turn, decreases lipid mobilization (Rebuffe-Scrive et

al., 1991). As discussed in a later section, elevated cortisol concentrations, combined

with depressed levels of progesterone in females and testosterone in males, result in an

even riper environment for the development of truncal VAT.

Hormonal balance is essential to maintaining the human body in a non-diseased

state. In addition to the accumulation of adipose tissue, there are several other health

consequences which stem from an imbalance of hormonal activity. Several of these

health consequences will be discussed in the following section.

Link Between Abdominal Obesity and Health Outcomes

The question of why the accumulation of VAT is associated with so many

detrimental health outcomes is an important one to address. There are several

physiological characteristics of abdominal adipose tissue which lead to its potent health

consequences, especially in regard to hyperinsulinemia, insulin resistance, and

dyslipidemia.

FFA concentrations are elevated in obesity and notably so in abdominal obesity

(Bjorntorp, 1988). This elevation in FFA is attributed to the fact that VAT is very

lipolytically active. The lipolytic sensitivity of abdominal adipose tissue to

catecholamines is higher than that of the gluteo-femoral region, while at the same time,

the abdominal depot is less sensitive to the inhibitory effect of insulin (Bjorntorp, 1997).

Furthermore, insulin receptors have been shown to be less abundant in visceral fat

compared to subcutaneous adipose tissue (Bolinder, Engfeldt, Ostman, & Haner, 1983).

For these reasons, upper body fat depots would be expected to release more FFA into the

circulation.

An initial consequence of this elevation in adipose tissue mobilization is that

venous drainage from the abdominal region exposes the liver to high FFAs and glycerol

levels (Carey, 1998). In turn, this results in increased hepatic glucose and triglyceride

production and decreased insulin clearance by the liver (Carey, 1998). Direct

measurements of hepatic insulin uptake in perfused liver of obese rats confirm a

decreased insulin uptake (Stromblad & Bjorntorp, 1986). These same researchers also

demonstrated a negative relationship between insulin clearance and triglyceride content

of the liver. In humans, Peiris, Meuller, Struve, Smith, and Kissebah (1987) found that

insulin clearance was correlated negatively with WHR. In addition, insulin itself can

decrease hepatic insulin uptake, most likely as a consequence of down-regulation of

insulin receptors (Bjorntorp, 1988). Along with decreased insulin clearance, increased

secretion of insulin by the pancreas may contribute to the hyperinsulinemia state of

abdominally-obese individuals (Bjorntorp, 1988). Moreover, the WHR has been shown

to be related positively to insulin secretion in obese women (Kissebah & Peiris, 1989).

The hyperinsulinemic state so common to abdominal obesity, therefore, appears to derive

both from a diminished hepatic clearance of insulin and an elevated insulin release by the

pancreas.

Due to the frequent occurrence of insulin resistance associated with the

accumulation of intra-abdominal fat, viscerally-obese individuals are also at increased

risk of developing NIDDM. Computerized tomography (CT) scans of men and women of

varying body weight have revealed a positive relationship between the volume of VAT

and plasma insulin and glucose levels in response to an oral glucose challenge (Despres

& Lamarche, 1993). In addition, the relationship between intra-abdominal fat stores and

insulin resistance has been shown to be independent of total fat mass (Lamarche, 1998).

While the precise mechanism(s) underlying this insulin resistance have not been

elucidated, it has been hypothesized that excess circulating FFA may induce insulin

resistance (Randle, Garland, Hales, & Newsholme, 1963). Elevated levels of cortisol,

which are characteristic of abdominal obesity, may also promote insulin resistance

(Bjorntorp, 1997). Additionally, when insulin levels are high, as they are in obesity,

there is a down-regulation in insulin receptor number. This decrease in receptor number

decreases the sensitivity of adipose and muscle tissue to insulin and can lead to insulin

resistance (Mark et al., 1993). Interestingly, high levels of testosterone in females and

low levels of testosterone in men, both companions of abdominal obesity, have been

linked to insulin resistance (Bjorntorp, 1997). In sum, the evidence for preventing the

development of abdominal obesity in an attempt to curtail disease development is strong.

As noted earlier, the excess exposure of the liver to FFA results in an increased

production of triglycerides (TG). The hepatic production of triglycerides and very low-

density lipoprotein (VLDL) and small dense low-density lipoproteins (LDL) is substrate-

dependent (Lamarche, 1998). Hence, an excess volume of abdominal adipose tissue

would support the increased production of both VLDL and LDL. The associated

hyperinsulinemia also contributes to the overproduction of these lipoproteins (Kissebah

& Peiris, 1989). Concurrent with these disturbances in lipoprotein production is the

presence of an inverse relationship between plasma high density lipoprotein (HDL) and

CT scans of VAT (Lamarche, 1998). In addition, WHR correlates positively with plasma

TG and negatively with HDL levels (Kissebah & Peiris, 1989). Kissebah and Peiris

(1989) have speculated that combined hyperinsulinemia and insulin resistance may result

in the decreased activity of lipoprotein lipase (LPL) in adipose tissue and an increase in

the activity of hepatic lipase. This would result in a reduction in the conversion of HDL3

to HDL2 and total HDL cholesterol. This hypothesis is supported by the inverse

relationship between HDL concentrations and level of hyperinsulinemia (Kissebah &

Peiris, 1989). Together, these hormonal alterations form a dyslipidemic profile primed

for the development of cardiovascular disease and NIDDM (Assmann & Schulte, 1992).

These hormonal aberrations not only increase the risk of disease, but also contribute to

the accumulation of abdominal VAT. This developmental pathway will be discussed in

the following section.

Hormonal Aberrations and the Accumulation of VAT

There are a number of hormonal deviations associated with abdominal obesity

that can influence fat accumulation and mobilization. Excess fat accumulation,

specifically in the abdominal region, is associated with hypogonadism in men,

hyperandrogenism and low progesterone in women, and an increased sensitivity of the

hypothalamo-adrenal axis in both men and women (Bjorntorp, 1991a; 1992b; Jern,

Bergbrant, Bjorntorp, & Hansson, 1992).

As a consequence of an elevated sensitivity of the hypothalamo-adrenal axis,

individuals with abdominal obesity are known to have a relative hypercortisolism

compared to leaner counterparts (Bjorntorp, 1992a). Cushing’s disease, which is

characterized by excess cortisol secretion and an accumulation of fat in the central

abdominal region, serves as a strong example of how alterations in cortisol levels

influence body fat distribution (Bujalska, Kumar, & Stewart, 1997; Rebuffe-Scrive et al.,

1988). In particular, the intra-abdominal fat region has a high density of glucocorticoid-

cortisol receptors, which may help to explain why excessive levels of cortisol lead

specifically to VAT accumulation (Bjorntorp, 1996; Rodin, 1992). Among females who

typically accumulate fat in the gluteo-femoral region, progesterone competes with

cortisol for glucocorticoid receptors, therefore acting as a protective agent against

visceral fat accumulation (Bjorntorp, 1992a). In men, the actions of testosterone

counteract those of cortisol in minimizing abdominal fat deposition. Thus, the sex steroid

hormones act to buffer the net influence of cortisol and help explain why hypogonadism

in males and the absence or low levels of progesterone in females favors central adipose

deposition.

While it is not clear how cortisol influences adipose accumulation, the effect of

excess cortisol on adipose accumulation, especially when occurring in conjunction with

sex steroid derangements, is known. Current research on this topic is focused on possible

origins of these hormonal alterations. One area receiving attention is the study of

environmental events known to trigger the release of cortisol. Conclusions drawn from

epidemiological and cross-sectional research suggest that certain forms of stress may be

associated with an abdominal distribution of body fat. This hypothesis will be

highlighted in the next section of this review.

Perceived Stress and Abdominal Obesity

The Defeat Reaction and the Accompanying Hormonal Response

It has been speculated that factors which lead to increased cortisol production may

provide a pathogenic environment for the development of abdominal obesity (Bjorntorp,

1991b; Henry & Grim, 1990). The first researchers to posit a link between environmental

factors and the regulation of body fat deposition were Bjorntorp (1988) and Rebuffe-

Scrive (1988). Based on the knowledge that intra-abdominal tissue contains a high

density of glucocorticoid receptors, these researchers hypothesized that repeated arousal

of the defeat reaction would cause endocrine abnormalities and ultimately lead to the

accumulation of centralized body fat.

The defeat reaction occurs in response to the perception or threat of a loss of

control (Henry & Grim, 1990). The opportunity to exert control or the extent to which an

individual perceives events as lying within his or her sphere of influence is recognized as

a major determinant of the perceived stressfulness of person-environment interactions

(Frankenhaeuser, 1981). When an individual perceives that events and outcomes are

independent of his or her actions (external locus of control), a state of helplessness or

“learned helplessness”?may develop (Maier & Seligman, 1976). Conversely, an

individual who can regulate his or her environment may be able to maintain physiological

and psychological activation at optimal levels over a wide range of environmental

conditions (Frankenhaeuser, 1981).

Physiologically, conditions characterized by perceived unpredictability,

uncertainty, and a lack of control trigger the hypothalamic-adrenal axis and are

accompanied by a pronounced increase in the secretion of cortisol (Frankenhaeuser,

1981; Henry & Grim, 1990). Hormonal secretions from the hypothalamus, in turn,

regulate the release of hormones from the pituitary gland. Stimuli perceived as a threat

trigger the release of corticotropin-releasing hormone (CRH) from the hypothalamus

(Asterita, 1985). This hormone travels through the hypothalamic-hypophyseal portal

vessels that lead from the hypothalamus to the anterior pituitary (Asterita, 1985). CRH

then stimulates the secretion of adrenocorticotropin (ACTH) from the anterior pituitary,

which subsequently causes the rapid release of cortisol from the adrenal gland. When

blood cortisol levels are high, they normally exert a negative feedback on the

hypothalamus to lower CRH production (Asterita, 1985). Under conditions of intense

perceived stress, this direct- feedback loop is suppressed. Specifically, under intense

stress, high cortisol levels will not inhibit the further release of CRH and ACTH. This

phenomenon has been demonstrated in a number of experimental situations involving

research animals. Shively and Kaplan (1984), for instance, continuously scrambled the

male members of male and female monkey groups to prevent familiarity and the

development of stable relationships amongst the animals. The subordinate animals

developed larger and heavier adrenal glands, indicating greater hormone release. Holst

(1986), working with pairs of male tree shrews, an intensely territorial species,

demonstrated that adrenal weights and cortisol levels were highest in the defeated and

submissive animal in each pair. Additionally, Henry and Stephens (1988) documented

that under conditions of persistent social disorder, the least successful mice in social

groups developed adrenal hypertrophy.

In addition to increasing cortisol production, the defense reaction also influences

the secretion of sex steroid hormones in both males and females (Bjorntorp, 1992a;

Henry & Stephens, 1977). It is hypothesized that the disturbance in the production of sex

steroid hormones is secondary to the increase of CRH, which inhibits the release of

gonadotropin-releasing hormones, and leads to low concentrations of sex steroid

hormones (Henry & Grim, 1990; Marin & Bjorntorp, 1993; Olster & Ferin, 1987). In

sum, the endocrine profile of an individual exposed to chronic situations of perceived

stressful and uncontrollable situations mirrors that of an abdominally-obese individual.

Research Support for the Stress-Abdominal Obesity Hypothesis Drawn from

Human Studies

As an initial step in examining the question of stress- induced central body fat

distribution, data from population studies of 1400 women and 1000 men in Gothenburg,

Sweden were re-analyzed to target relationships between the WHR and a number of

environmental variables (Lapidus et al., 1989; Larsson et al., 1989). The reanalyzed data

indicated that higher WHR values were associated with low social class, poor education,

and low-paying manual labor. Additionally, WHR was related to what might be viewed

as symptoms of stress, including psychiatric and psychosomatic disease (e.g. peptic

ulcers), higher use of social welfare facilities, a greater number and duration of leaves

from work, sleeplessness, consumption of drugs for the treatment of anxiety and

depression, alcohol consumption, and cigarette smoking (Lapidus et al., 1989; Larsson et

al., 1989). Other researchers have also demonstrated that an elevated WHR is associated

with low concentrations of sex steroid hormones in both men and women (Seidell et al.,

1990; Hartz, Rupley, & Rimm, 1984).

Based on these preliminary findings, Bjorntorp (1991b) suggested that individuals

with a low socioeconomic background, poor education, and a job associated with low

levels of control might experience a submissive stress reaction with a response along the

HPA axis. In other words, Bjorntorp (1991b) proposed that the WHR might be a somatic

indicator of these psychosocial phenomena and their sequella.

The relationship between perceived stress and WHR has also been examined in

Type I and Type II diabetics. Lloyd, Wing, and Orchard (1996) demonstrated that

persons with Type I diabetes who perceived higher levels of stress also had higher WHR

values. As an extension of this research, Bell, Summerson, Spangler, and Konen (1998)

found that women with Type II diabetes who have higher levels of perceived stress also

have an elevated WHR. Neither of the above studies, however, evaluated the role of

cortisol as a possible mediator of the relationship between the WHR and levels of

perceived stress.

Recent methodological advances have provided a stepping stone from which

human research in this area has progressed. As noted in Chapter One, Rosmond,

Dallman, and Bjorntorp (1998) documented a link between cortisol, perceived stress, and

anthropometric variables in 284 middle-aged men. By repeatedly taking salivary

samples, these researchers were able to assess cortisol concentrations and perceived stress

levels throughout a standard workday and quantify variability in diurnal cortisol curves.

Decreased variability in morning and evening cortisol concentration values is a

consequence of frequent stimulation of the HPA axis (Dallman, 1993) and chronic or

frequent stressful stimuli have the potential of overstimulating the HPA axis and

flattening the cortisol diurnal curve. Rosmond et al. (1998) found that the relationships

between stress-related cortisol release and the WHR and sagittal diameter were stronger

in men exhibiting decreased diurnal cortisol variation. While research on this topic is

limited and true causal inferences cannot yet be made, this latest investigation more

firmly establishes the validity of Bjorntorp’s hypothesis (1991) and provides new

methodological techniques which may be applied to future studies in this area.

Due to the limited work conducted with human participants the role that

psychosocial stress may play in human abdominal fat deposition is a novel area for

research. The research of Lapidus et al. (1989) and Rosmond et al. (1998) provide a

basis from which new research can be initiated. Both sets of investigators targeted a

number of variables found to be related to the WHR and also provided techniques to

measure these variables. Incorporation of these variables and methodological advances

into future research will provide valuable replication of these premiere studies.

Research Support for the Stress-Abdominal Obesity Hypothesis Drawn from

Animal Studies

Due to the ease of random assignment, manipulation of the experimental

conditions, and the ability to control environmental and genetic influences, animal

models have produced valuable and definitive advances in the research literature.

Decreased sex hormones and increased cortisol levels are well-known responses to

uncontrollable conditions which induce submission and defeat in animals (Rebuffe-

Scrive, Walsh, McEwen, & Rodin, 1992). Monkeys stressed to submission show

increased adrenal weights, decreased sex steroid levels, and centralized adipose tissue

accumulation (Shively & Clarkson, 1988). Due to the high density of glucocorticoid

receptors in internal fat stores, Rebuffe-Scrive and colleagues (1992) postulated that

uncontrollable stress would lead to increased fat deposition in the mesenteric region of

male rats. In rats, the mesenteric region includes the deep internal fat tissue that

surrounds the liver and is the counterpart of the human intra-abdominal fat store.

Stressors used in this experiment included rotation from speeds of zero to 100 revolutions

per minute and restraint in a plexiglass container. To decrease habituation to the

experimental protocol, the regimen was alternated randomly between the two stress

conditions, in addition to being increased in frequency, severity, and duration. The

protocol lasted for 28 days, during which time control animals were pair fed to account

for the influence of stress on food intake and body weight. Results demonstrated that the

mesenteric fat depot responded differently to the chronic stress than other fat depots in

the male rats. Fat cell weight was significantly higher only in the mesenteric region and

LPL activity was doubled in the stressed rats in comparison to the control animals. The

stressed rats also had significantly higher cortisol levels and statistically significant lower

levels of free testosterone than the control animals. Additionally, cortisol levels only

showed statistically significant correlations with the size of the mesenteric fat pad. To

determine if the effect of the stress was due only to glucocorticoids or if other

mechanisms were involved, Rebuffe-Scrive et al. (1992) conducted a follow-up study,

wherein rats were given exogenous administrations of corticosterone through either pellet

implantation or the water supply. Under both conditions, fat cell size and weight and

LPL activity increased preferentially in the mesenteric region (Rebuffe-Scrive et al.,

1992).

Significance of the Study Population

The current investigation will sample from the population of Black females. As

noted previously, the prevalence of obesity is elevated among this segment of the

population (DiPietro, 1995). This finding alone provides adequate reason to study the

origin of obesity in this sample. Further justification for sampling from this ethnic group

is derived from the understanding that Black females may also perceive higher levels of

stress than other segments of society.

In a recent study, a sample of Black females specified during an interview that

the dual burdens of racism and sexism were strong, stress-provoking agents (Walcott-

McQuigg, 1998). As detailed by these women, the lack of positive media images and

isolation associated with being the only Black female in a department or in management

at work were some of the difficulties associated with being a black female (Walcott-

McQuigg, 1998). Warren (1997) investigated this “double jeopardy” minority status and

noted that Black women reported more depressive symptoms than white women.

Hauenstein (1996) also suggested that Black women might be at greater risk to

experience depression and feel devalued in American society.

Despite their elevated occurrence of obesity, it has been documented that fewer

Black females perceive themselves as being overweight compared to White women

(Dawson, 1988). It also appears that in the Black culture, less significance is placed on

being slim as an indicator of attractiveness (Allan, 1993). Moreover, Black women

generally do not demonstrate the same concerns about diet and weight management so

often voiced by White females (Kumanyika, 1987).

The previous discussion highlights the importance of the accumulation of VAT in

Black females. This minority group is characterized not only by an elevated prevalence

of adipose accumulation, but also by an elevated potential to experience stressful life

conditions. Knowledge of the health risks associated with excess adipose tissue must be

shared equally across the population, especially because the health concerns associated

with obesity may be underplayed in the Black culture. Explanations of why Black

females may be prone to accumulate adipose tissue and how this accumulation of fat may

be altered by lifestyle changes are needed to help curb the prevalence of disease.

Measurement Issues Related to Assessing Abdominal Obesity and Cortisol

Concentrations

Assessing Fat Distribution and VAT

The most powerful means to determine intra-abdominal fat include computerized

tomography (CT) and magnetic resonance imaging (MRI) (Lamarche, 1998). These

direct measures of visceral fat are not likely to be used as routine clinical tools for risk

management, however, because they are expensive and require the use of specialized

equipment and trained technicians. Additionally, the use of CT involves exposing

patients to radiation (Lamarche, 1998). As a result of these concerns, several

anthropometric substitutes have been developed as indicators of body fat distribution and

used in the prediction of VAT volume. The traditional measure of body fat distribution is

the WHR, wherein waist circumference is divided by hip circumference to produce a

single reference value. Ratios greater than .86 in women and .95 in men are associated

with an increased risk of developing cardiovascular and metabolic diseases (American

College of Sports Medicine, 1995). While the WHR has been shown to be useful in

screening populations, it appears to be less accurate at indicating central fat mass

compared to other simple anthropometric measures, such as waist circumference and

sagittal diameter (Samaras & Campbell, 1997). The WHR is influenced by factors, such

as frame size and gluteal muscle mass, which may contribute to the potential inaccuracy

of this measure (Bjorntorp, 1992). Research using CT has shown that waist

circumference has a stronger relationship with intra-abdominal fat content (degree of

obesity and accumulation of VAT) than WHR (Lemieux, Prud’home, Bouchard,

Tremblay, & Despres, 1996; Pouliot, Despres, Lemieux, et al., 1994).

Pouliot, Depres, Lemieux and colleagues (1994) found that the common variance

between WHR and VAT as assessed by CT did not reach 50% in either men or women.

Conversely, use of only the waist circumference allowed the prediction of up to 75% of

the variance in VAT. In 1997, Despres concluded that waist circumference appeared to

be the best anthropometric correlate of VAT currently available. Moreover, waist girth

was more closely related to elevated blood pressure and elevated total to HDL-cholesterol

than the WHR in a sample of 10,054 Canadians (Angel, Reeder, Chen, et al., 1994).

Studies using CT have also suggested that VAT can be predicted from sagittal diameter

(Kvist, Chowdhury, Grangard, Tylen, & Sjostrom, 1988), defined as the distance between

the examination table and the highest point of the abdomen when an individual is in a

recumbent position (Kvist et al., 1988). Measurement of sagittal diameter has been

particularly useful in monitoring changes in intra-abdominal fat over time (Lamarche,

1998). In summary, waist circumference and sagittal diameter are simple and

inexpensive anthropometric measures that can be used to estimate VAT levels.

Assessing Cortisol Concentrations

The use of saliva as a means to determine cortisol concentrations is becoming

increasingly popular. There are now over 400 studies indicating that saliva is a reliable

reflection of plasma cortisol levels (Kirschbaum & Hellhammer, 1989). Because cortisol

is responsive to perceptions of stress, it is essential that apprehension and/or anxiety

associated with the sampling procedure be minimized. Clearly, venipuncture is a

procedure that elicits elevated emotions in many individuals. Therefore, ease of sampling

is one of the clear advantages of salivary hormonal assessment.

Another hindrance associated with plasma sampling is the need for trained

laboratory personnel and the inconvenience associated with repeated sampling. Salivary

samples, on the other hand, can be obtained easily and at almost any frequency by

participants outside of the laboratory environment. In addition, salivary samples can be

acquired in less than one minute and there is no evidence of altered cortisol levels post-

sampling (Kirschbaum et al., 1989). Due to the high stability of cortisol in saliva,

samples can also be stored and transported at room temperature (Aardal & Holm, 1995).

Salivary samples are stable at room temperature for at least seven days and are unaffected

by freezing at -20 degrees Celsius for up to nine months (Aardel et al., 1995). Taken

together, the advantages associated with salivary sampling make it a valuable alternative

procedure to blood analysis of cortisol.

Cortisol diffuses freely into the acinar cells of the salivary gland and then passes

easily into the saliva. Salivary hormone concentrations, therefore, are unaffected by

salivary flow rate (Kirschbaum & Hellhammer, 1994). Neither minimal stimulation by

medications causing “dry mouth” nor maximal stimulation of flow by application of citric

acid to the tongue significantly alter salivary cortisol concentrations (Cook, Harris,

Walker et al., 1986; Kahn, Rubinow, Davis et al., 1988). Furthermore, equilibrium

between serum and saliva is reached in less than five minutes (Vining, McGinley,

Maksvytis, & Ho, 1983). Addressing this point, Walker, Joyce, Davis et al. (1984)

demonstrated that the time lag between peak plasma cortisol and peak saliva cortisol

levels is one to two minutes.

The saliva assay measures unbound, free cortisol (Scerbo & Kolko, 1994), which

is the fraction that reaches the target tissue to elicit glucocorticoid effects. In other

words, it is the free fraction of cortisol that exerts physiological effects (Kirschbaum et

al., 1994). Although the concentration of cortisol in saliva is much lower than that found

in blood, both measures show significant covariation. This difference in cortisol

concentrations between blood and saliva is caused by the enzyme 11-hydroxysteroid-

dehydrogenase, which converts cortisol to cortisone and is present in large amounts in

saliva (Kirschbaum et al., 1989).

As mentioned previously, salivary cortisol concentrations can be used to

accurately reflect serum unbound cortisol concentrations throughout the physiological

range (Vining et al., 1983). Research confirming the association between cortisol in

saliva and serum is drawn from a variety of populations ranging from newborns (Gunnar,

Connors, & Isensee, 1989) to elderly subjects (Reid, Intrieri, Susman, & Beard, 1992).

Following exogenous cortisol administration, simultaneous saliva and blood measures

produced a correlation coefficient of 0.96 in healthy elderly subjects (Tunn, Mollmann,

Barth, Derendorf, & Krieg, 1992). Other published correlations between cortisol in these

two fluids range from 0.71 to 0.96, with most investigators reporting correlation

coefficients of 0.90 or greater (Kirschbaum et al., 1994). This indicates that at least 80%

of the total variation in cortisol concentration can be accounted for by salivary sampling.

Beyond the sampling medium, the normal variation in cortisol levels throughout

the day must also be considered when assessing cortisol levels. Cortisol has a regular

circadian rhythm which rises during the night, reaches its peak in the morning hours, and

falls throughout the day to the lowest level in the evening (Sherwood, 1991). When

investigating stress-related alterations in the circadian hormone profile, Kirschbaum et al.

(1989) have suggested four optimal time periods during which salivary samples should be

obtained. Based on circadian patterns, time periods when relatively small changes in

unstimulated cortisol values are observed include 8-9 a.m., 11-12 a. m., 3-4 p.m., and 8-

10 p.m.

In addition to expected physiological variation, several other factors need to be

accounted for when measuring cortisol. First, it is generally accepted that nicotine will

elevate cortisol concentrations (Pomerleau & Rosecrans, 1989). As such, this is an

intervening variable that should be controlled when measuring cortisol concentrations in

either blood or saliva. Aardal et al. (1995) suggest that study participants avoid smoking

within 60 minutes of data collection. Korbonits, Trainer, Nelson and colleagues (1996)

have also demonstrated that blood cortisol levels peak 40-60 minutes following food

ingestion and then return to baseline within 100 minutes. Accordingly, it is essential that

study participants refrain from food and beverage consumption for 100 minutes prior to

collection of saliva. Physical activity also results in elevated cortisol levels (Kirschbaum

et al., 1989). It is firmly established in the literature that salivary cortisol levels are

elevated in normal individuals during physical activities performed at exercise intensities

exceeding 70% VO2 max (Urhausen, Gabriel, & Kindermann, 1995). Therefore, it is

important that participants avoid performing any activity requiring heavy exertion prior

to salivary collection.

A final variable to consider when quantifying cortisol levels in adult females is

menstrual cycle phase. Previous research (Kirchengast et al., 1996; Korbonitis et al.,

1996) suggests that saliva samples be obtained during the follicular phase (e.g., before

ovulation occurs). Menstruation typically occurs during the first five days of the

menstrual cycle and is followed by the proliferative phase, which lasts until the 14th day

of the cycle, when ovulation occurs and the luteal phase begins (Totora & Grabowski,

1993). Therefore, the menstrual cycle can be used as a distinct marker indicating that an

adult female is in the follicular phase. Kirchengast et al. (1996) suggest that hormone

assessments be made during the 7-10th days of the cycle or within five days of the

completion of the menstrual cycle.

The collection of salivary samples offers the opportunity for frequent, non-

invasive assessment of cortisol under free- living conditions. When collecting salivary

samples, it is important to ensure that study participants do not eat, smoke, or exercise

within 60 to 100 minutes of sampling, and that females are within five days of the

completion of their menstrual cycles.

Summary

It has recently been suggested that increased cortisol levels may be linked to

abdominal obesity and psychosocial stress. Researchers are now beginning to investigate

the possibility that the hormonal response to psychosocial stress may provide a biological

pathway for the development of abdominal obesity and the symptoms and disorders

associated with this condition. Stress reactions in response to feelings of defeat or a loss

of control, which have been implicated in stimulating the hypothalamopituitary-

adrenalcortical axis, are the focus of ongoing study. These observations may provide

links between stress, somatic symptoms, and disease precursors. The proposed

investigation aims to provide an initial "snapshot" of the relationship among perceived

stress, cortisol concentrations, and abdominal fat deposition in Black, premenopausal

women. This is a promising area of research that has the potential for providing insights

into the etiology of several cardiovascular disease and a better understanding of the

physiological mechanisms underlying the association between stress and health in adult

females.