CALIDAD DEL SERVICIO EN ALIMENTOS Y BEBIDAS

3.10.1 Physiology

Waist – hip ratio (WHR) is simply the waist circumference divided by the hip circumference. It is measure of fat distribution and visceral fat mass, unlike body mass index (BMI = weight/height2), which is a measure of generalised obesity, measuring the sum of fat mass and fat free mass, regardless of the distribution. It seems likely that BMI is dependent on the regulation of energy balance via the HPA and leptin resistance, whereas localisation of body fat is probably regulated by other factors, such as endocrine perturbations [253].

There is high (but not perfect) correlation between BMI and WHR [254, 255] and increased health risks are associated with abnormalities of each, independently.

Figure 3.3 indicates the health risk associated with different body profiles. In general, low BMI (leanness) is associated with better health than high BMI, but being lean and having greater central fat is a particularly unhealthy profile, which predicts premature mortality [254].

Figure 3.3 Body fat distribution and health risk [256]

In the general population, around 65-70% of body fat is subcutaneous, 20% intra- abdominal and 12% intra-muscular [257]. Premenopausal women preferentially deposit fat in the gluteofemoral region and carry on average only 8% of their total body fat mass in the abdominal region [258]. Increased abdominal adiposity may be due to increased intra-abdominal fat or increased subcutaneous fat (lying between the abdominal muscles and the skin). Approximately 80% of intra-abdominal fat is visceral fat, while 20% is retroperitoneal [257] and it appears to be increased visceral fat that is particularly associated with poorer health outcomes [259].

Pathogenesis of central fat deposition

Fat deposition is a complex process controlled by brain neuropeptides, leptin, the SNS and the HPA. Obesity in turn affects the SNS and the HPA axis through secretion of IL-6 from adipose tissue.

During puberty, under the control of androgens and estrogens, characteristic male and female fat distribution is established. Males develop a more central fat distribution, whereas females tend to store more fat peripherally, [260, 261] although the preferential site of fat deposition is determined by genetic and environmental factors [254, 260]. Estimates of the contribution of genetic factors to the variation in fat distribution vary from around 25% [260] to 50% [262] and these genetic factors probably work through alterations in the HPA and ANS.

Perhaps the most striking example of central body fat deposition is that seen in Cushing’s syndrome, the result of diseases that cause elevated levels of cortisol. This typical pattern of fat deposition associated with cortisol excess has led to considerable research investigating the role of cortisol and the HPA axis in the development of central adiposity in non-Cushingoid individuals. Such people display increased excretion of (urinary) free cortisol that is proportional to WHR, and blunting of dexamethasone suppression (a test of an intact cortisol negative feedback on ACTH), with depressed levels of circulating sex steroids and growth hormone [263]. The HPA appears to be hyper-responsive and it is hypothesized that increased levels of cortisol stimulate the increased deposition of body fat. Sex steroids and growth hormone (GH), which have opposite effects on fat deposition, are depressed, further favoring

net fat deposition [263]. Indeed, GH and female sex hormones not only mobilize fat, but they cause it to be stored peripherally (at the hips) rather than centrally [254].

Bjorntorp theorizes that stress-induced abnormalities in the HPA are causative of central fat deposition. The defeat response to stress is characterized by elevated secretion of cortisol in response to ACTH stimulation [264] and decreased secretion of growth hormone and sex steroid hormones [263]. In humans with central adiposity, there is physiological evidence of both the fight/flight and defeat responses to stress, with hypertension, high haemoglobin concentration and increased plasma free fatty acids (FFAs) [265].

Similar outcomes have been observed in rats and in non-human primates under chronically stressful conditions. Studies in cynomolgus monkeys indicate that socially subordinate animals who are the recipients of constant aggression by dominant monkeys experience a defeat reaction associated with this chronic stressor exposure, with activation of the HPA and accumulation of excess visceral adipose tissue in a dose-dependent manner [163, 254, 261, 266-268]. There is concomitant development of insulin resistance, dyslipidemia, hypertension, impaired glucose tolerance and early signs of coronary atherosclerosis [162].

In humans, chronic over-activity of the SNS may manifest as essential hypertension [263], while chronic activation of the HPA axis results in the loss of the normal diurnal rhythm of cortisol excretion, with lower morning values and blunted responses to stressor exposure [163, 269]. In association with blunting of the HPA

axis, the central glucocorticoid receptors (GCRs) may become atrophied and less dense, with diminished cortisol feedback inhibition of CRH and ACTH and an abnormal dexamethasone suppression test. The SNS may become overactive in compensation [163]. Consequently, adults with central fat distribution may have greater cardiovascular reactivity to stress (such as blood pressure reactivity) than those with peripherally distributed fat [254].

In adipocytes, glucocorticoids (in the presence of insulin) inhibit the fat-releasing effects of insulin (by inhibiting hormone sensitive lipase) [163], and promote fat storage by enhancing the activity of lipoprotein lipase [270]. The net result is an ever- increasing accumulation of fat and enlargement of adipocytes. Visceral fat is morphologically different from peripheral fat with higher levels of GCRs (fourfold higher in intraabdominal fat than in subcutaneous fat [270]), greater blood flow and denser innervation. Thus cortisol acts via the GCRs preferentially at visceral sites to increase fat deposition [166, 263, 271-273]. This is summarized in Box 3.4.

Box 3.4. Cortisol effects on visceral fat Binding to glucocorticoid receptors (High density of receptors)

Lipoprotein lipase activated →Triglycerides accumulated

(Gene transcription & enzyme stabilization)

Lipid mobilisation inhibited → Triglycerides retained

The reproductive system is inhibited at all levels by components of the HPA axis: CRH suppresses hypothalamic LHRH, while cortisol inhibits the actions of pituitary and gonadal sex hormones on target tissues [273]. Consequently, women with central obesity may have a higher incidence of oligomenorrhea, elevated free testosterone levels [163], and low levels of sex hormone binding globulin [261]. Men with elevated WHR are prone to relative hypogonadism and abnormally low levels of testosterone [273].

Growth hormone (GH) and the sex hormones have opposite effects to cortisol on lipid accumulation, inhibiting lipoprotein lipase and stimulating lipolysis [163]. Thus the prolonged combination of elevated cortisol levels and low secretions of sex steroids and GH (as in the defeat response to stress) will tend to favour the accumulation of body fat, with preferential accumulation in intraabdominal visceral depots [163], insulin resistance and ultimately the metabolic syndrome [274]. Figure 3.4 summarizes the pathophysiology of central obesity.

Other hypotheses regarding the pathogenesis of central fat distribution include a primary adrenal [261], hypothalamic or pituitary [166] abnormality with hyper- responsiveness of the HPA and excessive secretion of CRH and cortisol. Depression of the normal negative feedback mechanism could be the result of downregulation of GCR due to HPA hyperactivity or a genetic polymorphism at the level of the coding sequence of the GCR gene [166].

Figure 3.4 Pathophysiology of central obesity (adapted from [261])

SNS: sympathetic nervous system FFA: free fatty acids

VLDL: very low density lipoproteins CVD: cardiovascular disease

ACTH: adrenocorticotrophin hormone

FSH/LH: follicle stimulating hormone/luteinizing hormone

Androgens Smoking STRESS AROUSAL ACTH FSH/LH Progesterone Abdominal obesity FEMALE CARCINOMA Cortisol SNS Hypertension FFA VLDL CVD STROKE Insulin Resistance DIABETES

3.10.2 Association with disease or risk factors for disease

Why is the distribution of body fat important?

Vague (1956) was the first to propose that the level of body fat was less important to health than the distribution of this fat, following studies of health outcomes in obese mice (cited in [259]). Subsequent studies have confirmed that centrally deposited fat is a greater risk factor for a number of diseases, including cardiovascular disease and diabetes, than peripherally deposited fat [162].

WHR and the metabolic syndrome

The metabolic syndrome is a cluster of signs including insulin resistance, dyslipidemia (elevated cholesterol and triglycerides, elevated LDL and depressed HDL cholesterol), glucose intolerance, hypertension and a hypercoagulable state. Individuals with elements of the metabolic syndrome are at increased risk of cardiovascular disease. Elevated WHR is an important feature of the metabolic syndrome and may be the driving force behind the associated abnormalities [257, 275].

WHR and cardiovascular disease

In cross-sectional studies, elevated WHR and excessive visceral fat or central body fat distribution are associated with hypertension [276, 277], peripheral arterial disease, cardiovascular disease [239, 259, 278, 279] and cerebrovascular disease [261, 280] in men and women [259, 263, 281, 282]. In many studies this association is stronger than the association between these diseases and BMI and persists after

adjustment for total cholesterol level, blood pressure, diabetes, BMI, smoking, alcohol consumption, education and race [282-285].

Increased central fat deposition is also associated with other measures of cardiovascular risk, such as increased levels of fibrinogen, total cholesterol, decreased HDL cholesterol and elevated triglycerides [257, 278, 279, 286-288]. The association between central adiposity and disease appears to be dose-related such that higher WHR predicts higher disease risk and this is independent of sex, age and race [256, 285, 289].

McGill noted that visceral adiposity at postmortem, in young persons dying of external causes, was associated with increased vascular fatty streaks and atherosclerosis in the coronary artery [239].

Association with diabetes mellitus

There is a strong association between central fat deposition and (non-insulin dependent) diabetes mellitus Type 2, in adults [261, 290, 291]. In the Iowa Women’s Health Study, women simultaneously in the highest quintiles of BMI and WHR had a relative risk (RR) of (self-reported) incidence of diabetes of 29 (95% CI 18-46) compared to women in the lowest combined quintiles. Even women in the lowest quintile for BMI had increased risk of developing diabetes if they had elevated WHR (RR = 5.1, 95% CI 3.9 - 6.8) [289].

Association with malignancy

There is a positive association of central fat tendency with breast and endometrial cancers in women [259, 261] with a relative risk (RR) for developing breast cancer of 1.3 (95% CI 1.1 - 1.5) and for endometrial cancer RR = 2.0 (95% CI 1.4 - 2.8) in women in the highest quintile of WHR compared to the lowest [289]. Among women with a family history of breast cancer, the relative risk for developing breast cancer for those in the highest quintile of WHR was 3.24 [292]. Hyperinsulinemia is commonly seen in association with central adiposity [293] and insulin is a known growth-promoting hormone for normal breast epithelium and human breast cancer cells. In addition, visceral adipose tissue synthesizes significant amounts of insulin- like growth factors, which may facilitate the response of breast cancer cells to estradiol [292].

Association with other diseases

Increased central adiposity is associated with sleep apnoea (in men) [294] and renal impairment [295] even in those with normal BMI. WHR predicts total mortality, as well as mortality from the specific diseases already outlined [289]. The relative risk for stroke in a large study of male health professionals was 2.33 (95% CI 1.25-4.37) in the highest quintile of WHR compared to the lowest [280].

WHR and BMI are also associated with psychiatric ill-health, including use of psychopharmacological drugs, personality disorders, depression, anxiety and psychosomatic disease, but this relationship appears to be stronger for WHR in women than men [261, 296, 297]. Non-obese depressed women had higher cortisol

levels and a twofold increase in intra-abdominal fat compared to controls (non- depressed) [270].

Obesity is increasing in the populations of developed countries over time, but the increase in central obesity appears to be more marked than the increase in generalized obesity [298].

3.10.3 Measurement issues

A number of different measurements have been used to evaluate central fat deposition, including waist - hip ratio, waist circumference and waist to height ratio [276, 279]. Waist circumference is often more highly correlated with BMI than with WHR and may reflect both general and abdominal obesity, thus adding little further information beyond that contributed by BMI [276, 289]. WHR appears to predict disease in adults better than other simple measurements [259].

Waist and hip measurements may each provide different health information [261] and it seems likely that it is the combination of increased waist circumference with relatively decreased hip circumference that is predictive of disease, rather than just increased waist circumference [299]. Hip circumference includes the measurement of the large muscles in the gluteal region, and muscle tissue is a major regulator of systemic insulin sensitivity [160]. A small hip circumference, indicating a small muscle mass may mean less quantity of this important tissue [160]. The combination of high waist circumference (with excessive visceral adipose tissue and release of FFAs into the circulation) and low hip circumference (with lower gluteal muscle

mass), may be particularly important in relation to insulin resistance. Thus, while waist circumference may be an important indicator of increased cardiovascular disease (CVD) risk, both hip and waist circumference may carry information for insulin resistance.

WHR is highly correlated with more advanced methods of determining visceral fat mass, e.g. magnetic resonance imaging, dual energy X-ray absorptiometry and CT scan [259, 289, 300].

There is considerable variability in the position of the measurements taken. Waist measurements have been taken at the level of the umbilicus [267], midway between the costal margin and the iliac crest [282], 2.5 cm above the umbilicus [289], at the mid-level between the processus xiphoideus and the umbilicus [233] and at the smallest diameter between the rib cage and the iliac crest [274]. Hip measurements have been taken at the level of greater trochanters [267], the largest girth between the waist and thigh [274], the widest point between the hip and the buttock [282] or at the level of maximal protrusion of the gluteal muscles [276, 289].

For consistency all subjects must be similarly clothed although there is a high correlation (r = 0.994) between measurements taken over clothing and those taken over skin [298]. Some researchers have measured the tension of the measuring tape to ensure consistency in technique [287]. Despite the variation in measurement technique, the associations with disease are robust and were apparent even when participants self-measured using a paper tape [262, 296]. Comparisons of different

versions of the WHR measurement suggest that they measure disease risk similarly [301].

Upper limits of WHR vary by gender and in different populations. A survey of Australian populations in 1993 returned a mean of 0.89 for men and 0.76 for women aged 40-44 years [301]. Increased CVD risk may be present when WHR exceeds 0.85 [259].

3.10.4 Population distribution

Age and Gender

Central fat distribution increases with age and is correlated with the level of decline of the sex hormones. Higher WHR is more common in males and post-menopausal women and is correlated with parity [302].

Genetics and intrauterine effects

The greatest risk for developing abdominal obesity and insulin resistance occurs in children who are born small but gain weight rapidly between ages 4 to 7 [257, 303]. Higher birth weight is associated with higher adult BMI, while maternal and paternal obesity predict childhood obesity in their children [303].

3.10.5 Behavioural variation

Smoking

Most studies indicate that central body fat tendency is more prevalent among smokers [259, 265, 286, 289, 297, 304], possibly associated with hypothalamic arousal, hypercortisolemia and depressed levels of sex hormones similar to that seen in the defeat response to stress. Male smokers tend to have increased WHR but decreased BMI, that is, they are leaner than non-smokers but tend to deposit fat centrally [259, 301].

Alcohol

Alcohol intake has an inconsistent association with WHR [259, 301, 302].

Physical activity

Physical activity is negatively associated with both abdominal adiposity and BMI in a dose-response manner [301, 305], although Visser found that this association was restricted to intensive exercise training (but not low or moderate intensity exercise) [305].

3.10.6 Psychosocial and SES variation

The literature examining associations between psychosocial factors and WHR and the SES variation in WHR is summarized in Table 3.4. These findings are covered in greater detail, in the context of my own findings, in the Discussion, Chapter 6.

Table 3.4 Summary of SES and psychosocial variation in waist-hip ratio. Direction of effect Findings Reference s Childhood SES Inverse association Highly consistent [281, 286, 306-310] Adult SES Inverse

association

Consistent. May be strongest with education. 30% of studies report direct relationship in men. Inconsistent results with employment status and for housing. Subjective SES may be more strongly associated than objective indicators, for women. [150, 226, 256, 265, 287, 290, 297, 301, 304, 308, 311-314] Acute stress Direct

association

May be an interaction with BMI in men [254, 315, 316] Chronic

stress

Direct association

Strain in social relationships, perceived inequality and constraint, work stress

[254, 263, 310] Affiliation and extroversion Direct association [297] Self-esteem Inverse association [254] Social support Inverse association

Lack of social support associated with increased risk of the metabolic syndrome.

[317]

3.10.7 An evolutionary perspective

Vague referred to central fat tendency as an android fat distribution, more commonly seen in males. Gynoid fat distribution, more common in women is characterised by localisation of fat in the thighs and buttocks. Abdominal adipose tissue is metabolically active and relatively easily mobilized to provide energy in time of need, such as in response to stressor exposure, and may therefore have survival value [160]. Fat stores in the gluteofemoral area, more common in premenopausal females, may have evolved primarily as an energy store (during pregnancy and lactation), insulator or to have a reproductive function [256]

3.10.8 Conclusion

WHR is a stable, easily measured, biological marker with strong associations with cardiovascular disease and insulin resistance. There is considerable evidence that central fat deposition may be a consequence of activation of the neuroendocrine stress response, particularly to stressors perceived as being overwhelming.

In addition to an association with the metabolic syndrome and a number of endocrine aberrations, central fat deposition is also associated with adverse behaviours such as smoking, excessive alcohol ingestion, sedentary lifestyle, high daily caloric intake and weight gain during adulthood as well as psychiatric disorders, such as depression and anxiety, peptic ulcers and increased use of stimulants [263, 273]. Bjorntorp suggests that this constellation of behavioural and metabolic abnormalities is the human equivalent of the defeat reaction seen in cynomolgus monkeys [160].