International Textbook of Obesity - part 5 pptx

This suggested the possibility ofa relationship between stress and pat-terns offat distribution that is associated with in-creased coronary artery atherosclerosis in monkeys and increase

Figure 15.1 Behavioral characteristics ofsocially dominant

(Dom; white bars) and subordinate (Sub; black bars) female monkeys Subordinates receive more aggression, are groomed less—that is they spend less time in positive affiliative behavior, spend more time alone, and they spend more time vigilantly scanning their social group than dominants Freq./h : fre- quency per hour; % time : percentage oftime spent (Based on

data from Shively et al (13,14))

PSYCHOSOCIAL FACTORS THAT

INFLUENCE CORONARY ARTERY

ATHEROSCLEROSIS AND CHD RISK

IN FEMALE MONKEYS

Social Status

Cynomolgus monkeys typically live in large social

groups that are characterized by complex social

relationships Complex social living includes the

possibility ofsocial stress effects on health A major

social organizing mechanism ofmonkey society is

the social status hierarchy (12) Female monkeys

with low social status, or subordinates, are

behav-iorally and physiologically different from

domi-nants

The distinguishing behavioral characteristics of

subordinates are depicted in Figure 15.1

Subordi-nate females are the recipients of about three times

the hostility or aggression oftheir dominant

counterparts They are groomed less, i.e they spend

less time in positive affiliative behavior They spend

more time vigilantly scanning their social group

than dominants The purpose ofthe vigilant

scann-ing appears to be to track and avoid dominants in

order to avoid aggressive interactions

Subordi-nates also spend significantly more time alone than

dominant females (13—15) Primates typically

com-municate non-verbally by touch, facial expressions

and body language or postures Although human

primates also are able to communicate with

lan-guage, they still rely heavily on non-verbal

com-munication When a female monkey spends time

alone, it means that the monkey is not in physical

contact or within touching distance ofanother

monkey Rather, the monkey is socially isolated

This is intriguing given the observations in human

beings that suggest that social support is associated

with reduced CHD risk, and observations in

mon-keys suggesting that social isolation increased

cor-onary artery atherosclerosis and heart rate (16—18).

Thus, it seems that subordinates are subject to

hos-tility and have very little social support

Physiological characteristics ofsubordinates that

distinguish them from dominants include

differen-ces in measurements ofadrenal function Following

dexamethasone suppression, the adrenal glands of

subordinate females hypersecrete cortisol in

re-sponse to an adrenocorticotropic hormone

chal-lenge, and are also relatively insensitive to cortisol

negative feedback (15) Since the hypersecretion ofcortisol is typically viewed as indicative ofa stressedindividual, these findings imply that, in general,subordinate females are stressed females

Subordinate females also have a greater numberofabnormal menstrual cycles than dominant fe-males (8) Progesterone concentrations are lowerduring the luteal phase, and estradiol concentra-tions are lower in the follicular phase of the men-strual cycles ofsubordinate females Moderatelylow luteal phase progesterone concentrations indi-cate that although ovulation may have occurred,the luteal phase was hormonally deficient Very lowluteal phase progesterone concentrations indicate

an anovulatory cycle (19,20) Thus, stressed, dinate females have poor ovarian function com-pared to dominants Subordinate females with poorovarian function have more coronary arteryatherosclerosis than their dominant counterparts(Figure 15.2) Indeed, the coronary artery atheros-clerosis extent in these subordinate, stressed females

subor-is comparable to that found in both ovariectomizedfemales and males (5,8)

The effects ofstress on ovarian function inwomen are difficult to evaluate because ofthe diffi-culties in characterizing menstrual cycle qualityover long periods oftime However, the results ofseveral studies are consistent with the hypothesisthat stress can have a deleterious effect on ovarian

function in women (21—23) Furthermore,

Figure 15.2 Coronary artery atherosclerosis (measured as

plaque extent) in males and in females in different reproductive

conditions Among females: gray bars : ovariectomized females;

black bars : intact socially subordinate females with poor

ovar-ian function; white bars : intact socially dominant females with

good ovarian function (Based on data from Hamm et al (5),

Adams et al (8))

istic pathways relating stress to impaired

reproduc-tive function in female primates have been

identifi-ed, suggesting that the stress—ovarian function

im-pairment hypothesis is plausible from a

physio-logical perspective Activation ofthe

hypothalamic-pituitary-adrenal axis, endogenous opioid

path-ways, increased prolactin release, and changes in

sensitivity to gonadal steroid hormone feedback

have all been proposed to mediate the effects of

behavioral stress on the reproductive system

(24—30) Intriguingly, women with hypothalamic

amenorrhea also have increased

hypothalamic-pi-tuitary-adrenal activity similar to that observed in

subordinate female cynomolgus monkeys (31) The

relationship between poor ovarian function during

the premenopausal years and CHD risk is also

difficult to ascertain in women due to the double

challenge ofcharacterizing ovarian function, and

detecting an adequate number ofclinical CHD

events However, La Vecchia reported that women

with a history ofirregular menstrual cycles are at

increased risk for CHD (32)

Ovarian hormones (particularly estradiol) are

also associated with the function ofthe coronary

arteries In response to neuroendocrine signals,

cor-onary arteries either dilate or constrict to modulate

the flow ofblood to the heart Inappropriate

coron-ary artery constriction, or vasospasm, early in life

may change flow dynamics, injuring the epithelium

and exacerbating atherosclerosis Coronary

vasos-pasm later in life in the presence of exacerbated

atherosclerosis may increase the likelihood

ofmy-ocardial infarction In cynomolgus monkeys, the

coronary arteries ofnormal cycling females dilate in

response to acetylcholine infused directly into thecoronary artery, whereas those ofovariectomizedfemales constrict The dilation response can be re-stored in ovariectomized females by administeringestradiol, i.e estrogen replacement therapy (33,34).The coronary arteries ofdominant females withgood ovarian function dilate in response to an infu-sion ofacetylcholine, whereas those ofsubordinatefemales with poor ovarian function constrict in re-sponse to acetylcholine (35) Thus, female primateswith poor ovarian function may be at increasedCHD risk for two reasons: (1) impaired coronaryartery function, and (2) increased atherogenesis.Ovarian function declines at menopause, particu-larly the production ofestradiol and progesterone.Importantly, clinically detectable events occur mostfrequently during and after the menopausal decline

in ovarian function Thus, the impact ofpremenopausal ovarian function on CHD risk may

be temporally separate from the clinical tion ofCHD However, atherogenesis is a dynamicprocess that occurs over a lifetime We hypothesizethat atherogenesis during young and middle adult-hood may be accelerated among socially stressedwomen These women enter the menopausal yearswith exacerbated atherosclerosis During the estro-gen-deficient menopausal years, exacerbatedatherosclerosis, combined with a more atherogeniclipid profile and increased likelihood ofcoronaryvasospasm, result in increased CHD among womenwho experienced excessive premenopausal socialstress

manifesta-Social Status, manifesta-Social Stress, and

Depression

Social stress is believed to precipitate depression

(36—40) Unfortunately, depressive disorders are

prevalent and the rate ofoccurrence is increasing(41) The results ofseveral studies suggest that lowsocial status is associated with increased risk ofdepression, although the nature ofthe relationship

is unclear (42,43) In one prospective study in whichlow social status predicted first onset ofmajor de-pressive disorder, a lack ofsocial support (socialisolation) appeared to mediate this relationship, atleast in part (44) Thus, social support may reducerisk of depression following stressful life events(45,46)

205 SOCIAL STRESS IN PRIMATES

Figure 15.3 The effects oflow social status on the prevalence of

behavioral depression in female monkeys (Based on data from

Shively et al (47))

The hypothesis that social subordination is

stressful, and results in a depressive response in

some individuals, was examined in the following

experiment Forty-eight adult female monkeys were

fed an atherogenic diet, housed in small social

groups, and social status was altered in halfofthe

animals such that halfofthe subordinates became

dominant, and halfofthe dominants became

subor-dinate (Figure 15.3)

Current subordinates hypersecreted cortisol,

were insensitive to negative feedback, and had

sup-pressed reproductive function Current

subor-dinates received more aggression, engaged in less

affiliation, and spent more time alone than

domi-nants Furthermore, they spent more time fearfully

scanning the social environment and displayed

more behavioral depression than dominants

Cur-rent subordinates with a history ofsocial

subordi-nation were preferentially susceptible to a

behav-ioral depression response The results ofthis

experiment confirm that the stress ofsocial

subordi-nation causes hypothalamic-pituitary-adrenal and

ovarian dysfunction, and support the hypothesis

that chronic, low-intensity social stress may result

in depression in susceptible individuals (15,47)

Interim Summary

Low social status in female primates is associated

with worsened coronary artery atherosclerosis

These females are the recipients of sion, and they are also relatively socially isolated.Females with low social status are also preferen-tially susceptible to a depressive response to socialstress, particularly ifthey have a history ofsocialsubordination

hostility/aggres-Social stress increases the risk ofCHD and cipitates bouts ofdepression in human beings Lowsocioeconomic status is associated with increasedrisk ofdepression and CHD The relationship be-tween socioeconomic status and health in humanbeings is linear; there is no apparent threshold Theupper class has better health than the upper middleclass, and so on down the hierarchy Risk ofdisease

pre-is increased even among relatively low social statusemployed individuals with adequate health care,nutrition, and shelter Thus, the health gradientdoes not appear to be due to poverty, per se (48).Perhaps the reason low social status is associatedwith increased risk ofdisease in human beings isbecause low social status is stressful Like the mon-keys, human primates with low social status haverelatively little control over their lives, and lowcontrol is a source ofchronic stress that could en-gender physiological responses that are deleterious

to health

REGIONAL ADIPOSITY AND CORONARY ARTERY ATHEROSCLEROSIS IN FEMALES

We examined the relationship between socialstatus, social stress, and central obesity in a series ofstudies ofsocial group-living cynomolgus monkeys

In all ofthe experiments discussed below, adultmonkeys were fed a moderately atherogenic dietthat contained between 0.25 and 0.39 mg choles-terol/calorie and 40% of calories from fat (primarilysaturated fat) These monkeys were housed in smallsocial groups offour to six animals ofthe samegender

The initial investigation ofregional adiposity andcoronary artery atherosclerosis was a retrospectivenecropsy study of36 adult female cynomolgus mon-keys (49) Whole body and regional adiposity weredetermined using anthropometric measurements.Whole body adiposity did not predict the extent ofcoronary artery atherosclerosis However, the rela-tive amount ofsubcutaneous fat deposited on the

trunk (estimated by the ratio ofsubscapular:triceps

skinfold thickness) versus the periphery was

asso-ciated with coronary artery atherosclerosis extent

Females in the top halfofthe distribution

ofsub-scapular:triceps skinfold ratio had more than three

times as much coronary artery atherosclerosis than

females in the lower half of the distribution (49)

REGIONAL ADIPOSITY AND

METABOLIC ABERRATIONS

Female cynomolgus monkeys with high central fat

have higher glucose and insulin concentrations in

an intravenous glucose tolerance test than females

with relatively low central fat They also have

high-er blood pressure and total plasma cholesthigh-erol

con-centrations, and lower HDL cholesterol

concentra-tions compared to low central fat females (50) In

women, central obesity has been linked with a

metabolic syndrome consisting ofimpaired glucose

tolerance, raised serum triglycerides and low levels

ofHDL cholesterol (51)

SOCIAL SUBORDINATION AND

REGIONAL ADIPOSITY

To determine characteristics ofsubordinate females

which increase their risk ofcoronary artery

atheros-clerosis, whole body and regional adiposity were

evaluated using anthropometric measurements in

75 adult female cynomolgus monkeys (52,53)

Sub-ordinate females were more likely than dominants

to be in the top halfofthe distribution ofthe

sub-scapular:triceps skinfold ratio This suggested the

possibility ofa relationship between stress and

pat-terns offat distribution that is associated with

in-creased coronary artery atherosclerosis in monkeys

and increased risk ofcoronary heart disease in

women Since that observation, we have attempted

to further our understanding of the potential

rela-tionship between stress and fat distribution

SOCIAL STRESS AND REGIONAL

ADIPOSITY IN MALES

Since truncal fat patterns are associated with

an-drogenic hormone profiles, it is possible that the

androgenic fat distribution pattern observed morefrequently in social subordinates is due to ovariandysfunction To begin to address this possibility, therelationship between stress and fat distribution pat-terns was studied in male monkeys Coronary ar-tery atherosclerosis is exacerbated in male cyno-molgus monkeys when their social groups arerepeatedly disrupted Social disruption has beenachieved in several experiments by altering the con-stituency ofsocial groups frequently (e.g every 4weeks) for a 2-year period Generally, the monkeysrespond to alterations in group membership by in-creased aggression and decreased affiliation (54).Thus, repeated social reorganization was used asthe stressor in the following study of males

The monkeys were assigned to treatment groupsusing a method ofstratified randomization thatmatched the groups for pretreatment plasma cho-lesterol concentrations Pretreatment anthropo-metric measures were used to control for small(non-significant) differences in adiposity that werepresent prior to treatment Computed tomographywas used to measure intra-abdominal and subcu-taneous abdominal fat in forty monkeys and re-gional skinfold thicknesses were also measured (55).Males that lived in the stress condition produced byrepeated social reorganization had significantlyhigher ratios ofintra-abdominal:subcutaneous(IA: SQ) abdominal fat (56)

This experiment provides important evidencesupporting the hypothesis that social stress can al-ter regional fat deposition The stressor was ma-nipulated by the experimenter rather than resultingfrom social group living, as in the previous observa-tion ofan association between social status andregional fat deposition These findings also indicatethat stress can alter fat distribution patterns inde-pendent ofovarian function; however, the mechan-ism(s) that might relate these two factors remains to

be determined

MECHANISMS MEDIATING THE RELATIONSHIP BETWEEN SOCIAL STRESS AND REGIONAL ADIPOSITY

To identify potential mechanisms through whichsocial stress might alter fat deposition patterns, astudy was recently completed in which behavior,the function of the hypothalamic-pituitary-adrenal

207 SOCIAL STRESS IN PRIMATES

Table 15.1 Associations between abdominal fat distribution

and behavioral characteristics offemale monkeys

IA: SQ, Intra-abdominal to subcutaneous abdominal fat ratio as

measured using computed tomography.

Figure 15.4 Association between abdominal fat distribution

and heart rate in female monkeys (Heart rates were recorded 2 months and 24 months following social group formation.) IA: SQ, Intra-abdominal to subcutaneous abdominal fat ratio as measured using computed tomography; HR, heart rate in beats per minute (bpm) ) Low IA: SQ; high IA: SQ

Figure 15.5 Dexamethasone suppression test Relationship

be-tween insensitivity to cortisol negative feedback and abdominal fat distribution Females with low IA: SQ intra-abdominal to subcutaneous fat reduce cortisol secretion by 79%, whereas those with high IA: SQ abdominal fat reduce cortisol secretion

by 54% ( :0.05) IA: SQ, intra-abdominal to subcutaneous dominal fat ratio as measured using computed tomography White line, low IA: SQ; solid line, high IA: SQ

ab-axis, and the sympathetic nervous system were

characterized in female cynomolgus monkeys

Ab-dominal fat mass was characterized by computed

tomography as previously described (49,55) The

monkeys lived in their social groups for 2 years,

and social behavior was recorded throughout this

time period Females above the mean ofthe ratio of

IA: SQ abdominal fat mass were compared to

fe-males below the mean

Females with high IA: SQ abdominal fat ratios

spent less time in affiliative social interaction, were

more frequently the victims of aggression, and were

socially subordinate compared to females with low

IA: SQ abdominal fat ratios (Table 15.1)

There was also a modest correlation between

behavioral depression and the IA: SQ ratio

(Spear-man’s rho: 0.26, P : 0.05, 1-tailed test),

suggest-ing that females with relatively greater amounts of

intra-abdominal fat were more likely to display

be-havioral depression Heart rate, a non-invasive

in-dicator ofsympathetic nervous system activity, was

measured while the animals were in their social

groups, using a telemetry system, from 15:00 h to

8:00 h the following day for three consecutive days

Heart rates ofthese animals are generally lowest at

night, increase during the time ofday when there is

the most activity in their building, and decrease in

the afternoon after the activity level in the building

declines Two months following the formation of

social groups, there were no differences in high

ver-sus low IA: SQ females However, by 24 months,

differences between these groups had emerged The

heart rates ofall females were similarly elevatedduring the day; however, in the afternoon and night,heart rates ofthe females in the high IA: SQ ab-dominal fat group were higher than those in the low

group (P: O 0.05; Figure 15.4)

Hypothalamic-pituitary-adrenal (HPA) functionwas assessed using a dexamethasone suppressiontest Suppression ofserum cortisol in response todexamethasone was greater in females in the lower

halfofthe distribution ofIA: SQ abdominal fat

mass (P: 0.05; Figure 15.5) This observation

sug-gests that the central regulatory areas ofthe HPA

axis offemales with a relatively low IA: SQ ratio are

more sensitive to circulatory cortisol

concentra-tions than those with relatively high IA: SQ Taken

together, these data suggest that females with

rela-tively greater amounts ofvisceral fat are also

char-acterized by behavioral and physiological

at-tributes indicative ofchronic stress Furthermore,

the sympathetic nervous system and the HPA axis

may mediate the relationship between social stress

and regional adiposity Our findings in cynomolgus

monkeys support the hypothesis proposed by some

that stress and a hypersensitive HPA axis are

cen-tral abnormalities in abdominal obesity ofhuman

beings (51)

SUMMARY

In primates, abdominal obesity is associated with

low social status, the metabolic syndrome, and

in-creased risk ofmorbidity and mortality due to

de-pression and cardiovascular disease Data from

stu-dies ofmonkeys suggest that social stress may be an

underlying cause We hypothesize that the stress of

social subordination or social instability causes

in-creased sympathetic nervous and HPA function

The chronic stimulation ofthese two systems leads

to increased blood pressure and heart rate, and

imbalances in sex steroid production which result in

injury to the artery wall, and deposition offat in the

viscera Visceral fat depots in turn exacerbate the

metabolic effects ofstress Some ofthese

physiologi-cal stress responses affect the function of the brain,

resulting in depression

ACKNOWLEDGEMENT

The studies reported here were funded, in part, by

grants HL-39789, HL-14164, and HL-40962 from

the National Institutes ofHealth, National Heart,

Lung, and Blood Institute, Bethesda, MD

REFERENCES

1 Brunner EJ, Marmot MG, Nanchahal K, Shipley MJ,

Stans-feld SA, Juneja M, Alberti KGMM Social inequality in

coronary risk: central obesity and the metabolic syndrome.

Evidence from the Whitehall II study Diabetologia 1997; 40: 1341—1349.

2 Wingard DL, Suarez L, Barrett-Connor E The sex tial in mortality from all causes and ischemic heart disease.

differen-Am J Epidemiol 1983; 117: 19—26.

3 Vanecek R Atherosclerosis ofthe coronary arteries in five

towns Bull WHO 1976; 53: 509—518.

4 Tejada C, Strong JP, Montenegro MR, Restropo C, Solberg

LA Distribution ofcoronary and aortic atherosclerosis by

geographic location, race, and sex Lab Invest 1968; 18: 509—526.

5 Hamm TE Jr, Kaplan JR, Clarkson TB, Bullock BC Effects ofgender and social behavior on the development ofcoron-

ary artery atherosclerosis in cynomolgus macaques clerosis 1983; 48: 221—233.

Atheros-6 Jewett DA, Dukelow WR Cyclicity and gestation length of

Macaca fascicularis Primates 1972; 13: 327—330.

7 Mahoney CJ A study ofthe menstrual cycle in Macaca irus with special reference to the detection of ovulation J Reprod Fertil 1970; 21: 153—163.

8 Adams MR, Kaplan JR, Clarkson TB, Koritnik DR Ovariectomy, social status, and atherosclerosis in cynomol-

gus monkeys Arteriosclerosis 1985; 5: 192—200.

9 Kannel WB, Hjortland MC, McNamara PM, Gordon T Menopause and risk ofcardiovascular disease: the Framin-

gham study Ann Intern Med 1976; 85: 447—452.

10 Adams MR, Kaplan JR, Manuck SR, Koritnik DR, Parks

JS, Wolfe MS, Clarkson TB Inhibition of coronary artery atherosclerosis by 17-beta estradiol in ovariectomized mon-

keys Arteriosclerosis 1990; 10: 1051—1057.

11 Bush TL, Barrett-Connor E, Cowan LD, Criqui MH, lace RB, Suchindran CM, Tyroler HA, Rifkind BM Car- diovascular mortality and noncontraceptive use ofestrogen

Wal-in women: results from the Lipid Research ClWal-inics Program

Follow-up Study Circulation 1987; 75: 1102—1109.

12 Shively CA The evolution ofdominance hierarchies in human primate society In: Ellyson SL, Dovidio JF (eds)

non-Power, Dominance, and Nonverbal Behavior New York: Springer-Verlag, 1985: 67—88.

13 Shively CA, Kaplan JR, Adams MR Effects ofovariectomy,

social instability and social status on female Macaca cicularis social behavior Physiol Behav 1986; 36: 1147—1153.

fas-14 Shively CA, Manuck SB, Kaplan JR, Koritnik DR Oral contraceptive administration, interfemale relationships, and

sexual behavior in Macaca fascicularis Arch Sex Behav 1990; 19: 101—117.

15 Shively CA Social subordination stress, behavior and tral monoaminergic function in female cynomolgus mon-

cen-keys Biol Psychiatry 1998; 44: 882—891.

16 Shumaker SA, Hill DR Gender differences in social support

and physical health Health Psychol 1991; 10: 102—111.

17 Shively CA, Clarkson TB, Kaplan JR Social deprivation and coronary artery atherosclerosis in female cynomolgus

19 Wilks JW, Hodgen GD, Ross GT Luteal phase defects in the

rhesus monkey: The significance ofserum FSH: LH ratios J

209 SOCIAL STRESS IN PRIMATES

Clin Endocrinol Metab 1976; 43: 1261—1267.

20 Wilks JW, Hodgen GD, Ross GT Endocrine characteristics

ofovulatory and anovulatory menstrual cycles in the rhesus

monkey In: Hafez ESE (ed.) Human Ovulation Amsterdam:

Elsevier/North-Holland Biomedical Press, 1979: 205—218.

21 Matteo S The effect ofjob stress and job interdependency on

menstrual cycle length, regularity and synchrony.

Psychoneuroendocrinology 1987; 12: 467—476.

22 Barnea ER, Tal J Stress-related reproductive failure J in

Vitro Fertil Embryo Transfer 1991; 8: 15—23.

23 Gindoff PR Menstrual function and its relationship to

stress, exercise, and body weight Bull NY Acad Med 1989;

65: 774—786.

24 Hayashi KT, Moberg GP Influence ofthe

hypothalamic-pituitary-adrenal axis on the menstrual cycle and the

pitu-itary responsiveness to estradiol in the female rhesus monkey

(Macaca fascicularis) Biol Reprod 1990; 42: 260—265.

25 Abbott DH, O’Byrne KT, Sheffield JW, et al

Neuroendo-crine suppression ofLH secretion in subordinate f emale

marmoset monkeys (Callithrix jacchus) In: Eley RH (ed.)

Comparative Reproduction in Mammals and Man

Proceed-ings of the Conference of the National Center for Research in

Reproduction Nairobi, Institute ofPrimate Research,

Na-tional Museums ofKenya, 1989: 63—67.

26 Abbott DH, Saltman W, Schultz-Darken NJ Hypothalamic

switches regulating fertility in primates Paper presented at

the XIVth Congress ofthe International Society

ofPrimatol-ogy, Strasbourg, France, 1992.

27 Gindoff PR, Ferin M Endogenous opioid peptides

modu-late the effect ofcorticotropin-releasing factor on

gonadot-ropin release in the primate Endocrinology 1989; 121:

837—842.

28 Ferin M Two instances ofimpaired GnRH activity in the

adult primate: The luteal phase and ‘stress’ In:

Delemarre-van de Waal HA, et al (eds) Control of the Onset of Puberty

III Amsterdam: Elsevier Science Publishers BV, 1989:

265—273.

29 Biller BM, Federoff HJ, Koenig JI, Klibanski A Abnormal

cortisol secretion and responses to corticotropin-releasing

hormone in women with hypothalamic amenorrhea J Clin

Endocrinol Metab 1990; 70: 311—317.

30 Cameron JL Stress and behaviorally induced reproductive

dysfunction in primates Semin Reprod Endocrinol 1997; 15:

37—45.

31 Nappi RE, Petraglia F, Genazzini AD, D’Ambrogio G,

Zarta C, Genazzani AR Hypothalamic amenorrhea:

evi-dence for a central derangement of

hypothalamic-pituitary-adrenal cortex axis activity Fertil Steril 1993; 59: 571—576.

32 La Vecchia C, Decarli A, Franceschi S, Gentile A, Negri E,

Parazzini F Menstrual and reproductive factors and the risk

of myocardial infarction in women under fifty-five years of

age Am J Obstet Gynecol 1987; 157: 1108—1112.

33 Williams JK, Adams MR, Klopfenstein HS Estrogen

modu-lates responses ofatherosclerotic coronary arteries

Circula-tion 1990; 81: 1680—1687.

34 Williams JK, Adams MR, Herrington DM, Clarkson TB.

Short-term administration ofestrogen and vascular

re-sponses ofatherosclerotic coronary arteries J Am Coll

Car-diol 1992; 20: 452—457.

35 Williams JK, Shively CA, Clarkson TB Vascular responses

ofatherosclerotic coronary arteries among premenopausal

female monkeys Circulation 1994; 90: 983—987.

36 Brown GW, Harris TO Aetiology ofanxiety and depressive disorders in an inner-city population 1 Early adversity.

Psychol Med 1993; 23: 143—154.

37 Brown GW Life events and affective disorder: Replications

and limitations Psychosom Med 1993; 55: 248—259.

38 Brown GW, Harris TO, Eales MJ Aetiology ofanxiety and depressive disorders in an inner-city population 2 Comor-

bidity and adversity Psychol Med 1993; 23: 155—165.

39 Brown GW, Moran P Clinical and psychosocial origins of

chronic depressive episodes I: A community survey Br J Psychiatry 1994; 165: 447—456.

40 Brown GW, Harris TO, Hepworth C Loss, humiliation and entrapment among women developing depression: a patient

and non-patient comparison Psychol Med 1995; 25: 7—21.

41 Fombonne E Increased rates ofdepression: update

ofepi-demiological findings and analytical problems Acta Psychiatr Scand 1994; 90: 145—156.

42 Murphy JM, Olivier DC, Monson RR, Sobol AM, man EB, Leighton AH Depression and anxiety in relation to

Feder-social status Arch Gen Psychiatry 1991; 48: 223—229.

43 Cole DA, Carpentieri S Social status and the comorbidity of

child depression and conduct disorder J Consult Clin Psychol 1990; 58: 748—757.

44 Bruce ML, Hoff RA Social and physical health risk factors for first-onset major depressive disorder in a community

sample Soc Psychiatry Psychiatr Epidemiol 1994; 29: 165—171.

45 Aneshensel CS, Stone JD Stress and depression: A test ofthe

buffering model ofsocial support Arch Gen Psychiatry 1982; 39: 1392—1396.

46 Serban G Stress in affective disorders Methods Achieve Exp Pathol 1991; 15: 200—220.

47 Shively CA, Laber-Laird K, Anton RF The behavior and physiology ofsocial stress and depression in female cyno-

molgus monkeys Biol Psychiatry 1997; 41: 871—882.

48 Marmot MG, Davey Smith G, Stansfeld S, et al Health

inequalities among British Civil Servants: The Whitehall II

50 Shively CA, Clarkson TB Body fat distribution and

atheros-lerosis In: Howard DF Jr (ed.) Nonhuman Primate Studies

on Diabetes, Carbohydrate Intolerance, and Obesity New York: Alan R Liss, 1988: 43—63.

51 Bjo¨rntorp P Body fat distribution, insulin resistance and

metabolic diseases Nutrition 1997; 13: 795—803.

52 Shively CA, Clarkson TB Regional obesity and coronary artery atherosclerosis in females: a nonhuman primate

model Acta Med Scand (Suppl) 1988; 723: 71—78.

53 Shively CA, Kaplan JR, Clarkson TB Carotid artery atherosclerosis in cholesterol-fed cynomolgus monkeys: The effects oforal contraceptive treatments, social factors and

regional adiposity Arteriosclerosis 1990; 10: 358—366.

54 Kaplan JR, Manuck SB, Clarkson TB, Lusso FM, Taub

DM Social status, environment, and atherosclerosis in

cynomolgus monkeys Arteriosclerosis 1982; 2: 359—368.

55 Laber-Laird K, Shively CA, Karstaedt N, Bullock BC

As-sessment of abdominal fat deposition in female cynomolgus

monkeys Int J Obes 1991; 15: 213—220.

56 Jayo J, Shively CA, Kaplan JR, Manuck SB Effects of exercise and stress on body fat distribution in male cynomol-

gus monkeys Int J Obes 1993; 17: 597—604.

211 SOCIAL STRESS IN PRIMATES

Centralization of body fat stores has proven to be

an index of several serious diseases and their

precur-sor states, indicated by risk factors Historically this

is an observation which originates from

anthropol-ogists in the early twentieth century Kretschmer (1)

noticed the difference in disease associations with

different body build, mainly from the aspect of his

own specialty, which was psychiatry He observed

that the pychnic type frequently suffered from gout,

atherosclerotic disease and stroke, and he saw early

signs of glucose intolerance as well as abnormal

pharmacological reactions of the autonomic

nerv-ous system in comparison with, particularly, the

leptosomic body build type Both these types were

also different from the aspect of food intake, where

the pychnic type increased more readily in body

weight Kretschmer made anthropometric

measure-ments from which the waist-to-hip circumference

ratio (WHR) can be calculated Such calculations

show that the pychnic type had a WHR which is

well within the risk zone for disease, as recently

suggested by the WHO (2) The pychnic type was

also more prone to develop depressive symptoms

while the leptosomic type often had a schizoid

per-sonality

Jean Vague in Marseille (3) is another pioneer

who already 50 years ago saw the differences

be-tween gynoid and android obesity and the risk for

complications in the latter Vague was focusing

mainly on adipose tissue distribution but also made

observations on other diseases than obesity

All these sharp-sighted clinical observations havebeen confirmed and extended in modern sciencewith more refined methods, as will be briefly re-viewed in this chapter

Centralization of body fat stores can be measured

in a number of ways The gold standard methods forobtainingabsolute masses of various body fat storesare the imaging techniques These methods are,however, complicated and expensive to use in epi-demiological work, where simpler surrogatemeasurements have to be employed Such methodsinclude skinfolds, which, however, do not measureintra-abdominal fat masses Various circumferencemeasurements such as waist circumference or theWHR provide an estimate of internal fat masses.The WHR has probably been somewhat better an-chored in prospective studies of disease than thewaist circumference, although the latter is slightlyeasier to measure The abdominal sagittal diameterseems to provide the most accurate estimation of theimportant visceral, intra-abdominal fat masses (4).Utilizingsuch measurements, it has now becomeincreasingly clear that body fat centralization is apowerful index of prevailingpreviously establishedrisk factors for disease such as insulin resistance,dyslipidaemia and hypertension, is found with highprevalence in already established disease, and is apowerful independent risk factor for disease in pros-pective studies The abnormalities associated withbody fat centralization span a wide range of somaticdiseases in metabolism and energy intake, such

as obesity, cardiovascular and cerebrovasculardiseases Furthermore, respiratory, haematological

International Textbook of Obesity Edited by Per Bjorntorp.

Copyright © 2001 John Wiley & Sons Ltd Print ISBNs: 0-471-988707 (Hardback); 0-470-846739 (Electronic)

and psychiatric diseases as well as cancer show

associations with centralization of body fat This is

also the case for personality characteristics, alcohol

abuse, socioeconomic and psychosocial handicaps

It is thus apparent that centralization of body fat

embraces a large cluster of human life conditions,

health and disease

Only from this wide array of conditions does it

seem unlikely that central fat distribution could be a

causative factor It seems more likely that

central-ization of body fat is an index of perturbations in

profound, central regulation of several vital systems

in the body Such regulations usually have their

origin in the hypothalamic—limbic areas of the

brain, which regulate vital functions in endocrine,

metabolic and haemodynamic systems via

neuroen-docrine and autonomic signals to the periphery

This is orchestrated by the central nervous system

into appropriate reactions to maintain homeostasis

or allostasis When various factors challenging

these counterbalancingmechanisms become too

se-vere, homeostasis or allostasis can no longer be

maintained, and disease and disease symptoms will

appear in the longrun

In this chapter certain new developments within

this area will be overviewed The input into this

research emanated originally from the obesity field,

where Jean Vague’s pioneering work has attracted

too little attention Obesity will, however, only be

briefly touched upon here, primarily with emphasis

on novel findings Instead an outlook into other

diseases and conditions will be offered Some, but

not all, of these fields are related to obesity,

suggest-ingthat body fat centralization has a much more

fundamental significance for human disease than

only in the obesity field

By approachingvarious problems in biomedical

research with epidemiological techniques on a

population basis, it is possible to obtain a wide view

on several diseases and their development, provided

that a sufficient number of well-selected variables

are examined With this method many conditions

can be analysed to search for potential pathogenetic

pathways and generate hypotheses for further

re-search Selectingout only one phylogenetic

charac-teristic for examination in case control studies

limits the focus on the particular selected variable,

for example obesity In our research we have

there-fore frequently based observations on

epi-demiological studies to obtain a wider outlook on

health problems

OBESITY THE HYPOTHALAMIC-PITUITARY- ADRENAL (HPA) AXIS

It must now be considered established that central,abdominal obesity is the malignant form of obesity.This condition seems to be associated with variousperturbations of the function of the HPA axis.About one-quarter of a male population, selected atrandom, and all 52 years of age, have signs of anelevated diurnal cortisol secretion, associated withabdominally localized excess of body fat, measuredwith the sagittal, abdominal diameter, as well assigns of metabolic derangements, characteristic ofthe metabolic syndrome (5) It seems possible toexplain the central fat accumulation as well as themetabolic derangements via effects of cortisol (6,7).The elevated cortisol secretion is seen most clearlywhen the endogenous activity of the HPA axis ismost pronounced, that is before noon (Figure 16.1).Duringthis period reports of perceived stress werealso most prevalent (unpublished) This observation

is in agreement with results of controlled animalexperiments, where chronic stress facilitates thisparticularly active phase of HPA axis activity (8),and indicates that the men examined were exposed

to a stressful environment not only duringthe day

in their ordinary life when their cortisol secretionwas measured, but also duringa period precedingthe examination

In about 10% of the population the HPA axisdisplays a depressed activity with less diurnal vari-ation, a ‘burn-out’ condition (see Figure 16.1) Thecortisol secretion is about 75% of controls, and thesecretion is again most perturbed during the highactivity phase of the HPA axis This is also in agree-ment with controlled animal experiments of chronicstress (9), and might be the end result of a develop-ment in stages from repeated stress challenges, as inthe group of men mentioned above, to eventualburn-out

In spite of not beingelevated, cortisol secretion inthis condition is associated with central obesity andits well-known associated risk factors, includinghy-pertension and elevated pulse rate Interestingly, inthis group secretions of testosterone and growthhormone are depressed (5), probably a consequence

of the challenges on their central regulation by theHPA axis perturbations (10)

Such a burned-out HPA axis has been observed

Figure 16.1 Saliva cortisol concentrations in controls (solid

line), men with high stress-related cortisol (dashed line), and men

with a burn out hypothalamic-pituitary-adrenal axis (dotted

line) (from references 5 and 17 and unpublished data) *, P
0.05;

**, P
0.01; ***, P
0.0001 (in comparison with controls)

previously in conditions of severe stress such as in

war veterans, holocaust victims, chronic pain, and

‘vital exhaustion’ (9), but it can apparently also be

found in the general population It appears that

psychosocial and socioeconomic handicaps as well

as alcohol abuse might be involved, but there are

most likely other factors involved, yet to be

identifi-ed, some of which probably have a genetic

back-ground (11,12) This condition should have a high

priority for further research because of its serious

endocrine abnormalities with associated malignant

risk factor pattern

In this condition the pathogenetic factors at play

are unlikely to include cortisol secretion, because

total cortisol secretion is low It seems possible that

the diminished secretions of sex steroid and growth

hormones are involved Deficiencies in these

hor-mones would be expected to be followed by similar

consequences as elevated cortisol, because these

hormones counteract and balance the effects of

cor-tisol in both the regulation of visceral fat mass and

insulin resistance (6,7)

Another putative pathogenetic pathway might be

via elevation of central sympathetic nervous

activ-ity, as indicated by the elevations of blood pressure

and heart rate in this condition (5) Animal

experi-ments clearly show that when the activity of the

HPA axis is insufficient or burned out the

sympath-etic nervous system is activated in compensation to

maintain homeostatic conditions (13) This would

also be expected to be followed by increased ization of free fatty acids with hepatic and muscularinsulin resistance (14,15) as well as dyslipidaemiaand perhaps diminished hepatic clearance of insulin

mobil-as consequences (16) Free fatty acids are clearlyelevated in abdominal obesity (16)

It should be observed at this point that although

it seems possible to understand the pathogeneticpathways to centralization of body fat with asso-ciated risk factors via elevated cortisol secretion (5),

as seen in the group of men with elevated related cortisol secretion, the evidence suggests thatthe kinetics of diurnal cortisol secretion are at least

stress-of equal importance In the men with elevatedstress-related cortisol secretion the increase in diur-nal cortisol is limited, about 20% in comparisonwith controls The nocturnal cortisol secretion wasnot measured, and might have been elevated also toadd to the total increase in cortisol secretion It isnoteworthy, however, that this group of men hadsignificantly lower cortisol secretion in the earliestmeasurement after wakingup (Figure 16.1) Onemight speculate that this is a sign of progressiontowards the low secretion of the group with theburned-out axis

The burned-out condition is by itself the mostpowerful evidence suggesting that the perturbedregulation of the HPA axis rather than elevatedcortisol secretion might be the crux of the matter inattempts to understand how cortisol secretion isassociated with metabolic abnormalities

The men in the group examined who had a mal HPA axis function with little or no exposure toperceived stress (see Figure 16.1) also had normalsecretions of testosterone and growth hormone aswell as normal blood pressures and heart rate This

nor-is an apparent picture of normally regulated oendocrine and autonomic functions in severalaxes Here cortisol secretion was negatively corre-lated to various risk factors This means that, forexample, absence of body fat centralization wasassociated with high cortisol, particularly whenmeasured duringthe physiological stimulation ofcortisol by food intake (17) This is an unexpectedfinding; one would actually have expected the op-posite correlation We have interpreted this to meanthat a normally functioningneuroendocrine system

neur-as indicated by a high plneur-asticity of the HPA axis,and normally functioning gonadal and growth hor-mone secretions, is associated with signs of bodilyhealth, a ‘mens sana in corpore sano’ (18)

215 CENTRALIZATION OF BODY FAT

Taken together these observations indicate that

neuroendocrine regulations are of fundamental

importance for somatic health Regulation of HPA

axis activity probably occupies a central role,

where cortisol secretion might be considered both

as an index of neuroendocrine health or

abnor-mality, with cortisol itself as a trigger of somatic

perturbations in some, but not all conditions

These considerations, based on observations,

probably explain the complicated results in

previ-ous attempts to measure cortisol secretion in

obes-ity

We have recently finalized a similar population

study in women as that reported in men (5,11,17)

The situation is different in several important

as-pects Hyperandrogeneticity (HA) is a prominent,

important abnormality in women related to

cen-tralization of body fat The highest quintile of free

testosterone is strongly associated with abdominal

obesity and conventional risk factors for

cardiovas-cular disease, stroke and type 2 diabetes mellitus

We have previously shown that HA is a powerful,

independent risk factor for these diseases, as well as

certain cancers (19), and is therefore probably a

major predisposingcondition for disease in women

The mechanism of action is likely to be induction of

insulin resistance in muscles, which then triggers

metabolic disease (20)

The nature of this HA has been examined in the

new population study It seems highly likely that its

origin is at least partly adrenal, because steroids

secreted mainly by the adrenals such as

dihyd-roepiandrosterone sulphate and cortisol are

elev-ated in parallel Furthermore, concentrations of free

testosterone show associations with features of

sal-iva cortisol concentrations, which have been shown

in men to be particularly associated with risk

fac-tors, namely low morningcortisol and

food-in-duced cortisol secretion (21) Associations between

HA and depressive traits have been found, similar

to the associations of cortisol secretion in this

con-dition, previously disclosed in men (22) Further

potential background factors are currently being

examined

There are, however, probably also other factors

involved First, the aromatase gene shows

polymor-phisms in microsatellite areas, associated with

elev-ated androgens and decreased 17 oestradiol,

which are the expected consequences of a defect

function of the aromatase enzyme, which converts

testosterone to 17 oestradiol A poorly functioning

aromatase would therefore be followed by elevatedtestosterone This may add to HA

Furthermore, the androgen receptor gene shows

a diminished number of trinucleotide repeats (CAG,glutamine) in the first exon, which might be asso-ciated with increased androgen sensitivity

The polymorphisms found are thus localized tomicrosatellites with tetra- or trinucleotide repeats inthe genes examined Such abnormalities have intro-duced a new dimension in the research of geneticabnormalities in polygenic diseases Monogenicdiseases with mutations in exons usually have an all

or nothingphylogenetic consequence The ingfeature with microsatellite polymorphisms isthat they often express themselves as quantitativeabnormalities of more or less importance (23) Forexample, we see in the example mentioned above avariation in the expression of HA dependingon thenumber of or lack of nucleotide repeats

interest-In summary, the currently performed analyses inwomen suggest that not only adrenal cortisol butalso adrenal androgens are associated with centralobesity with its risk factors and diseases Adrenalandrogens might well be also elevated in men withcentral obesity, but would be expected to be fol-lowed by minor or no peripheral consequences,because they would add only a minor fraction totestosterone produced in the gonads The insulinresistance followingHA in women is probably thetrigger for at least the metabolically related dis-eases Additional factors seem to be involved asdisease-generating triggers in women with HA, in-cludingandrogen metabolism and sensitivity

OBESITY CORTISOL METABOLISM

Cortisol is subjected to metabolic transformations

in the periphery, which are of importance for theimpact of cortisol on peripheral target tissues Thisarea is reviewed in the chapter by Walker and Seckl(Chapter 18), where detailed references can befound, and is only discussed here in relation to thecentral perturbations of HPA axis activity, re-viewed in the precedingsection

There are two main systems regulating cortisolmetabolism One is the 5 reductases which transfercortisol to tetrahydrocortisone, which is an essen-tially inactive metabolite excreted in the urine Theother system is the 11-hydroxysteroid dehyd-

rogenases (HSD), which consist of the HSD1,

vertingcortisone to cortisol, and the HSD2,

con-vertingcortisol to cortisone In humans cortisone is

a much less powerful glucocorticoid than cortisol

There is evidence for an increased activity in

obesity of 5-reductase and HSD2, which

inacti-vates cortisol This would be expected to result in

less active occupancy of the central glucocorticoid

receptors (GR) which regulate cortisol secretion by

a negative feedback mechanism (9), and an elevated

cortisol secretion would be the expected outcome

In a recent study cortisol measurements have been

adjusted for the body mass index (BMI), in an

at-tempt to examine cortisol secretion without the

influence of adipose tissue inactivation This

result-ed in a visualization of elevatresult-ed cortisol secretion in

obesity (24) Consequently peripheral inactivation

of cortisol might explain the elevated cortisol

secre-tion in obesity

It is, however, apparently not possible to explain

why cortisol secretion is particularly elevated in

centrally localized obesity, since an elevated cortisol

secretion alongthis mechanism would be expected

to be dependent on total mass of adipose tissue

irrespective of its localization Furthermore, if

corti-sol is rapidly inactivated in the peripheray, this

would not be expected to result in peripheral

conse-quences of hypercortisolism, as seen in central

obes-ity

Local elevations of the HSD1, which has been

reported to occur in visceral fat depots, might have

local effects but it seems difficult to imagine that a

secretion of cortisol from visceral fat would have

systemic effects, due to the small mass of this tissue

Cortisol from such elevated secretion would

pre-sumably also be inactivated peripherally It is also

difficult to understand the relationships, if any,

be-tween mechanisms, workingon the regulatory

centres of the HPA axis, described in the preceding

section, and peripheral metabolism of cortisol It

might be considered that the peripheral enzymes

involved in cortisol metabolism are secondarily

modified by obesity-related factors such as cotisol

itself, insulin and other hormone secretions which

are abnormal

These peripheral conversions of cortisol add to

the complexity of this field, but are clearly

import-ant for the understandingof the problems involved

Parallel studies of both the central and peripheral

mechanisms, regulating the net concentration of

circulatingglucocorticoids and their interaction

with peripheral target tissues, would be needed tounderstand potential interactions and the resultingoutcome

OBESITY PERINATAL FACTORS

Perinatal factors are likely to be involved in theproblem of centralization of body fat stores Thisidea originates from studies by Barker (25), whofound that children born small for gestational agefrequently develop centralization of body fat andassociated metabolic syndrome, ‘the small babysyndrome’ Although originally based on statisticalobservations from populations where intrauterineundernutrition was suspected, this hypothesis hasgained considerable support from the results of re-cent studies

Subjects with the small baby syndrome and dominal preponderance of body fat stores have re-cently been reported to have elevated cortisol secre-tion (26) This might correspond to the group ofmen we have studied with elevated stress-relatedcortisol and centralization of body fat (8,17).There are experimental studies which indicatepotential mechanisms The HPA axis can be sensi-tized by intrauterine exposure to immune stress orcytokine exposure or to lipopolysaccharides (27,28),and the handlingof newborns has also been shown

ab-to be of importance (29) Recent studies have vided further interestinginformation, probably ex-plainingthe effects of prenatal exposure to lipo-polysaccharides These bacterial endotoxins stimu-late the secretion of cytokines Prenatal exposure tointerleukin-6, tumour necrosis factor  or dex-amethasone, a synthetic glucocorticoid whichpasses the placental barrier, is followed by perma-nent sensitization of the HPA axis, leptin-resistantobesity and insulin resistance (30) It seems likelythat leptin-resistant obesity is caused by the in-creased corticosterone secretion from the HPA axis,because a similar condition develops after elevatedcorticosterone exposure in adult rats (31) This isprobably applicable also to humans because it iswell known from clinical experience that patientstreated with glucocorticoids overeat and becomeobese Furthermore, in recent experiments we havebeen able to show that women taking25 mgpred-nisolone daily for a week increase their food intake

pro-in spite of elevated leptpro-in concentrations (32)

217 CENTRALIZATION OF BODY FAT

These interesting developments suggest that

cen-tralization of body fat and also the development of

obesity might be affected not only by cortisol in

adulthood, but also by prenatal factors Infections

during pregnancy might speculatively be involved

in such developments

It is thus apparent that perinatal factors are

criti-cal for the development of obesity and

central-ization of body fat stores with its metabolic

associ-ates in adult life Evidence suggests that this might

at least partly be mediated via programming of the

regulation of the HPA axis It will be of interest in

the future to find out to what extent ‘the small baby

syndrome’ is involved in the overall prevalence of

centralization of body fat and the metabolic

syn-drome in adult life

HYPERTENSION

There is now considerable evidence indicatingthat

primary hypertension is frequently associated with

centralization of body fat mass (33) and the

meta-bolic syndrome (34,35) From the statistical

corre-spondence between elevated blood pressure and

insulin arose the suggestion that elevated blood

pressure might be caused by hyperinsulinaemia or

its precursor, insulin resistance This contention is

supported by experimental work showingthat the

central sympathetic nervous system is activated by

insulin (36)

New evidence is, however, not in agreement with

this chain of events (37) In statistical calculations

with blood pressure as the independent variable,

HPA axis perturbations take over all statistical

power, and blood pressure is no longer dependent

on insulin This suggests that some factor related to

HPA axis activity is a major determinant of blood

pressure This is probably the activity of the

sym-pathetic nervous system, which shows signs of

par-allel activation when the HPA axis is not

function-ingnormally (38) This is a well-described

pheno-menon with interactions between the central

regu-lation of the HPA axis and the synpathetic nervous

system at several levels In fact, it is difficult to

activate one of these axes without interferingwith

the other, due to this tight coupling of their

regula-tory centres (10)

It therefore seems likely that the relationship

be-tween insulin levels and blood pressure is due to a

parallel activation of the HPA axis and the pathetic nervous system at central levels It seemslikely that the sympathetic nervous system is re-sponsible for blood pressure elevation and the HPAaxis for insulin resistance with hyperinsulinaemiafollowingas described above The HPA axis is pre-sumably also responsible for the centralization ofbody fat as also discussed above

sym-In the case of primary hypertension; ization of body fat stores seems to be a sign ofcentral neuroendocrine disturbances where elev-ated blood pressure is probably due to a parallelactivation of the sympathetic nervous system, alsooccurringat a central level It may well be, however,that insulin amplifies this autonomic activation.For further discussion of this problem, see review in

central-Bjo¨rntorp et al (35).

MENTAL DEPRESSION

Much to our initial surprise we found in populationstudies that subjects with traits of depression andanxiety often had centralized fat depots (39) Thishas also been found in our most recent studies inboth men and women (22, and data in preparation).These traits are depressed moods, frequent use ofantidepressant drugs and anxiolytics as well as vari-

ous sleepingdifficulties (39—41) This has now also

been confirmed from other laboratories (42) As isalmost invariably the case, this centralization ofbody fat is followed by the metabolic syndrome, aswell as frequently, by hypertension

These findings are of interest from at least twoaspects Full-blown melancholic depression is acondition with severe perturbations of severalneuroendocrine axes, includingactivation of theHPA axis with poor suppression of cortisol secre-tion by dexamethasone, elevated activity of thesympathetic nervous system and inhibition of thehypotalamic-gonadal axis and growth hormonesecretion (43) These are exactly the same neuroen-docrine perturbations that occur in people withcentralization of body fat (see above) Consequent-

ly, we believe that depressive traits might be a nificant pathogenetic factor which via the neuroen-docrine perturbations will lead to body fatcentralization and the metabolic syndrome.Another aspect of interest is that depression isclearly followed by an increased risk for somatic

disease and premature mortality, also when suicide

is taken into account Prospective studies have

demonstrated that the risk for cardiovascular

dis-ease and type 2 diabetes mellitus is clearly incrdis-eased

in subjects with frequent episodes of depression,

and the risk power is comparable to that of

conven-tional somatic metabolic risk factors such as

dys-lipidaemia, insulin resistance and hypertension

(44,45) Unfortunately, such risk factors have not

been extensively followed in these prospective

stu-dies of depression A very recent study has,

how-ever, clearly shown that visceral fat masses are

elev-ated in patients with repeelev-ated depressive periods

(46)

Taken together this evidence strongly suggests

that depressive traits or clinically manifest

melan-cholic depression are associated already at early

stages with centralization of body fat masses, and

metabolic and circulatory risk factors for prevalent,

serious, somatic disease This probably provides an

explanation for the increased somatic morbidity

and mortality in depression The pathogenetic

mechanisms are most likely provided by the central

multiple neuroendocrine and auto-nomic

perturba-tions, that occur in depression as well as in subjects

who present with centralization of body fat This

field has recently been summarized, and the reader

is referred to this review for detailed references and

further discussion (47)

This has interestingtherapeutic implications

Mental depression is improved or cured by modern

pharmacological treatment This is also followed by

a correction of the neuroendocrine and autonomic

abnormalities If the proposed chain of

patho-genetic events presented above is correct, then

metabolic and haemodynamic abnormalities,

fol-lowingthe neuroendocrine and autonomic

abber-ations, would also be expected to be improved

Unfortunately, this does not seem to have been

systematically followed in psychiatric literature

We have therefore recently finalized a pilot study

with this problem in focus Men with elevated

WHR were treated with an antidepressant

inhibit-ingserotonin reuptake, but without effects on

en-ergy balance This was followed by an apparent

normalization of the signs of a perturbed activity of

the HPA axis as well as a decreased activity of the

sympathetic nervous system, and signs of metabolic

correction such as improved glucose tolerance and

insulin sensitivity Interestingly, these men did not

show any pathological scores in several depression

scales, and these scores did not change with ment, perhaps suggesting that metabolic improve-ments may occur without parallel mental changes(48) In addition, these results suggest that theserotonergic system is involved in neuroeuroendo-crine regulation, which is an established phenom-enon (10)

treat-ALCOHOL AND SMOKING

There are several reports in the literature that bacco smokingas well as elevated alcohol con-sumption is associated with centralization of bodyfat stores This is dramatically apparent in the socalled pseudo-Cushingsyndrome which is due toalcohol abuse Both tobacco smokingand alcoholintake above a limit of a couple of drinks per dayare followed by an activation of the HPA axis,providinga possible link to centralization of bodyfat stores (49)

to-PSYCHOSOCIAL AND SOCIOECONOMIC FACTORS

Psychosocial factors have been found to be ciated with an elevated WHR in both men andwomen The relationships seem stronger in men,with factors such as livingalone and divorce.Socioeconomic handicaps are also involved, includ-ingpoor education, physical type of work, low so-cial class and low income (50,51) This has alsorecently been observed in the Whitehall studies with

socioeconomic status on the one hand and an ated WHR associated with the metabolic syndrome

elev-on the other (52) In a similar treatment of our data

we find the same relationships, which are associatedwith perturbations of the HPA axis In addition,exposure time for such handicaps seems to worsenthe symptoms (53)

It seems likely that exposure to such economic and psychosocial handicaps provides abackground which would frequently expose suchindividuals to a stressful environment, and activatethe stress systems in the lower part of the brain,followed by the neuroendocrine and autonomiccascade of events, eventually leadingto central fataccumulation, the metabolic syndrome and disease

socio-219 CENTRALIZATION OF BODY FAT

This then might provide an explanation for the

social inequality of disease

PERSONALITY

Accumulating evidence now also suggests that body

fat tends to be stored in central depots in certain

types of personalities This includes both normal

variants and what has been defined as personality

disorders (54,55) Men with an elevated WHR

fre-quently score high on items of ‘novelty seeking’, and

sometimes display antisocial, histrionic and

explos-ive personalities Personality disorders include

schizoid and avoidant, dependent and passive,

ag-gressive characteristics Men with such

personali-ties might be expected to react to their surroundings

in a way that induces stress Examinations also

show that they have high values on various

re-ported stress items such as difficulties in control of

not only important things in life but also day to day

problems and annoyances, and have a high total

score in stress questionnaires These findings are in

concert with the reports of frequent perceived stress

periods, associated with perturbed neuroendocrine

functions, which supposedly are followed by

cen-tralization of body fat (5,17) as discussed in a

pre-cedingsecretion

ENDOCRINE DEFICIENCIES

Men with low testosterone, women after

meno-pause and both men and women with growth

hor-mone deficiency without involvement of HPA axis

perturbations tend to have abdominal obesity (49)

These hormones prevent accumulation of body fat

in intra-abdominal depots, and deficiency would

then be expected to be followed by enlargement of

these depots The mechanisms whereby this occurs

have been largely elucidated, and substitution with

the deficient hormone is followed by a specific

de-crease of visceral fat as well as improvement of the

factors included in the metabolic syndrome (6) The

prevalence of such conditions seem to be in the

order of 10% in the middle-aged male population

(56)

CANCER

Cancer is also predicted by increased proportions ofthe central fat stores This was first reported in asmall number of endometrial carcinomas (57), andhas subsequently been reported also for breast car-cinoma (58) and confirmed in a larger study ofendometrial carcinomas (59) Since these reportsseem to suggest that the carcinomas predicted arelocalized to tissues which are sensitive to sex steroidhormones, one might speculate that the abnormali-ties of steroid hormone secretion found in abdomi-nal obesity are also involved in this problem Ele-vated androgens are closely associated with central-ization of fat in women (21,60) as discussed in aprecedingsection, and probably originate from theadrenals as a consequence of a central drive of theHPA axis Such abnormalities indicate disturbedsecretions of sex steroid hormones which in an un-known way might be associated with these endoc-rine dependent carcinomas

GENETIC FACTORS

Genetic factors are clearly involved in the enon of central accumulation of body fat Suchfactors could be present locally in the adipose tis-sues in question, or in the regulatory mechanismsinvolved in adipose tissue distribution A majorfactor in this regard is probably the activity of theHPA axis, which has been shown to be stronglydependent on genetic factors (61)

phenom-A first target for examining molecular geneticfactors in men with elevated central body fat hasbeen the gene locus of the glucocorticoid receptor(GR), because the men with perturbed diurnal corti-sol secretion discussed in the section on obesityoften show abnormalities in the suppression of theHPA axis by dexamethasone (5) We then foundthat a known polymorphism of the GR gene locus,situated in the first intron, was associated with cen-tralization of body fat as well as insulin resistanceand, furthermore, an exaggerated stimulated corti-sol secretion (62) Furthermore, another polymor-phism in the promoter region is associated withelevated basal cortisol secretion (63) There are thusgenetic markers for centralization of fat depots inthis gene, probably, if functionally significant, act-ingvia regulation of the HPA axis

In women additional polymorphisms, localized

in microsatellites of genes involved in androgen

metabolism and sensitivity seem to be involved (21),

as discussed in a precedingsection

Other polymorphisms of potential general

inter-est for the syndrome of elevated central fat are those

involved in the regulation of the sympathetic

nerv-ous system Such polymorphisms have been found

in the beta-2-adrenergic receptor and in the

dopamine-2 receptor, both associated with elevated

blood pressure Polymorphisms of the leptin

recep-tor are, however, apparently protective for

hyper-tension in obesity (64—66).

These early findings demonstrate that the

syn-drome of central fat accumulation is associated with

several gene polymorphisms, indicating a complex

genetic background of the syndrome

WHY DOES FAT ACCUMULATE

PREDOMINANTLY IN CENTRAL

DEPOTS?

The mechanistic, mainly endocrine background to

visceral fat accumulation has been reviewed

else-where (6) One may wonder from a teleological

viewpoint why humans in a wide variety of

condi-tions store an excess fraction of body fat in central

depots

These depots are equipped with a very sensitive

fat mobilization system, which becomes even more

efficient by a dense innervation and a rich blood

flow to remove mobilized free fatty acids to the

portal circulation, and subsequently after hepatic

extraction, to systemic circulation Accumulation of

depot fat in these portally drained depots thus

serves as an easily available substrate for important

liver and peripheral functions in, for example,

muscles The substrate delivery to the periphery is

in the form of both free fatty acids and very low

density lipoprotein triglycerides, synthesized in the

liver (for review see Bjo¨rntorp (16) The

accumula-tion of central fat is more pronounced in men than

women Specific localization of fat accumulation

seems to have a clear survival value, particularly in

men, who were particularly dependent on their

muscles for survival in ancient times

One may also look upon this phenomenon as a

reserve depot for periods when the surrounding

milieu is threatening, and where much available

energy is best stored in easily mobilizable depots,

and not, for example, in the gluteo-femoral depot ofwomen, which seems to be constructed for specificchild-bearingpurposes (6) Such a construction is,however, outdated in current urbanized societies.When this excess is not used for purposes of energydelivery to muscles after longer stressful periodswith accumulation of central fat, these depots re-main intact as a sign of long-term environmentalpressures, which lead to disease by mechanismsinvolvingneuroendocrine and autonomic mechan-isms, as discussed above

GENERAL SUMMARY

This overview has attempted to summarize brieflythe multitude of conditions in which central, vis-ceral fat is accumulated in excess In all these situ-ations there seems to be a neuroendocrine back-ground affecting the HPA as well as other centralhormonal axes, often coupled to the autonomicnervous system This parallel activation is charac-teristic of an arousal reaction of centres in the lowerparts of the brain, constructed for adaptation tosurroundingpressures in order to maintain homeo-stasis or allostasis The widespread occurrence ofelevated central body fat masses suggests by itselfthat vital, common pathways are activated Theassociations between central fat and such diverseconditions as heart disease, stroke, diabetes, obes-ity, hypertension, cancer, depression, anxiety, en-docrine disturbances, personality aberrations, alco-hol abuse, socioeconomic and psychosocial handi-

pathogenetic denominator It seems likely that thisdenominator is a central arousal, induced by factors

in a competitive, hectic society Central fat lation may be considered mainly as a convenientlyobservable indicator of a chronic exposure to suchdamaging factors, remaining as an outdated sur-vival mechanism Figure 16.2 illustrates a hypoth-esis of the putative pathways linkingincreased cen-tral fat mass to diseases and conditions which havebeen shown to be statistically associated with it

accumu-ACKNOWLEDGEMENT

The studies from the author’s laboratory referred tohave been performed in collaboration with a large

221 CENTRALIZATION OF BODY FAT

Figure 16.2 Hypothetical explanation to the statistical associations between central fat depot enlargement and a large cluster of

different factors Various stress factors, includingpsychosocial and socioeconomic handicaps (psycho-soc, soc-ec), depression, anxiety, alcohol and smokingwhich, dependent on personality characteristics, via corticotrophin-releasinghormone (CRH) and endocrine perturbations, includingcortisol, sex steroid and growth hormones (endocr) direct fat to central depots, and constitute risk factors for endocrine type of cancers In addition, CRH, together with the functionally, tightly connected central sympathetic nervous system (SNS), generates metabolic and haemodynamic (hypertension) risk factors for disease

number of international and Swedish researchers

Their names are found in the reference list

REFERENCES

1 Kretschmer E Physique and Character New York:

Har-court, 1936 (Originally published in German 1921.)

2 WHO Obesity Preventing and Managing the Global

Epi-demic Report of a WHO consultation on obesity, Geneva,

3—5 June 1997, WHO/NUT/NCD/98 1 Geneva: WHO,

1998.

3 Vague J La diffe´renciation sexuelle facteur de´terminant des

formes de l’obesite Presse Med 1947; 30: 339—341.

4 Kvist H, Chowdhury B, Granga  rd U, Tyle’n U, Sjo¨stro¨m L Total and visceral adipose tissue volumes derived from measurements with computed tomography in adult men and

women: predictive equations Am J Clin Nutr 1988; 48: 1351—1361.

5 Rosmond R, Dallman MF, Bjo¨rntorp P Stress-related sol in men: Relationships with abdominal obesity and en-

corti-docrine, metabolic and hemodynamic abnormalities J Clin Endocrinol Metab 1998; 83: 1853—1859.

6 Bjo¨rntorp P The regulation of adipose tissue distribution in

humans Int J Obes 1996; 20: 291—302.

7 Bjo¨rntorp P Neuroendocrine perturbations as a cause of

insulin resistance Diabetes/Metabolism Res Rev 1999; 15:

1—15.

8 Dallman MF, Akana SF, Scribner KA Stress, feedback and

facilitation in the hypothalamo-pituitary-adrenal axis J

Neuroendocrinol 1992; 4: 516—526.

9 McEwen BS Protective and damaging effect of stress

medi-ators N Engl J Med 1998; 338: 171—179.

10 Chrousos GP, Gold PW The concept of stress and stress

system disorders Overview of physical and behavioral

ho-meostasis JAMA 1992; 267: 1244—1252.

11 Rosmond R, Lapidus L, Bjo¨rntorp P The influence of

occu-pational and social factors on obesity and fat distribution in

middle-aged men Int J Obes 1996; 20: 599—607.

12 Larsson B, Seidell J, Sva¨rdsudd K, Wilhelmsen L Tibblin G,

Bjo¨rntorp P Obesity, adipose tissue distribution and health

in men—the study of men born in 1913 Appetite 1989; 13:

37—44.

13 Plotsky PM, Cunningham Jr ET, Widmaier EP

Cate-cholaminergic modulation of corticotropin-releasing factor

and adrenocorticotropin secretion Endocr Rev 1989; 19:

437—458.

14 Randle P, Garland P, Hales N, Newsholme E The

glucose-fatty acid cycle: its role in insulin sensitivity and the

meta-bolic disturbances in diabetes mellitus Lancet 1963;

785—789.

15 Groop LC, Saloranta C, Shank M, Bonadonna RC,

Ferran-nini E, DeFronzo RA The role of free fatty acid metabolism

in the pathogenesis of insulin resistance in obesity and

non-insulin dependent diabetes mellitus J Clin Endocrinol Metab

1971; 72: 96—107.

16 Bjo¨rntorp P ‘Portal’adipose tissue as a generator of risk

factors for cardiovascular disease and diabetes

Arterioscler-osis 1900; 10: 493—496.

17 Rosmond R, Holm G, Bjo¨rntorp P Food-induced secretion

in relation to anthropometric, metabolic and

hemo-dynamic variables in men Int J Obes 2000; 24: 416—422.

18 Chrousos GP, Gold PW Editorial: A healthy body in a

healthy mind—and vice versa—the damaging power of

‘un-controllable’ stress J Clin Endocrinol Metab 1998; 83:

1842—1845.

19 Bjo¨rntorp P Abdominal fat distribution and disease: An

overview of epidemiological data Ann Med 1992; 24: 15—18.

20 Bjo¨rntorp P Metabolic implications of body fat

distribu-tion Diabetes Care 1991; 14: 1132—1143.

21 Baghaei F, Rosmond R, Eriksson E, Holm G, Bjo¨rntorp P.

Disease predictors and adrogens in women 2000 Submitted

for publication.

22 Rosmond R, Bjo¨rntorp P Endocrine and metabolic

aberra-tions in men with abdominal obesity in relation to

anxio-depressive infirmity Metabolism 1998; 47: 1187—1193.

23 Comings DE Polygenic inheritance and

micro/minisatel-lites Mol Psychiatry 1998; 3; 21—31.

24 Walker BR, SoderbergS, Lindahl B, Olsson T Independent

effects of obesity and cortisol in predictingcardiovascular

risk factors in men and women J Intern Med 2000; 247:

198—204.

25 Barker DJP Fetal origins of coronary heart disease BMJ

1995; 311: 171—174.

26 Phillips DIW, Barker DJP, Fall CHD Elevated plasma

cortisol concentrations; a link between low birthweight and

the insulin resistance syndrome? J Clin Endocrinol Metab

1998; 83: 757—760.

27 Reul JMHM, Stec J, Wiegers GJ Prenatal immune

chal-lenge alters the hypothalamic-pituitary axis in adult rats J Clin Invest 1994; 93: 2600—2607.

28 Nilsson C, Larsson B-M, Jennische E, Eriksson E, Bjo¨rntorp

P, York DA, Holma¨ngA Prenatal endotoxin exposure sults in leptin resistance, obesity and insulin resistance in adult male rats 2000 Submitted for publication.

re-29 Meaney MJ, McCormick CM, Smythe JW, Sharma S Sex specific effects of prenatal stress on hypothalamic-pituitary- adrenal responses to stress and brain glucocorticoid density

in adult rats Brain Res 1995; 84: 55—61.

30 DahlbergJ, Nilsson C, Jennische E, Hoi-Por Ho, Eriksson

E, Niklasson A, Bjo¨rntorp P, Albertsson-Widkland K, ma¨ngA Prenatal cytokine exposure results in obesity, and sex-specific programming of neuroendocrine axes in adult

Hol-rats 2000 Amer J Physiol, in press.

31 Zakrzewska KE, Cusin J, Sainsbury A, Rohner-Jeanrenaud

F, Jeanrenaud B Glucocorticoids as counterregulatory mones of leptin: toward an understandingof leptin resis-

hor-tance Diabetes 1997; 46: 717—719.

32 Udde´n J, Bjo¨rntorp P, Arner P, BarkelingB, MeurlingL, Hill M, Ro¨ssner S Glucocorticoids induce leptin resistance

in women 2000 Submitted for publication.

33 Krotkiewski M, Bjo¨rntorp P, Sjo¨stro¨m L, Smith U Impact

of obesity on metabolism in men and women J Clin Invest 1983; 72: 1150—1162.

34 Ferrannini E, Buzzigoli G, Bonadonna R Insulin resistance

in essential hypertension N Engl J Med 1987; 317: 350—357.

35 Bjo¨rntorp P, Holm G, Rosmond R, Folkow B Hypertension and the metabolic syndrome Closely related central origin?

in Swedish men J Hypertens 1998; 16: 1721—1726.

38 LjungT, Holm G, FribergP, Dallman MF, Mc Ewen BC, Bjo¨rntorp P The activity of the hypothalamic-pituitary- adrenal axis and the sympathetic nervous system in men

with elevated waist/hip circumference ratio Obes Res 2000; 8: 487—495.

39 Lapidus L, Bengtsson C, Ha¨llstro¨m T, Bjo¨rntorp P Obesity, adipose tissue distribution and health in women—results

from a population study in Gothenburg, Sweden Appetite 1989; 12: 25—35.

40 Rosmond R, Bjo¨rntorp P Psychiatric ill-health in women

and its relationship to obesity and fat distribution Obes Res 1998; 6: 338—345.

41 Rosmond R, Lapidus L, Ma  rin P, Bjo¨rntorp P Mental distress, obesity and body fat distribution in middle-aged

men Obes Res 1996; 4: 245—252.

42 WingRR, Matthews KA, Kuller LH, Meilahn EN, tinga P Waist to hip ratio in middle-aged women Associ- ations with behavioral and psychosocial factors and with

Plan-changes in cardiovascular risk factors Arterioscler Thromb 1991; 11; 1250—1257.

43 Gold PW, Chrousos GP The endocrinology of melancholic and atypical depression: relation to neurocircuitry and so-

223 CENTRALIZATION OF BODY FAT

matic consequences Proc Assoc Am Physicians 1999; 111:

22—34.

44 Pratt LA, Ford DE, Crum RM, Arnenian HK, Fallo JJ,

Eaton WW Depression, psychotropic medication, and risk

of myocardial infarction Prospective data from the

Balti-more ECA follow-up Circulation 1996; 94: 3123—3129.

45 Eaton WW, Armenian H, Gallo J, Pratt L, Ford DE

De-pression and risk for onset of type II diabetes A prospective

population-based study Diabetes Care 1996; 19: 1097—1102.

46 Thakore JH, Richards PJ, Reznek RH, Martin A, Dinan TG.

Increased intra-abdominal fat mass in patients with major

depressive illness as measured by computed tomography.

Biol Psychiatry 1997; 41: 1140—1142.

47 Bjo¨rntorp P Epidemiology of the relationship between

de-pression and physical illness In Thakore JH (ed.) Dede-pression,

a Stress-related Disorder: Physical Consequences and

Poten-tial Treatment Implications Petersfield, UK: Wrightson,

2001, in press.

48 LjungT, AhlbergA-C, Holm G, FribergP, Andersson B,

Eriksson E, Bjo¨rntorp P Treatment of men with abdominal

obesity with a serotonin reuptake inhibitor A pilot study.

2000; submitted for publication.

49 Bjo¨rntorp P Visceral obesity: a ‘civilisation syndrome’ Obes

Res 1993; 1: 206—222.

50 Rosmond R, Bjo¨rntorp P Psychosocial and socioeconomic

factors in women and their relationship to obesity and

re-gional fat distribution Int J Obes 1999; 23: 138—145.

51 Rosmond R, Lapidus L, Bjo¨rntorp P The influence of

occu-pational and social factors on obesity and body fat

distribu-tion in middle-aged men Int J Obes 1996; 20: 599—607.

52 Brunner EJ, Marmot MG, Nanchatal K Social inequality

and coronary risk: central obesity and the metabolic

syn-drome Evidence from the Whitehall II study Diabetaologia

1997; 40: 1341—1349.

53 Rosmond R, Bjo¨rntorp P Occupational status, cortisol

se-cretory pattern and visceral obesity in middle-aged men.

1999; Obes Res 2000; 8: 445—450.

54 Rosmond R, Eriksson E, Bjo¨rntorp P Personality disorders

in relation to anthropometric and metabolic factors J

En-docrinol Invest 1999; 22: 279—288.

55 AhlbergA-C, LjungT, Rosmond R, McEwen BC, Holm G,

Bjo¨rntorp P, A  kesson H-O Mood disturbance and

anthro-pometric measurements in men 2000, submitted for

publica-tion.

56 Rosmond R, Bjo¨rntorp P The interactions between

hy-pothalamic-pituitary-adrenal axis activity, testosterone,

in-sulin-like growth factor I and abdominal obesity with

me-tabolism and blood pressure in men Int J Obes Relat Metab Disord 1998; 22: 1184—1196.

57 Lapidus L, Helgesson OR, Merck C, Bjo¨rntorp P Adipose tissue distribution and female carcinomas A 12 year follow-

up of the participants in the population study of women in

Gothenburg, Sweden Int J Obes 1988; 12: 361—368.

58 Folsom AR, Kaye SA, Prineas RJ, Potter JD, Gapstur SM, Wallace RB Increased incidence of carcinoma of the breast associated with abdominal adiposity in postmenopausal

women Am J Epidemiol 1990; 131: 794—803.

59 Schapira DV, Kumar NB, Lyman GH, Cavanagh D, Roberts WS, LaPolla J Upper-body fat distribution and

endometrial cancer risk JAMA 1991; 266: 1808—1811.

60 Peiris AN, Mueller RA, Struve MF, Smith GA, Kissebah

AH Relationship of androgenic activity to splanchnic lin metabolism and peripheral glucose utilisation in

insu-premenopausal women J Clin Endocrinol Metab 1987; 64: 162—169.

61 Linkowski P, Van Ouderbergen A, Kerkhofs M, Bosson D, Mendlewicz J, Van Cauter E Twin study of the 24-h cortisol profile: Evidence for genetic control of the human circadian

clock Am J Physiol 1993; 264: E173—181.

62 Rosmond R, Chagnon YC, Holm G, Chagnon M, Pe´russe L, Carlsson B, Bouchard C, Bjo¨rntorp P A Bcl I restriction fragment length polymorphism of the glucocorticoid recep- tor gene locus is associated with abdominal obesity, leptin and dysregulation of the hypothalamic-pituitary-adrenal

axis Obes Res 2000; 8: 211—218.

63 Rosmond R, Chagnon YC, Chagnon M, Pe´russe L, Bouchard C, Bjo¨rntorp P A polymorphism of the 5 -flank- ing region of the glucocorticoid receptor gene locus is asso-

ciated with basal cortisol secretion in men Metabolism 2000,

in press.

64 Rosmond R, Ukkola U, Chagnon M, Bouchard C, torp P Polymorphisms of the 2-adrenergic receptor gene (ADRB2) in relation to cardiovascular risk factors in men.

Bjo¨rn-1999, submitted for publication.

65 Rosmond R, Rankinen T, Chagnon M, Pe´russe L, Bouchard

C, Bjo¨rntorp P A polymorphism of the dopmine 2 receptor

2 gene locus is associated with elevated blood pressure and

personality disorders in men J Clin Endocrinol Metab 2000;

Obesity and Hormonal

Abnormalities

Renato Pasquali and Valentina Vicennati

Endocrinology, S Orsola-Malpighi Hospital, Univesity Alma Mater Studiorum, Bologna, Italy

INTRODUCTION

Obesity is associated with multiple alterations in

the endocrine system, including abnormal

circula-ting blood hormone concentrations, which can be

due to changes in the pattern of their secretion

and/or metabolism, altered hormone transport

and/or action at the level of target tissues In recent

years a great stimulus in both basic and clinical

research has, on one hand, produced a great deal of

knowledge on the pathophysiology of obesity, and,

on the other, led to the discovery of new hormones,

such as leptin (1) and orexins (2)

This chapter reviews the main alterations in the

classic endocrine systems, specifically those related

to the hypothalamic-pituitary-gonadal (HPG) axis,

the growth hormone/insulin-like growth factor 1

(GH/IGF-1) axis, and the

hypothalamic-pituitary-adrenal (HPA) axis The discussion will focus on

human endocrinology, and animal studies will be

referred to only when relevant to the organization

of current knowledge Several other endocrine

sys-tems will not be discussed, and readers are referred

to extensive recent reviews in the field (3,4)

The recent discovery of the product of the ob

gene, leptin, has pointed to the role of adipose tissue

as an endocrine organ, capable of interacting with

the central nervous system and other peripheral

tissues by an integrated network, mainly devoted to

the regulation of the energy balance and fuel stores

The impressive growth of knowledge that has lowed the discovery of leptin in 1996 is under con-tinuous investigation Other chapters of this bookreview this exciting topic, which will probably rad-ically modify our clinical and therapeutic approach

fol-to obesity and related metabolic disorders in thenext few years

THE HPG AXIS IN FEMALES (Table 17.1) Sex Steroid and Gonadotropin Concentration and Metabolism

An increase in body weight and fat tissue is ciated with several abnormalities of sex steroid bal-ance in premenopausal women They involve bothandrogens and estrogens and their main transportprotein, sex hormone-binding globulin (SHBG).Changes in SHBG, which binds testosterone anddihydrotestosterone (DHT) with high affinity andestrogens with lower affinity, also lead to an alter-ation of androgen and estrogen delivery to targettissues The concentrations of SHBG are regulated

asso-by a complex of factors, which include estrogens,iodothyronines and growth hormone (GH) asstimulating, and androgens and insulin as inhibi-ting factors (5) The net balance of this regulation isprobably responsible for decreased SHBG con-

International Textbookof Obesity Edited by Per Bjorntorp.

Copyright © 2001 John Wiley & Sons Ltd Print ISBNs: 0-471-988707 (Hardback); 0-470-846739 (Electronic)

Table 17.2 Main alterations of the hypothalamic-pituitary-gonadal axis in obese women

Effect of obesity on sex hormones Increased SHBG-bound and non SHBG-bound androgen production rate and

metabolic clearance rate Reduced SHBG synthesis and concentrations Increased percentage free testosterone fraction Normal gonadotropin secretion

Increased estrogen production rate Altered active/inactive estrogen balance Impact of central obesity Worsened androgen imbalance

Treatment with androgens increases visceral fat in postmenopausal women Obesity, hyperandrogenism and PCOS Half PCOS women are overweight or obese

Obesity may have a pathogenetic role in the development of hyperandrogenism in PCOS

Obese women with PCOS have a prevalence of visceral fat distribution Hyperinsulinemia represents a pathogenetic factor of hyperandrogenism

The metabolic syndrome is part of the obesity—PCOS association

Effects of weight loss In simple obesity, improvement of androgen and SHBG imbalance

In obese women with PCOS reduction of hyperinsulinemia and insulin resistance, hyperandrogenism, and improvement of all clinical features, including fertility rate

PCOS, polycystic ovary syndrome; SHBG, sex hormone-binding globulin.

centrations in obesity, in inverse proportion to the

increase in body weight (4,5) Body fat distribution

has important effects on SHBG concentrations in

obese women In fact, those with central obesity

usually have lower SHBG concentrations in

com-parison to their age- and weight-matched

counter-parts with peripheral obesity (6) Insulin seems to

play a dominant role in this context Numerous

epidemiological studies have, in fact, demonstrated

a significantly negative correlation between insulin

and SHBG blood levels, suggesting a cause—effect

relationship (7) Moreover, studies in vitro have

shown that insulin inhibits SHBG hepatic synthesis

(8), and suppression (9) or stimulation (10) of insulin

secretion in vivo has been found to be inversely

associated with changes in SHBG concentrations,

at least in hyperandrogenic obese women Not

sur-prisingly, reduced SHBG concentrations are

there-fore commonly associated with obesity, particularly

in the central phenotype, type 2 diabetes,

hyperan-drogenic states such as polycystic ovary syndrome

(PCOS), and cardiovascular atherosclerotic

dis-eases (11), all conditions characterized by

hyperin-sulinemia and insulin resistance On the other hand,

not all obese women have reduced levels of SHBG,

in spite of similar circulating androgen and estrogen

concentrations, similar body weight and pattern of

fat distribution It has been suggested, for example,

that dietary factors may help to explain these

dis-crepancies In fact, a significantly negative

correla-tion has been found in premenopausal women tween lipid intake and SHBG levels (12) Moreover,experiments performed in men have demonstratedthat high lipid intake significantly decreased SHBGconcentrations, although contradictory data havebeen reported by others (see reference 12 for review).Although the urinary excretion rate of 17-keto-steroids may be higher than normal in obesewomen (4), the levels of the main androgens areusually high only in obese women with amenorrheaand are normal in those with regular menstrualcycles (12) Gonadotropin secretory dynamics alsoappear to be normal in eumenorrheic obese women(4) The reduction of SHBG increases the metabolicclearance rate of circulating SHBG-bound steroids,specifically testosterone, DHT and androstane 3,

be-17
-diol (3A-diol), which is the principal activemetabolite of DHT (13) However, this is compen-sated for by elevated production rates The metab-olism of the androgens not bound to SHBG is alsomodified by obesity In fact, both production ratesand metabolic clearance rates of dehydroepiandros-terone (DHEA) and androstenedione are equallyincreased in obesity (14,15), so that no difference inblood concentrations of the hormones is usuallyfound in comparison to normal-weight individuals.Androgen production and metabolism may alsoshow several differences in relation to the pattern of

body fat distribution Kirschner et al (15), for

example, found that premenopausal women with

central obesity had higher testosterone production

rates than those with peripheral obesity, whereas no

differences in androstenedione and DHT

produc-tion rate values were found Accordingly, metabolic

clearance rates of testosterone and DHT were

sig-nificantly higher in the former than in the latter The

maintenance of normal circulating levels of these

hormones in obesity suggests the presence of a

servo-mechanism of regulation which adjusts both

the production rate and the metabolic clearance

rate of these hormones to body size In women with

obesity, the rates of androgen production increase

but, due to the appreciable quantity of circulating

blood passing through the adipose tissue,

andro-gens may be cleared (metabolized) not only in the

liver but also in the fat In turn, this will result in a

reduction in hormone uptake by

androgen-sensi-tive tissues Although speculaandrogen-sensi-tive, this hypothesis

may explain why most obese women seem to be

protected against the biological effects of excessive

androgen production, such as hirsutism and

men-strual disturbances (13)

Obesity can also be considered a condition of

exaggerated estrogen production It has been

dem-onstrated that the conversion of androgens to

es-trogen in peripheral tissues is significantly

corre-lated with body weight and the amount of body fat

(16) Several other factors can contribute to this

condition of ‘functional hyperestrogenism’ (12)

Due to reduced SHBG synthesis and lower

circu-lating SHBG concentrations in obesity, the free

estradiol fraction increases, thus increasing

expo-sure of target tissues to this hormone Moreover,

the metabolism of estrogens is altered in obese

women A decreased formation of inactive

es-tradiol metabolites, such as 2-hydroxyestrogens,

which are virtually devoid of peripheral estrogen

activity, is observed, together with a higher than

normal production of estrone sulfate (which

repre-sents an important reservoir of active estrogens,

particularly estrone), due to the concurrent

reduc-tion of its metabolic clearance and increased

pro-duction rate The final result of these metabolic

derangements on estrogens is an increased ratio of

active to inactive estrogens in obese women In

spite of these alterations, blood estrogen

concen-trations are usually normal or only slightly

elev-ated in both premenopausal and postmenopausal

obese women (3,4) This may be related to the fact

that enlarged body fat may act as deposits for

ex-cess formed estrogen, thus contributing to

main-tain normal circulating hormone concentrations.Most sex steroid and SHBG alterations can beimproved by reducing body weight (17)

The Impact of Body Fat Distribution

Due to the greater reduction of SHBG tions, percentage free testosterone fraction tends to

concentra-be higher in centrally oconcentra-bese women than in thosewith peripheral obesity (18) Moreover, there arehardly ever systematic differences in the concentra-tions of principal C19 androgens between womenwith central and peripheral obesity, although theformer may have higher androstenedione levelsthan the latter (19) This may be due to the fact thatandrogen production rates are higher in womenwith central obesity than in their peripheralcounterparts (see above) An inverse correlationexists between waist-to-hip ratio (WHR) (or otherindices of body fat distribution) and testosterone orSHBG concentrations, regardless of body fatnessand body mass index (BMI) (18) Therefore, a condi-tion of ‘relative functional hyperandrogenism’ may

be present in women with the central obesityphenotype This may play an important role in thedevelopment of visceral fat deposits Androgen re-ceptors are expressed in the adipose tissue, with ahigher density in intra-abdominal than subcu-taneous deposits, at least in rats (20) In concert with

GH and catecholamines, testosterone activates thelipolytic cascade particularly in the visceraladipocytes, thus favoring increased free fatty acidrelease (20) These events are suggested as participa-ting in the development of insulin resistance andcompensatory hyperinsulinemia, both conditionsinvariably associated with central obesity In-creased insulin levels can in turn produce an inhibi-tion of SHBG synthesis, which further aggravatesthe androgen imbalance Since hyperinsulinemia

per se appears to play a role in the development of

visceral fatness, hyperinsulinemia and ‘functionalhyperandronism’ in the central obesity phenotypemay participate in a coordinated fashion to increasevisceral fat deposits in obese women This is furthersupported by the finding that exogenous androgenadministration in obese postmenopausal womenhas been shown to cause a significant gain in vis-ceral fat (as measured by computed tomographyscan) and a relatively greater loss of subcutaneousfat in comparison with placebo (21)

227 OBESITY AND HORMONAL ABNORMALITIES

Obesity and Hyperandrogenism in

Premenopausal Women: a Link with the

PCOS

Approximately half the women with PCOS are

overweight or obese (12) This association has

aroused a great deal of interest in recent years,

particularly since the discovery that PCOS women

are often hyperinsulinemic and that the degree of

hyperandrogenism may be positively and

signifi-cantly correlated with that of hyperinsulinemia (10)

The association between obesity and

hyperan-drogenism develops during puberty, and common

pathogenetic mechanisms primarily appear to

in-volve a dysregulation of insulin secretion and action

and also of the GH/IGF-I system (22) Recently,

however, it has been suggested that in obese women

with PCOS, higher than normal ovarian secretion

of androgens is associated with birthweight and

maternal obesity, suggesting that intrauterine

fac-tors may play a role in the development of the

syndrome later in life (23) Premenopausal women

with PCOS are clinically characterized by several

signs and symptoms related to hyperandrogenism

and hyperinsulinemia, including chronic

anovula-tion, hirsutism and acne Hyperandrogenism,

hy-perinsulinemia and insulin resistance and all

clini-cal features tend to be more severe in PCOS women

with abdominal body fat distribution (24) Altered

lipid profile represents another associated

meta-bolic characteristic

Pathophysiological aspects of the association

be-tween obesity and PCOS have been extensively

re-viewed in recent years (12,25,26) There may be

various mechanisms by which obesity may

influ-ence hyperandrogenism in premenopausal women

with PCOS The pivotal role of insulin was first

suggested on the basis of the significant positive

correlation observed between the degree of

hy-perandrogenism and that of hyperinsulinemia in

women with PCOS (9) In vitro studies have

subse-quently demonstrated that insulin is capable of

stimulating androgen secretion by the ovaries,

re-ducing aromatase activity in peripheral tissues and,

finally, reducing SHBG synthesis in the liver

(9,26,27) In vivo, numerous studies have

demon-strated that both acute and chronic

hyperin-sulinemia can stimulate testosterone production

and that suppression of insulin levels can conversely

decrease androgen concentrations (9,26) The fact

that hyperinsulinemia and insulin resistance are variably associated with obesity and, particularly,abdominal-visceral obesity, represents the basis forthe hypothesis supporting its role in the develop-ment of hyperandrogenism in PCOS women Suffi-cient data demonstrate that suppression of insulinlevels by diet (28,29) or chronic insulin sensitizingagent administration, such as metformin (23), trog-litazone (30), or -chiro inositol (31) can improve

in-not only the hyperandrogenic state but also thedegree of hirsutism and the fertility rate These dataobviously add further emphasis to the role of obes-ity-related hyperinsulinemia as a co-factor respon-sible for increased androgen production in obesePCOS women

As reported above, obesity is associated withsupranormal estrogen production Since estrogensexert a positive feedbackregulation upon gonado-tropin release, increased ovarian androgen produc-tion in obese PCOS women could be partly favored

by increased luteinizing hormone (LH) secretionsecondary to prevailing hyperestrogenemia (32).Obesity, as well as PCOS, is also characterized by

increased opioid system activity, and studies in vitro and in vivo have shown that
-endorphin is able tostimulate insulin secretion Moreover, the adminis-tration of
-endorphin can reduce LH release at thehypophyseal level in normal but not in PCOSwomen (33) The possibility that increased opioidactivity may favor the development of hyperin-sulinemia and, in turn, of hyperandrogenism, is fur-ther supported by the finding that both acute andchronic administration of opioid antagonists, such

as naloxone and naltrexone, suppresses both basaland glucose-stimulated insulin blood concentra-tions in a small group of obese women with PCOSand acanthosis nigricans (34) Finally, there aretheoretical possibilities that diet may play some role

in the development of the obesity—PCOS

associ-ation, although very few studies have addressed thisissue In fact, a higher than usual lipid intake hasbeen described in PCOS women by some authors(35) Mechanisms by which high lipid intake may beresponsible for altered androgen balance in suscep-tible women with obesity and PCOS have beensummarized above

Table 17.2 Main alterations of the hypothalamic-pituitary-gonadal axis in obese men

Effect of obesity on sex hormones Reduced testosterone (free and total), and C19 steroids

Reduced SHBG concentrations Reduced luteinizing hormone secretion Increased estrogen production rate Altered aromatase activity (?) Impact of body fat distribution Men with hypogonadism have typically enlarged visceral fat depots

Relationship with waist-to-hip ratio (and other indices of fat distribution) controversial Association between androstane 3, 17
-diol glucuronide and visceral fatness

Effects of weight loss Improved sex hormone imbalance (increase of testosterone)

SHBG can be restored to normal when near-normal body mass index is achieved Effect of testosterone therapy Reduction of visceral fat

Improvement of all parameters of the metabolic syndrome

SHBG, sex hormone-binding globulin.

Effects of Weight Loss and Reduction of

Insulin Concentrations in Obese

Hyperandrogenic Women with PCOS

There is long-standing clinical evidence concerning

the efficacy of weight reduction upon both the

clini-cal and endocrinologiclini-cal features of obese women

presenting PCOS Reduction of

hyperandro-genemia (namely testosterone, androstenedione,

and dehydroepiandrosterone sulfate (DHEA-S))

(28,29) appears to be the key factor responsible for

these effects However, weight loss primarily

im-proves insulin sensitivity and reduces

hyperin-sulinemia, and changes in testosterone and insulin

concentrations are significantly correlated,

regard-less of body weight variations (28,29) Recent

stu-dies have suggested that hyperinsulinemia may be

responsible for increased activity of the ovarian

cytocrome P450c17 system, which has been

im-plicated in ovarian hyperandrogenism in many

PCOS women (36) Reduction of insulin

concentra-tions by diet (37), metformin (27), or-chiro inositol

(31) has been demonstrated to reduce this enzyme

activity and, consequently, ovarian androgen

pro-duction Finally, weight loss and/or insulin

sensi-tizers also significantly improved ovulation and

fer-tility rate (28,29,31,37), further supporting the role

of hyperinsulinemia in the pathogenesis of

hyperan-drogenism in women with obesity and PCOS The

effects of dietary-induced weight loss on androgen

levels (except SHBG) seem to be peculiar to obese

hyperandrogenic women, since they have not been

reported in non-PCOS obese women (17)

THE HPG AXIS IN MALES (Table 17.2) Sex Steroid and Gonadotropin Concentration and Metabolism

Contrary to what occurs in obese women, withincreasing body weight testosterone (total and free)blood concentrations progressively decrease inobese men (36) Reduced testosterone levels are as-sociated with a progressive decrease of SHBG con-centrations as body weight increases (38) Sper-matogenesis and fertility are not affected in themajority of obese men, although they may be reduc-

ed in subjects with massive obesity (3) Serum tosterone is also inversely correlated with bodyweight in men with Kinefelter’s syndrome (3), thussupporting the causal relationship between obesityand hypotestosteronemia Serum levels of other sexsteroids have also been examined in obese men.Androstenedione concentrations are usually nor-mal or slightly reduced (39) and are not correlatedwith the degree of obesity (4) Likewise, concentra-tions of DHT are usually normal (4) Other C19steroids, such as DHEA and 3 A-diol and andros-tenediol (5-diol), may be reduced in obesity (39)

tes-As previously reported for women, estrogen duction rates are increased in male obesity in pro-portion to body weight, and blood concentrations

pro-of all major estrogens, particularly estrone, may benormal (3,4) or slightly increased (4) Altered estro-gen metabolism in obesity presumably reflects thearomatase activity of the adipose tissue, which isresponsible for active conversion of androgens intoestrogens

229 OBESITY AND HORMONAL ABNORMALITIES

tes-As... androgens and insulin as inhibi-ting factors (5) The net balance of this regulation isprobably responsible for decreased SHBG con-

International Textbookof Obesity Edited by Per Bjorntorp....

International Textbook of Obesity Edited by Per Bjorntorp.

Copyright © 2001 John Wiley & Sons Ltd Print ISBNs: 0-4 7 1-9 88707 (Hardback); 0-4 7 0-8 46739 (Electronic)