Dissertations / Theses on the topic 'Body composition'

Childhood obesity has been increasing rapidly. Previous work investigating body composition in obese children and adolescents has relied primarily on body mass index (BMI), or on measures that assume constant properties of fat-free mass (FFM). This limits our understanding of the effect of treatment. My PhD is divided into three aims. First, I explored differences in body composition between obese and non-obese children using multi-component models. Second, I investigated the effectiveness of two weight-loss programmes (a randomised controlled trial adopting the traffic-light program (TLP), and a pilot study evaluating treatment using Metformin). Third, I evaluated a bio-electrical impedance analysis (BIA) machine (TANITA BC-418 MA) as a clinical tool for assessing body composition in obese children.;Aim 1: obese children had greater hydration of FFM this limits the accuracy of using techniques that assume constant FFM properties. Taking this into account, obese children had increased fat mass (FM) and FFM, particularly FM in the abdomen region.;Aim 2: analyses from the TLP revealed that treatment and control groups significantly lost 0.1 BMI SDS during the trial but did not significantly differ for any of the body composition outcomes. A further analysis revealed that there was a significant reduction in BMI SDS and FM but an increase in FFM in the period when obese children were treated versus an increase in BMI SDS and FM in the period when they were left. Results from the Metformin programme revealed no significant change in body composition following 6 months or 1 year treatment.;Aim 3: Using a manufacturer's equations, TANITA was not accurate at assessing body composition or its change over time. My new equations had no systematic bias in relation to body fatness, although an error of the FFM estimate of 2.2 kg, and may be used to guide management in clinical practice.

The internet is a novel medium for delivering behavioural modification weight loss programs and owing to its widespread accessibility it has the potential to treat at the population level. This thesis aimed to evaluate the effects of a unique web-based behaviour change program (WBP) called 'Imperative Health' on weight loss in an overweight/ obese population at high risk of cardiovascular disease (CVO) using a randomised controlled trial. In total, 65 participants were randomly assigned to one of two groups for 12 months: the control group (usual care) or the intervention group (WBP). Assessments were conducted at baseline, 3-, 6- and 12-months. The primary outcome of this study was between-group change in weight loss at 3 months. Secondary outcomes included between-group change in the following health outcomes: CVD risk factors, health-related quality of life, emotional states and self efficacy. The intervention group achieved a mean weight loss of -4.4 ± 3.5 kg at 3 months; the control group lost -0.54 ± 3.0 kg, overall this accounted for a Significant mean weight difference of -3.66 kg after adjusting for baseline weight between the groups (P <0.001). No treatment effect was observed between the groups in the longer term (12 months). A substantial impact on the majority of the above health outcomes irrespective of weight loss was not observed. Web-based programs for weight loss have the potential to reach large proportions of the population and the ability to promote modest weight loss in the short term. It is difficult, however, to ascertain their long term effectiveness as a result of their associated high attrition and non-usage attrition rates. Solutions are required to enhance engagement levels in the longer term in order to maximise the potential of WBPs for weight management to treat obesity at the population level.

L'obesitat es defineix com un excés de greix en el cos, però generalment es diagnostica mitjançant mètodes que en realitat no poden mesurar o estimar el teixit adipós del cos, com l'índex de massa corporal (IMC). Hi ha moltes tècniques que poden diferenciar els compartiments corporals in vivo i després, el greix es pot estimar amb un nivell relativament alt de precisió, per exemple, absorciometria de raigs X de doble energia (DXA), pletismografia per desplaçament d'aire (ADP), dilucions isotòpiques, models multi-components, entre d'altres. El mètode de referència per avaluar la composició corporal in vivo és el model de 4-components. No obstant això, aquestes tècniques tenen algunes limitacions, principalment, totes són cares i inadequades per a la pràctica clínica. L'anàlisi d'impedància bioelèctrica (BIA) s'ha proposat com una tècnica adequada per avaluar la composició corporal en diverses poblacions, inclosos nens obesos. No obstant això, hi ha investigacions que van mostrar una baixa precisió de les mesures amb BIA en obesos. Els nostres resultats van mostrar que aquesta imprecisió podria ser deguda a l'ús de valors constants de les propietats de massa lliure de greix, hidratació i densitat, quan s’avalua la composició corporal mitjançant tècniques basades en 2-components (ADP, BIA). Tenint en compte aquest fet, aquest treball va proposar dos nous mètodes per avaluar la composició corporal en nens obesos: el primer va suggerir que utilitzar valors calculats de la densitat de la massa lliure de greix, amb la nova equació predictiva, en lloc dels valors constants publicats per mesurar la composició corporal mitjançant tècniques basades en 2-components (per exemple, ADP) millora la precisió de la tècnica; el segon mètode mostra una nova equació per calcular la massa lliure de greix a partir de les mesures d'impedància, millorant la precisió de les equacions del fabricant dels analitzadors d'impedància en població obesa.
La obesidad se define como un exceso de grasa en el cuerpo, pero generalmente se diagnostica mediante métodos que en realidad no pueden medir o estimar el tejido adiposo del cuerpo como el índice de masa corporal (IMC). Existen muchas técnicas que pueden diferenciar los compartimentos corporales in vivo y luego, la grasa se puede estimar con un nivel relativamente alto de precisión, por ejemplo, absorciometría de rayos X de doble energía (DXA), pletismografía por desplazamiento de aire (ADP), diluciones isotópicas, modelos multi-componentes, entre otros. El método de referencia para evaluar la composición corporal in vivo es el modelo de 4-componentes. Sin embargo, estas técnicas tienen algunas limitaciones, principalmente, todas son caras e inadecuadas para la práctica clínica. El análisis de impedancia bioeléctrica (BIA) se ha propuesto como una técnica adecuada para evaluar la composición corporal en diversas poblaciones, incluidos niños obesos. Sin embargo, existen investigaciones que mostraron una baja precisión de las medidas con BIA en obesos. Nuestros resultados mostraron que esta imprecisión podría deberse al uso de valores constantes de las propiedades de masa libre de grasa, hidratación y densidad, al evaluar la composición corporal mediante técnicas basadas en 2-componentes (ADP, BIA). Teniendo en cuenta este hecho, este trabajo propuso dos nuevos métodos para evaluar la composición corporal en niños obesos: el primero sugirió que usar valores calculados de la densidad de la masa libre de grasa, con la nueva ecuación predictiva, en lugar de los valores constantes publicados al evaluar composición corporal mediante técnicas basadas en 2-componentes (por ejemplo, ADP) mejora la precisión de la técnica; el segundo método muestra una nueva ecuación para calcular la masa libre de grasa a partir de las mediciones de impedancia, lo que mejora la precisión de las ecuaciones del fabricante de los analizadores de impedancia en la población obesa.
Obesity is defined as an excess of fat in the body but it is usually diagnosed by methods which cannot actually measure or estimate the adipose tissue of the body, i.e. body mass index (BMI). There are many existing techniques which can differentiate body compartments in vivo and then, fat can be estimated with a relative high level of accuracy, i.e. dual energy X-ray absorptiometry (DXA), air-displacement plethysmography (ADP), isotopic dilutions, multi-component models, among others. The gold standard method to assess body composition in vivo is the four-component model. However, these techniques have some limitations, and mainly, all of them are expensive and implausible for clinical practice. Bioelectrical impedance analysis (BIA) has been proposed as a suitable technique to assess body composition in a wide range of populations, including obese children. However, there are research evidences that showed a poor accuracy of BIA body composition assessments in this population. Our results showed that this lack of accuracy might be due to the assumption of constant values of the fat-free mass properties, hydration and density, when assessing body composition by 2-component based techniques (e.g. ADP and BIA). Considering this fact, this work proposed two new methods to assess body composition in obese children: the first one suggested that using calculated values of the density of the fat-free mass, with the new predictive equation, instead the published constant values when assessing body composition by 2-component based techniques (e.g. ADP) improves the accuracy of the technique; the second method shows a new equation to calculate the fat-free mass from whole-body impedance measurements, which improves the accuracy of the impedance analysers manufacturer’s equations in obese population.

The purpose of this study was to determine effects of diet composition and exercise on the loss and maintenance of loss of body weight and body fat; and the protection of muscle in obese subjects. Ten subjects who were members of the Ball State University (BSU) Adult Fitness Program and/or the BSU Weight Management Program, and who were at least 120 percent of ideal body weight, participated in the one-year study.Three-day diet records collected at four-month intervals were assessed using the Nutriplanner computer data base to determine energy and nutrient composition of subjects' diets. Measures of weight, triceps skinfolds (TSF), and arm circumference were collected at three-month intervals to assess changes in weight, arm fat area (AFA), and arm muscle area (AMA). Adherence to exercise was reported by participants.The findings of this study showed that obese subjects who consumed low-fat, high-carbohydrate diets while adhering to an exercise program lost, or tended to lose, more weight and arm fat area, while experiencing an increase in arm muscle area. Further studies in this area of research are needed to assess the effects of diet composition, exclusive of exercise, on changes in fat and muscle area. In addition, work is needed to develop more reliable means of assessing food intake over an extended period of time.

INTRODUCTION: Accurate body composition assessment is crucial for determining health consequences due to excess body fat (BF). While several techniques exist there are few that are accurate, non-invasive, fast, and comfortable for subjects. The Three Dimensional (3D) body scanner is a new body composition assessment method that might serve as another option for investigators and practitioners. The purpose of this study was to determine the accuracy of the 3D body scanner at measuring body composition using dual energy x-ray absorptiometry (DXA) and Air displacement plethysmography (Bod Pod) as criterion measures. The 3D body scanner was evaluated on its ability to work with differences in normal versus overweight subjects as determined by BMI. Also, a new prediction equation was created and compared to that of an existing equation used by the 3D body scanner developed by the Department of Defense (DoD).

METHODS: Eighty-Five male subjects (21.70 ± 2.28 yr old; 81.00 ± 12.21 kg; 25.37 ± 3.40 kg/m²) completed all body composition assessment techniques on the same day. Tests preformed included: DXA, Bod Pod, and 3D body scanning. Subjects did not eat or drink 2 hr previous to testing and did not exercise 4 hr previous to testing. Data was analyzed using SPSS version 17.0. Bland-Altmand plots, Pearson correlations, and a oneway ANOVA comparing means were performed. A prediction equation (3D MU) was created using a stepwise regression based on correlation to DXA.

RESULTS: Mean comparison of body composition techniques were as follows: DXA BF 16.30 ± 4.67; Bod Pod 12.17± 7.19; DoD 13.53 ± 6.43; 3D MU 16.49 ± 4.16. 3D MU had a SEE=3.09 over the entire sample compared to DoD SEE=3.67 and Bod Pod SEE=2.45. Although body volumes of Bod Pod and 3D Scanner were highly correlated (r = 0.984; p =0.001), the 3D Scanner underestimated body volume. Improvement in making consistent estimations of head, hand, and feet are necessary for the 3D body scanner to be used for body composition assessment.

CONCLUSION: Although the 3D body scanner shows promise as a method of evaluating BF, more work is needed before it can be considered an acceptable laboratory method of assessment. A 3D MU prediction equation was created that appears to be more accurate for young men than the current DoD equation. 3D body scanning shows potential as a method for determining body composition in overweight subjects.

Many of the current techniques of body composition assessment are either unsuitable for seriously ill patients or lacking precision. The disparity in electrical properties of fat and lean tissue suggests that the interaction between an alternating electromagnetic field and the human body could be a new potentially useful way to determine body composition. The purpose of the work described in this thesis was to develop an inexpensive, simple and safe non-invasive electrical technique of body composition assessment. A large coil, carrying an alternating current, produces an electromagnetic field within its volume. When a subject is placed inside the coil, the field configuration is disturbed and a corresponding change in the electrical behaviour of the coil can be observed. A theoretical analysis based on a helical waveguide model has allowed succesful prediction of the electrical behaviour of the coil when empty and also when it contains a simple homogeneous cylindrical phantom. A prototype helical waveguide system has been constructed, and the measured electric and magnetic field distributions in the coil compared with the theoretical predictions. The ability of the technique to discriminate between simple phantoms with electrical conductivities typical of muscle and of adipose tissue and between phantoms of the same conductivity but different volume has been confirmed experimentally. The suitability of the helical waveguide system for the determination of body composition in vivo has been assessed in a clinical trial involving 45 normal healthy volunteers (22 male, 23 female) ranging in age from 17 to 71 and in weight from 44 to 104 kg. Good correlations were found between changes in the electrical behaviour of the coil and fat-free mass predicted by three accepted techniques of body composition analysis.

In experiment one, week old Jersey bull calves (n=39) were assigned to one of four diets: 21/21 (n=8), 27/33 (n=8), 29/16 (n=9), MILK; or a baseline sacrifice group (n=6). Diets 21/21, 27/33, and 29/16 were milk replacers containing 21, 27, or 29% CP, and 21, 33, and 16% fat, respectively. Diet 21/21 was fed at 15% of BW. Diets 27/33, 29/16, and MILK supplied 180g CP/d. Calves were fed 4 wk. Weight, hip height, wither height, heart girth, and body length were measured weekly. Weekly plasma samples were analyzed for PUN, NEFA, and glucose. Calves were processed to estimate body composition. Feed efficiency and ADG were greatest for calves fed MILK, least for calves fed 21/21, and intermediate for calves fed 29/16 and 27/33. Calves fed 27/33 or MILK had the greatest gains of fat and percentage fat in the empty body. Body fat percentage of calves fed 29/16 or 21/21 was not changed by diet. Performance of calves fed 27/33 and 29/16 was similar except that calves fed 29/16 were leaner and calves fed 27/33 had a propensity for elevated NEFA. Feeding 180g of CP in the MR was beneficial to calf performance compared with diet 21/21. In experiment two, tissues from a subset of calves [21/21 (n=4), 27/33 (n=5), 29/16 (n=5), MILK (n=3), baseline (n=2)] were scanned using dual energy x-ray absorptiometry to estimate mass, fat, CP, and ash. Liver, organ, and carcass mass by DXA were correlated to scale weights (R²⁺= 0.99, 0.62, and 0.79, respectively). DXA was a poor predictor of percentage fat, CP, and ash (adjusted R²⁺ <0.10). Experiment three determined level of calf mortality in the United States; and identified opportunities to reduce mortality. Herds (n=88) were representative of the US Jersey population. Production averaged 7180 ± 757 kg milk annually. Herds averaged 199 births annually. Mortality was 5.0% from birth to 24 h (M24) of life and 6.7% from 24 h to 3 mo of life (M3). Level of mortality (M24) was highest in herds that calved on pasture. Lower levels of mortality (M3) were associated with use or maternity pens and earlier weaning.
Ph. D.

The objective of this investigation was to identify the appropriate isotopic fractionation factor for total body water (TBW) from ²H₂O enrichment in respiratory water vapor (RW) compared to serum (S), then to use the RW technique to estimate absolute TBW volumes and TBW relative to fat-free body mass (FFB) in three age groups (prepubescent, PP, age = 5-10 y; young adult, YA, age = 22-39 y; older adult, OA age = 65-84 y) of healthy white males and females. The effects of analytical technique (infrared spectrophotometry, IR versus isotope-ratio mass spectrometry, IRMS) and ambient relative humidity on estimates of TBW were also investigated. The composition of the FFB was estimated using a multi-component statistical model (body density, TBW and bone mineral density), and the errors associated with the traditional two-component formula for percent fat from body density were calculated. Our results demonstrated a significant (p < 0.0001) ²H₂O fractionation effect of 0.971 ± 0.005 (mean ± SEM, n = 36) for TBW from RW compared to S. Analysis by IR and IRMS were highly correlated (R² =.999) but IR values were significantly (p < 0.001) higher than IRMS. Deuterium enrichment in RW samples collected at ambient RH (∼20%) was significantly higher (Δ = 20.2 ± 4.5 ppm, mean ± SEM, p < 0.0005) than in RW samples collected at 100% RH, roughly equivalent to a 1.2 L (3.2%) difference in TBW. Total body water relative to FFB mass (W/FFB) was lower (p < 0.01) in YA males (71.0 ± 1.0%) and females (70.2 ± 1.3%) than in PP (boys = 73.1 ± 1.6%; girls = 72.2 ± 1.4%, mean ± SD). In OA, W/FFB was higher (p < 0.05) than in YA (OAM = 72.6 ± 1.1%; OAF = 72.2 ± 1.4%). The density of the FFB was 1.0996 and 1.0839 g/ml in OAM and OAF, respectively. Percent fat from density plus TBW and BMD was lower than from density alone in all groups but YA males, where it was 2.4 percent fat higher. In PP, the Siri density formula resulted in an overestimate of 5.8 ± 2.6 percent fat (mean ± SD, range = 1.4 to 13.6%). In OA females, the density formula overestimated percent fat by 4.4 ± 2.8% (range = 0 to 10.4%). In conclusion, RW corrected for isotopic fractionation will provide acceptable estimates of TBW, although the effects of analytical technique and RH should be controlled. The existence of age-related differences in FFB composition causes errors when the two-component model is used to estimate percent fat in PP and OA females.

Chronic kidney disease (CKD) is a significant public health issue. The uraemic milieu is associated with profound alterations in body composition and function. Therapeutic interventions to preserve renal function and to provide adequate homeostasis to improve outcomes in all stages of chronic kidney disease may promote other unwanted functional adversities which with careful attention to individualised treatment may be modifiable. The aim of this thesis is to clearly document these disorders of body composition and function and investigate whether commonly practiced interventions can indeed have additional deleterious impact. Our work involved subjects with different levels of CKD and included: • Antihypertensive therapy and falls in older persons with CKD 3/4. • Assessment of dynamic bone function in ERF subjects treated with haemodialysis and consequences of phosphate binder medication. • Distinguishing the dominant cardiac functional abnormalities in ERF subjects treated with haemodialysis and determination of the effects of haemodialysis on camitine depletion and its functional consequences (skeletal and myocardial). Key results included: • Antihypertensive therapy in older subjects with CKD was associated with a reduction in muscle mass over time and reduced overall function but no significant falls risk was noted. • Commonly utilised measurements to determine bone turnover in ERF subjects treated with haemodialysis do not appear to correlate with dynamic collagen formation rates. • Dobutamine-atropine stress with non-invasive assessment of cardiac parameters can be used to identify the dominant functional abnormalities that predispose to intradialytic hypotension in ERF subjects. • Skeletal muscle total carnitine decreases over the first 12 months of dialysis. Change in muscle total carnitine correlated weakly with exercise capacity. Carnitine replacement did not confer any measurable cardiovascular benefit over the first 12 months of dialysis. Body composition is highly variable over time in CKD. This is seen both in subjects receiving haemodialysis and in pre-dialysis patients. The interplay of these common alterations with the effects of treatments is potentially underestimated but should always be considered in the individualisation of patient care.

Sojourns at high altitude are often accompanied by weight loss and changes in body composition. The aim was to study body composition before and after 40 days high-altitude exposure. The subjects were four women and six men, non-smoking, healthy and active students and a scientist from Mid Sweden University in Östersund with a mean (SD) age of 26 (10) years. All subjects volunteered for a six-week trek to the Mount Everest Base Camp via Rolwaling in Nepal. Before the sojourn subject’s height was 177 (10) cm and weight was 71.9 (10) kg. Body composition was measured with Lunar iDXA at the Swedish Winter Sports Research Centre in Östersund before and after the trek. Total body mass (SD) decreased from 71.8 (10.0) kg before to 69.7 (9.4) kg after the trek (P=0.00). Total fat mass decreased from 14.7 (5.9) kg to 13.8 (4.6) kg (P=0.01). Fat percent decreased from 21.6 (7.9) % to 21.0 (7.2) % (P=0.03). Total lean mass decreased from 54.0 (10.0) kg to 52.9 (9.7) kg (P=0.01). Bone mineral content was unchanged, 3.04 (0.5) kg before and 3.03 (0.5) after (P=0.13). Thus both total body mass and total lean mass had decreased after a six week trekking in Nepal.

Both the quantity and distribution of body fat are associated with the expression of inflammatory markers. This thesis aims to examine (1) associations between measures of body composition (anthropometric and dual-energy X-ray absorptiometry (DXA)) with markers of endothelial dysfunction and inflammation, (2) the efficacy of body composition indices as predictors of inflammation in young adults and postmenopausal women, (3) the effects of modest weight change on markers of inflammation over 24 months in postmenopausal women and (4) the effects of a calcium rich marine-derived multi-mineral supplement administered alone or with short-chain fructo-oligosaccharides (scFOS) on cardiovascular disease (CVD) risk factors in postmenopausal women over 24 months. Results demonstrate that anthropometric measures of body composition are similarly associated with markers of inflammation in comparison to DXA and therefore may be reliable and clinically useful in the prediction of disease risk in young adults and postmenopausal women. Although sex differences were apparent, similar strong associations between leptin and body composition measures were evident among both men and women, highlighting a potential role for this marker in the assessment of inflammatory disease risk. Additionally, modest weight change in postmenopausal women has significant effects on pro-inflammatory marker concentrations over 24 months which may positively or negatively impact inflammatory status. With age, women experience changes in body composition, inflammation and bone health. Calcium supplements are routinely taken, particularly by postmenopausal women to meet the daily required intake and prevent osteoporosis. However, high calcium intakes via calcium containing supplements have been implicated as a risk factor for CVD. This thesis demonstrates that supplementation with a calcium rich marine-derived multi-mineral supplement alone or with scFOS, a prebiotic that enhances intestinal absorption, lowered low-density lipoprotein (LDL) and total cholesterol concentrations in postmenopausal women over 24 months and had no effect on body composition, blood pressure or systemic inflammation. Overall, this thesis demonstrates the interplay between body composition, nutrition and inflammation in young adults and postmenopausal women.

Hypertension is an important modifiable risk factor for cardiovascular disease. An important recent advancement in hypertension research is an understanding that hypertension often may have a developmental origin. Birthweight is associated with hypertension across the lifespan and adult cardiovascular disease, such that those at both ends of the spectrum are at increased risk. Nonetheless, birthweight is a crude surrogate of fetal growth and it may be that quantification of body composition, may more accurately identify the “at risk” individual. A causative mechanism linking birthweight and cardiovascular risk is yet to be identified but may involve changes to the structure and function of organs including the placenta which may impair development and predispose individuals to later cardiovascular disease. The aims of this thesis were to investigate the associations between placental function, body composition and cardiovascular autonomic function. Studies outlines in this thesis indicate different mechanism control fat mass and fat free mass in the newborn and that placental weight partly mediates the association of maternal factors with newborn body composition. While low birthweight has previously been shown to be associated altered autonomic function in the infant our studies suggests that body fatness may provide information beyond that obtained from birthweight assessment alone. Previous studies have shown altered blood pressure control in those born preterm, our studies found altered cardiovascular outcomes even in the late preterm newborn. Assessment of body composition in children and adolescents at rest and in response to an exercise test suggests worsening of autonomic control due to adiposity and may develop over time during childhood and adolescence. Collectively, these results emphasise the implications of altered in-utero and early life exposures on cardiovascular outcomes.

The aim of this research was to determine ethnic and cultural influences on body composition, lifestyle, and aspects of body image (perception, acceptability, and satisfaction) of younger (age 18-40 years) Australian and Japanese males, the latter including groups living in Australia and Japan. The sample sizes of the three groups were 68 Japanese living in Australia, 84 Japanese living in Japan, and 72 Australian Caucasian males respectively. The methodology included body composition assessments (by anthropometry and DXA), lifestyle and body image questionnaires, and dietary records. The study found significant p<0.05) ethnic differences in the %BF at given BMI levels and for Japanese the BMI values of 23.6kg/m2 and 28.6kg/m2 were found to be equivalent to 25 and 30 for Caucasians when used to classify individuals as "overweight" and "obese". Equations in common use for the calculation of body composition in Japanese males were evaluated using modern methods of body composition assessment and found to need considerable modification. New regression equations that represent BMI-%BF relationships for Japanese and Australians were proposed: Japanese: Log %BF = -1.330 + 1.896(log BMI), (R2 = 0.547, SEE = 0.09); Australians: Log %BF = -1.522 + 2.001(log BMI), (R2 = 0.544, SEE = 0.10). Equations were also developed to predict %BF for Japanese and Australian males from body composition assessments using anthropometry and DXA: Japanese: %BF = 0.376 + 0.402(abdominal) + 0.772(medial calf) + 0.217(age), (R2 = 0.786, SEE = 2.69); Australians: %BF = 2.184 + 0.392(medial calf) + 0.678(supraspinale) + 0.467(triceps), (R2 = 0.864, SEE = 2.37). Lifestyle factors were found to influence perceptions of body image.Australian males participate in physical activity more frequently than their Japanese counterparts (Australians = 98.6% involved in vigorous activity at least once per week, Japanese living in Japan = 85.7%, Japanese living in Australia = 72.1%). Significant differences p<0.05) in energy contribution patterns were found between the Japanese group (Protein: 14.4%, Carbohydrate: 50.4%, Fat: 28.1%) and Japanese living in Australia (JA: Protein: 16.3%, Carbohydrate: 47.3%, Fat: 32.3%) and the Australians (Protein: 17.1%, Carbohydrate: 47.9%, Fat: 30.6%). This shows that the Japanese living in Australia have adopted a more westemised diet than those living in Japan. Body Image assessments were done on all study groups using the Somatomorphic Matrix (SM) computer program and questionnaires, including the Ben-Tovim Walker Body Attitudes Questionnaires, (BAQ) the Attention to the Body Shape Scale (ABS), and the Eating Attitudes Test (EAT). Japanese males tended to overestimate their weight and amount of body fat, while Australian Caucasian males underestimated these parameters. The Japanese groups had higher scores on the selfdisparagement subscale and lower scores on the strengths and the attractiveness subscales of the BAQ questionnaire than Australian males. Australian males also had higher scores on the EAT total score and the dieting subscale of the EAT questionnaire than Japanese males. When all groups of subjects selected their perceived body image from the SM program menu, these results had no relationship with measured body composition values, suggesting that further development of this program is needed for use in these populations.

The aim of this research was to determine ethnic and cultural influences on body composition, lifestyle, and aspects of body image (perception, acceptability, and satisfaction) of younger (age 18-40 years) Australian and Japanese males, the latter including groups living in Australia and Japan. The sample sizes of the three groups were 68 Japanese living in Australia, 84 Japanese living in Japan, and 72 Australian Caucasian males respectively. The methodology included body composition assessments (by anthropometry and DXA), lifestyle and body image questionnaires, and dietary records. The study found significant p<0.05) ethnic differences in the %BF at given BMI levels and for Japanese the BMI values of 23.6kg/m2 and 28.6kg/m2 were found to be equivalent to 25 and 30 for Caucasians when used to classify individuals as "overweight" and "obese". Equations in common use for the calculation of body composition in Japanese males were evaluated using modern methods of body composition assessment and found to need considerable modification. New regression equations that represent BMI-%BF relationships for Japanese and Australians were proposed: Japanese: Log %BF = -1.330 + 1.896(log BMI), (R2 = 0.547, SEE = 0.09); Australians: Log %BF = -1.522 + 2.001(log BMI), (R2 = 0.544, SEE = 0.10). Equations were also developed to predict %BF for Japanese and Australian males from body composition assessments using anthropometry and DXA: Japanese: %BF = 0.376 + 0.402(abdominal) + 0.772(medial calf) + 0.217(age), (R2 = 0.786, SEE = 2.69); Australians: %BF = 2.184 + 0.392(medial calf) + 0.678(supraspinale) + 0.467(triceps), (R2 = 0.864, SEE = 2.37). Lifestyle factors were found to influence perceptions of body image.
Australian males participate in physical activity more frequently than their Japanese counterparts (Australians = 98.6% involved in vigorous activity at least once per week, Japanese living in Japan = 85.7%, Japanese living in Australia = 72.1%). Significant differences p<0.05) in energy contribution patterns were found between the Japanese group (Protein: 14.4%, Carbohydrate: 50.4%, Fat: 28.1%) and Japanese living in Australia (JA: Protein: 16.3%, Carbohydrate: 47.3%, Fat: 32.3%) and the Australians (Protein: 17.1%, Carbohydrate: 47.9%, Fat: 30.6%). This shows that the Japanese living in Australia have adopted a more westemised diet than those living in Japan. Body Image assessments were done on all study groups using the Somatomorphic Matrix (SM) computer program and questionnaires, including the Ben-Tovim Walker Body Attitudes Questionnaires, (BAQ) the Attention to the Body Shape Scale (ABS), and the Eating Attitudes Test (EAT). Japanese males tended to overestimate their weight and amount of body fat, while Australian Caucasian males underestimated these parameters. The Japanese groups had higher scores on the selfdisparagement subscale and lower scores on the strengths and the attractiveness subscales of the BAQ questionnaire than Australian males. Australian males also had higher scores on the EAT total score and the dieting subscale of the EAT questionnaire than Japanese males. When all groups of subjects selected their perceived body image from the SM program menu, these results had no relationship with measured body composition values, suggesting that further development of this program is needed for use in these populations.

No research has been performed comparing percent body fat measurements using dual energy x-ray absorptiometry (DEXA) and air plethysmography after weight loss. The purpose of this investigation was to compare body composition assessments using the Bod Pod® Body Composition System (BP) and the DEXA ProdigyTM before and after an 8-week weight loss diet and exercise program. Based on prior comparison studies, it was hypothesized that percent fat values would be significantly lower using the BP compared to the DEXA before and after weight loss. Also since both methods have been shown to be reliable, it was hypothesized that the BP and DEXA would detect similar changes in percent fat in response to a weight-loss intervention. Twelve women (42 ± 8 yrs) and ten men (40 + 11 yrs) had their percent body fat estimated using the BP and whole body DEXA ProdigyTM scanner. A similar significant (P < 0.05) decrease in percent body fat was observed using the BP and DEXA ProdigyTM (-2.2% and -1.8%, respectively) after weight loss. Percent body fat using the BP was significantly lower than the DEXA ProdigyTM before (36 ± 10.7% and 38.1 ± 9.4%, respectively) and after (33.8 ± 10.8% and 36.3 + 10.6%, respectively) the weight loss program. Significant Pearson correlation coefficients between the DEXA ProdigyTM and the BP were noted pre (r = 0.975) and post (r = 0.968) weight loss. These data indicate the BP provides significantly lower absolute %fat values within a range of body fat levels (20-56%) when compared to the DEXA. However, the BP and DEXA detect similar changes in %fat, fatfree mass, and fat mass from weight loss in men and women. Also, the BP consistently overestimated fat-free mass and underestimated %fat and fat mass compared to values obtained from DEXA.
School of Physical Education

The body composition of female collegiate athletes was measured using the Bod Pod® device. The sample consisted of 75 student athletes, aged 18 to 22 years old. Five sports at the university level were represented, including basketball, gymnastics, soccer, swimming & diving, and soccer. Participants were measured at the preseason and postseason periods. Overall, participants in all five sports were not found to change significantly in total body mass, fat mass, fat free mass, percent body fat, or body mass index from the preseason period to the postseason period at the alpha = 0.05 level. On average, the members from each of the different teams were found to be significantly different from each other for one or more variables. In general, basketball and volleyball players were found to be similar in body composition. The average member on the swimming & diving, soccer, and gymnastics teams was found to vary from the average team member on each of the other teams.

The purpose of this study was to determine the reliability of within-day and between-day measurements of body volume (BV), thoracic gas volume (Vtg), and body fat percentages (%fat) taken in the BOD POD. Fifty subjects (25 men, 25 women) of varying ages (20-81 yrs.), races, and body composition (2.2-51.1% fat) were tested under ideal conditions set forth by the manufacturer at the same time of day (± 3 hours) on three separate days within two weeks. On each day, testing was repeated in succession until three trials were obtained which met the manufacturer's criteria for acceptable BV and Vtg. Repeated measures ANOVA revealed no significant mean differences between the three within-day trials or for the first trial between each of the three days with the exception of between-day BV where day 1 was found to be 260 ml greater than days 2 and 3. Coefficient of variation (SD/mean* 100) and intra-class correlations were determined for each variable both within the three trials of each day and between the first trials of each of the days. For BV, the CV for the 150 cases of within-day measurements was.1 ± .01% while the between-day CV was .4 +.3%. For both within and between day BV, r = .99. The CV and correlations for Vtg was 2.8 ± 2.4% and .99 and 4.1 ± 2.6% and .98 for within-day and between-day measures, respectively. The mean absolute difference between measured Vtg and the predicted value (Vpred) from age, height, and weight was 0.471 ± 0.392 L which would result in body fat estimates that differ by 1.3 ± 0.3%. Estimates of body fat from Vpred were within +2% fat of those derived from measured Vtg values in 80% of the trials. Body fat percentage produced a within-day CV of 2.8 + 4.6% with an r = .99 and a between-day CV of 3.8 ± 5.9% with an r =.99. However, with the exclusion of 3 subjects with body fat <7%, the within-day and between-day CV decreased to 2.0 ± 1.8% and 2.7 ± 2.1%, respectively. The difference between the first two trials of each day was < 2% fat in 93% of the cases, and all 150 cases resulted in fat within + 2% in the three trials. In conclusion, these results support that BOD POD measurements of BV, Vtg, and % fat are reliable both within and between days for a wide range of adults. Also, to obtain the most reliable measurement of % fat in the BP, measuring thoracic gas volume and averaging the values obtained from two trials which agree within + 2 % fat is recommended.
School of Physical Education

Blood pressure (BP) measurement reliability, year-to-year BP tracking, distinguishing characteristics of upper quintile (UQ) vs lower four quintiles' (LQ) systolic BP (SBP) tracking and the relationships of fat distribution and body composition to SBP were examined in 57 youth. Subjects were measured on two occasions approximately one year apart. Longitudinal measures included auscultatory BPs, height, weight, body circumferences, skeletal widths, bioelectrical impedance and skinfolds. Inter-trial reliability of right/left arm averaged BP (RLBP) exceeded that of either limb alone; tracking magnitude was likewise greater with RLBP. Greater total body mass and fatness as well as larger anthropometric dimensions distinguished UQ from LQ SBP trackers. Fat distribution and SBP were not consistently associated with each other across study years. Irrespective of gender differences, fatness and fat free mass per unit height2 were independently related to within year SBP, yet only initial fatness was independently predictive of future SBP.

The measurement of total body volume (V) (excluding lung volume) together with total body mass (m) is required in order to determine body density (d = m/V). From this, and using certain simplifying assumptions, it is possible to derive body composition in terms of fat mass (FM) and fat free mass (FFM) for the two-compartment model. The standard method for determining body volume (and hence body composition) is the densitometric (underwater weighing) technique based on Archimedes' principle. Three variables, notably residual lung volume (RV), total body mass (m) and submerged body mass are measured. RV is normally determined using a gas dilution technique while total body mass is simply measured using an accurate weighing scale. The submerged body mass is measured while the subject is totally submerged in a tank of water. This method, although relatively accurate, requires substantial apparatus and is time consuming. An alternative method, based on a polytropic thermodynamic process, is described for body volume measurement and thereby for body composition assessment. Previous use of this method by Taylor, et al. (1985) and Gundlach and Visscher (1986) were successful, but complex in terms of operating system. The described system comprises of a Perspex, sealed chamber. A cycling piston communicates with the chamber and imposes a minute sinusoidal pressure variation which is then measured. With a subject situated inside the chamber an increased pressure variation, caused by the decreased chamber volume, is then measured and processed to yield the displaced, or body volume. Subject comfort, above all, is greatly enhanced, in comparison to the underwater weighing method. A substantial advantage of the method appears to be that RV need no longer be measured. Variables such as a rise of temperature and humidity caused by the subject, as well as pressure variations due to respiration, were expected and found. These were analyzed both theoretically and experimentally and where necessary the data were modified to account for these variables using a personal computer. Calibration and preliminary validation of the instrument has been carried out using underwater weighing, bioimpedance and skinfold analyses and the error of measurement assessed. It appears that the described plethysmographic method is capable of measuring body volume and thus compares favourably to the underwater weighing method. Even though other groups have succeeded in employing similar principles, a substantially simpler mechanism has been used here.

This study compared three indirect calorimetry determinations, as kcals/minute, over three consecutive days on 28 healthy, sedentary women of varying body composition. No significant within-individual variation for VO2, CO2, respiratory quotient (RQ), or Kcals/minute was found among the three days. A low coefficient of variation (3.4 +/- 3%) and a relatively small standard deviation in mean Kcals/day (1383 +/- 214) suggests possibly one or only a few measures are necessary for predicting resting metabolic rate (RMR) within a range applicable for clinical use. There were strong correlations of body weight and body composition variables (fat and LBM) with RMR. Knowing both LBM and fat mass increased the ability to predict RMR significantly over the prediction with either variable alone. The Harris Benedict equation over-predicted RMR by 11.1% compared to RMR measured by indirect calorimetry. When equations are based on body weight, rather than LBM, metabolic rate may be over-predicted in obese populations.

Thesis (Ph. D.)--Florida State University, 2009.
Advisor: Jasminka Z. Ilich-Ernst, Florida State University, College of Human Sciences, Dept. of Nutrition, Food and Exercise Sciences. Title and description from dissertation home page (viewed on Nov. 9, 2009). Document formatted into pages; contains viii, 76 pages. Includes bibliographical references.

The prevalence of obesity has increased worldwide at an alarming rate over the past four decades. Obesity is a major risk factor for various non-communicable diseases including cardiovascular disease, type 2 diabetes mellitus and certain types of cancer. In parallel with the global obesity epidemic, rates of maternal obesity and gestational diabetes mellitus have risen, both of which are associated with long-term adverse health outcomes for the affected offspring. There is a robust body of epidemiological evidence, supported by experimental animal models, that fetal growth has a profound influence on adult diseases. For instance, the environment to which a fetus is exposed in the womb contributes to long-term risks of cardiovascular disease in adulthood. Both fetal growth restriction (predominantly due to poor nutrition resulting from placental insufficiency) and excessive fetal growth (predominantly due to over nutrition resulting from GDM or maternal obesity) are associated with higher risk of cardiovascular disease during adulthood. Birth weight is a widely used marker of fetal growth, albeit crude. Measures of newborn body composition, such as fat mass, may identify ‘high risk’ infants more accurately. Postmenopausal women with obesity, are at a particularly high risk of cardiovascular disease. While on average obesity is associated with a greater severity of atherosclerosis and poorer arterial function, many postmenopausal women with obesity will never be affected by atherosclerosis-related cardiovascular disease. A better understanding of the risk factors for atherosclerosis in postmenopausal women with obesity will contribute to improved prediction of high-risk people and enable more effective disease management. Aims The studies undertaken as part of this thesis aimed to improve early identification of cardiovascular risk on the basis of advanced measures of body composition, and to trial preventive strategies for reducing the burden of atherosclerosis. Non-invasive measures of vascular health that enabled age-appropriate assessment of atherosclerosis across the life course was applied in order to achieve the following specific objectives. To examine whether a low Glycaemic Index (GI) diet during pregnancy could improve vascular health of the offspring of women with a high risk of gestational diabetes; To determine whether body composition in newborn infants, especially body fatness of neonates, is a stronger predictor of arterial wall thickness than birth weight; To determine the predictors of vascular structure in postmenopausal women with obesity; To determine theeffects of fast vsslow weight loss intervention strategies on vascular function. Results Chapter 3,published in the American Journal of Clinical Nutrition,reports the vascular sub-study of a randomized controlledtrial of a low-GI diet during pregnancy in women at high risk of gestational diabetes mellitus (GDM). The low-GI diet improved offspring arterial wall thickness at 1-year of age, independent of birth weight, body fatness and growth during infancy. This suggests that the vascular effects of a low maternal GI diet are acting via mechanisms that do not involve fetal growth. Chapter 4, published in Journal of Clinical Medicine,reports aortic intima-media thickness (AoIMT), an age-appropriate marker of vascular health, in newborns with high or low body fat, compared to those with average body fat. Infants with low body fat had higher AoIMT than controls, butinfants with high body fat showed no differences in AoIMT compared to controls. In contrast, AoIMT was significantly higher in those babies born at either end of the birth weight spectrum. The association of birth weight £10th percentile with AoIMT was independent of body fatness, but the association of birth weight >90thpercentile with AoIMT was weakened after controlling for body fatness.Accordingly, this suggests thatnewborns with low body fat have increased AoIMT, which is consistent with a higher risk of future cardiovascular disease, however birth weight appears to be a stronger predictor of aortic IMT. Chapter 5reports an analysis ofthe best predictors of atherosclerosisin postmenopausal women with obesity. We demonstrated that waist circumference (WC), subcutaneous adipose tissue(SAT) and intrahepatic lipid content were significantly associated with carotid intima-media thickness (CIMT) after controlling for age, lipid lowering medications, diastolic blood pressure and BMI. This suggests that waist circumference may represent a cost-effective measure for identifying those postmenopausal women with obesity who are at the highest risk of atherosclerosis-related cardiovascular disease. Finally,Chapter 6reports the vascular sub-study of a randomized controlled trial of weight loss in postmenopausal women with obesity. The trial examined the effects on arterial endothelial function measured by flow mediated dilatation (FMD), after 4-6 months on one of two intervention strategies; fast weight loss (very low energy diet-based diet), and slow weight loss (food-based diet). Both interventions produced meaningful weight loss, although greater weight loss was achieved at both 4 and 6 months in the fast weight loss intervention group. Only the fast weight loss group showed improvements in FMD. The strongest correlate of improvements in FMD over 4-6 months was concurrent reduction in VAT, suggesting that changes in this adipose tissue depot may mechanistically link weight loss with improvements in atherosclerotic cardiovascular disease. Conclusion The studies presented in this thesis suggest that body composition hasa profound influence on vascular health from birth through adulthood. Intervention strategies aimed at reducing cardiovascular risk should begin in the womb and focus on strategies that produce vascular health benefits. In order to treat obesity and cardiovascular disease, further research should take a holistic approach to integrate various evidences among different populations at risks with a focus on early detection and prevention. Keywords:atherosclerosis, body composition, fetal growth, obesity, vascular health

OBJECTIVE: The purpose of this study was to examine the relationship between strength training and body composition before and after controlling for several covariates. A cross-sectional study including 257 female subjects was conducted. METHODS: Subjects' level of involvement in strength training was determined via questionnaire. Body composition was assessed using dual energy X-ray absorptiometry (DXA). Diet was assessed using 7-d weighed food records. RESULTS: Strong linear relationships between subjects' level of involvement in strength training and body composition were identified. For each additional day of strength training reported per week, body fat was 1.32 percentage points lower (F = 14.8, p = 0.0002) and fat-free mass was 656.4 g (1.45 lb) higher (F = 18.9, p < 0.0001), on average. Likewise, the more time subjects spent lifting and the more intensely they trained, the better their body composition tended to be. Adjusting for differences in age, menopause status, objectively measured physical activity, energy intake, and protein intake tended to weaken each association. Controlling for differences in physical activity weakened each relationship the most. CONCLUSION: Women who strength train regularly tend to have significantly lower body fat percentages and significantly higher levels of fat-free mass compared to their counterparts, regardless of differences in several potential confounding variables.

There is an increase in the prevalence of obesity among people with schizophrenia thought to be due in part to the weight enhancing side-effects of medications commonly used to treat the symptoms of schizophrenia. Despite the deleterious health effects associated with obesity and its impact on quality of life and medication compliance, little is known about body composition and energy expenditure in this clinical group. The primary purpose of this thesis was to enhance understanding of body composition and energy expenditure, particularly resting energy expenditure in men with schizophrenia who take atypical antipsychotic medications. Unique to this investigation is the evaluation of clinical tools used to predict body composition and energy expenditure against reference methodologies in men with schizophrenia. Further, given the known links between obesity and physical activity, an additional but less comprehensive component of the thesis was a consideration of total and activity energy expenditure in addition to the interaction between psychiatric symptoms, side-effects of antipsychotic medications and physical activity also occurred as part of this thesis. Collectively, the goals of this thesis were addressed through a series of studies – the first two studies were related to the measurement and characteristics of body composition in men with schizophrenia, while the third and fourth studies were related to the measurement and characteristics of resting energy expenditure in men with schizophrenia. The fifth and sixth studies the utilised doubly labelled water technique to quantify activity and total energy expenditure in a small group of men with schizophrenia and explored the use of accelerometry in this cohort. The final study briefly considered the impact of psychiatric symptoms and self-reported medication side-effects on objectively measured physical activity. In the first study, thirty-one male adults previously diagnosed with schizophrenia and sixteen healthy male controls were recruited. Estimates of body composition derived from an anthropometry-based equation and from bioelectric impedance analysis (BIA) using deuterium dilution as the reference methodology to determine total body water were compared. The study also determined the validity of equations commonly used to predict body composition from BIA in the men with schizophrenia. A further aim was to determine the superiority of either BIA or body mass index (BMI) as an indicator of obesity in this cohort. The inclusion of the control group, closely matched for age, body size and body composition demonstrated that there was no difference in the ability of body composition prediction methods to distinguish between fat and fat-free mass (FFM) in controls and men with schizophrenia when both groups had similar body composition. However this study indicated that an anthropometry-based equation previously used in people with schizophrenia was a poor predictor of body composition in this cohort, as evidenced by wide limits of agreement (25%) and systematic variation of the bias. In comparison, the best predictor of percentage body fat (%BF) in this group was gained when impedance values were used to predict percentage body fat via the equation published by Lukaski et al (1986). Although percentage body fat was underpredicted using the Lukaski et al. (1986) equation, the mean magnitude was relatively small (1.3%), with the limits of agreement approximately 13%. Linear regression analysis revealed that %BF predicted using the Lukaski et al. (1986) equation explained 25% more of the variance in percentage body fat than BMI. Further, this study also indicated that BIA was more sensitive than BMI in distinguishing between overweight and obesity in this cohort of men with schizophrenia. Because of the almost exclusive use of BMI as an indicator of obesity in people with schizophrenia, the level of excess body fat may be in excess of that previously indicated. The second study extended the examination of body composition in men with schizophrenia. In this study, the thirty-one participants with schizophrenia (age, 34.2 ± 5.7 years; BMI, 30.2 ± 5.7 kg/m2) were individually matched with sedentary controls by age, weight and BMI. Deuterium dilution was used to distinguish between FFM and fat mass. The previous study had indicated that while BIA was a suitable group measure for obesity, on an individual level the technique lacked the precision required for investigating body composition in men with schizophrenia. Waist circumference was used as an indicator of body fat distribution. The findings of this study indicated that in comparison with healthy sedentary controls of similar body size and age, men with schizophrenia had higher levels of body fat which was more centrally distributed. Percentage body fat was on average 4% higher and waist circumference, on average 5 cm greater in men with schizophrenia than the sedentary controls of the same age and BMI. Further, this study indicates that the use of BMI to predict body fat in men with schizophrenia will result in greater bias than when it is used to predict body fat in other sedentary men. Commonly used regression equations to predict energy requirements at rest are based on the relationships between weight and resting energy expenditure (REE) and in such equations, weight acts as a surrogate measure of FFM. The objectives of study three were to measure REE in a small group of men with schizophrenia who were taking the antipsychotic medication clozapine and to determine whether REE can be predicted with sufficient accuracy to substitute for the measurement of REE in the clinical and/or research settings. Body composition was determined using deuterium dilution and REE was measured using a Deltatrac Metabolic Cart via a ventilated hood. The male participants, (aged 28.0 ± 6.7 yrs, BMI 29.8 ± 6.8 kg/m2) were weight stable at the time of the study and had been taking clozapine for 20.5 ± 12.8 months, with doses of 450 ± 140 mg/day. Of the six prediction equations evaluated, the equation of Mifflin et al. (1990) with no systematic bias, the lowest bias and the lowest limits of agreement proved to be the most suitable equation to predict REE in this cohort. The overestimation of REE can be corrected for by deducting 160 kcal/day from the predicted REE value when using the Mifflin et al. (1990) equations. However, the magnitude of the error associated with the prediction of REE for an individual is 370 kcal/day. The findings of this study indicate that REE cannot be predicted with sufficient individual accuracy in men with schizophrenia, therefore it was necessary to measure rather than predict REE in subsequent studies. In the fourth study, indirect calorimetry (Deltatrac Metabolic Cart via ventilated hood) and deuterium dilution were used to accurately determine REE, respiratory quotient (RQ) and FFM in 31 men with schizophrenia and healthy sedentary controls individually matched for age and BMI. Data from this study indicated that gross REE was lower in men with schizophrenia than in healthy sedentary controls of a similar age and body size. However, there was no difference between the groups in REE when REE was adjusted for FFM using the mathematically correct method (analysis of covariance with FFM as the covariate). There was however a statistically and clinically significant difference in resting, fasted RQ between men with schizophrenia and controls, suggesting that RQ rather than REE may be an important correlate worthy of further investigation in men with schizophrenia who take antipsychotic medications. Studies five and six involved the application of the doubly labelled water (DLW) technique to accurately determine total energy expenditure (TEE) and activity energy expenditure (AEE) in a small group of men with schizophrenia who had been taking the atypical antipsychotic medication clozapine. The participants were those who took part in study three. The purpose of these studies was to assess the validity of a commercially available tri-axial accelerometer (RT3) for predicting free-living AEE and to investigate TEE and AEE in men with schizophrenia. There was poor agreement between AEE measured using DLW and AEE predicted using the RT3. However, using the RT3 to measure inactivity explained over two-thirds of the variance in AEE. This study found that the relationship between current AEE per kilogram of body weight and change from baseline weight in men taking clozapine was strong although not significant. The sedentary nature of the group of participants in this study was reflected in physical activity levels, (PAL, 1.39 ± 0.27), AEE (435 ±352 kcal/day) and TEE (2511 ± 606 kcal/day) that fell well short of values recommended by WHO (2000) for optimal health and to prevent weight gain. Given the increasing recognition of the importance of sedentary behaviour to weight gain in the general community, further examination of the unique contributing factors such as medication side effects and symptoms of mental illness to activity levels in this clinical group is warranted. The final study used accelerometry (RT3) to objectively measure activity in a group of 31 men with schizophrenia who had been taking atypical antipsychotic medications for more than four months. The purpose of this study was to explore the relationships between psychiatric symptomatology, side-effects of medication and physical activity. Accelerometry output was analysed to provide a measure of inactivity and moderate intensity activity (MIA). The well-validated and reliable standardised clinical interview, the Positive and Negative Syndrome Scale (PANSS) was used as a measure of psychiatric symptoms. Perceived side-effects of medication were assessed using the Liverpool University Neuroleptic Rating Side-Effects Scale (LUNSER). Surprisingly, there was no relationship reported between any measures of negative symptoms and physical inactivity. However, self-reported measures of medication side-effects relating to fatigue, sleepiness during the day and extrapyramidal symptoms explained 40% of the variance in inactivity. This study found significant relationships between some negative symptoms and moderate intensity activity. Despite the expectation that as symptoms of mental illness reduce, inactivity may diminish and moderate intensity activity will increase, it may not be surprising that in practice this is an overly simplistic view. It may be that measures of social functioning and possibly therefore cognition may be better predictors of physical activity than psychiatric symptomatology per se.

There is an increase in the prevalence of obesity among people with schizophrenia thought to be due in part to the weight enhancing side-effects of medications commonly used to treat the symptoms of schizophrenia. Despite the deleterious health effects associated with obesity and its impact on quality of life and medication compliance, little is known about body composition and energy expenditure in this clinical group. The primary purpose of this thesis was to enhance understanding of body composition and energy expenditure, particularly resting energy expenditure in men with schizophrenia who take atypical antipsychotic medications. Unique to this investigation is the evaluation of clinical tools used to predict body composition and energy expenditure against reference methodologies in men with schizophrenia. Further, given the known links between obesity and physical activity, an additional but less comprehensive component of the thesis was a consideration of total and activity energy expenditure in addition to the interaction between psychiatric symptoms, side-effects of antipsychotic medications and physical activity also occurred as part of this thesis. Collectively, the goals of this thesis were addressed through a series of studies – the first two studies were related to the measurement and characteristics of body composition in men with schizophrenia, while the third and fourth studies were related to the measurement and characteristics of resting energy expenditure in men with schizophrenia. The fifth and sixth studies the utilised doubly labelled water technique to quantify activity and total energy expenditure in a small group of men with schizophrenia and explored the use of accelerometry in this cohort. The final study briefly considered the impact of psychiatric symptoms and self-reported medication side-effects on objectively measured physical activity. In the first study, thirty-one male adults previously diagnosed with schizophrenia and sixteen healthy male controls were recruited. Estimates of body composition derived from an anthropometry-based equation and from bioelectric impedance analysis (BIA) using deuterium dilution as the reference methodology to determine total body water were compared. The study also determined the validity of equations commonly used to predict body composition from BIA in the men with schizophrenia. A further aim was to determine the superiority of either BIA or body mass index (BMI) as an indicator of obesity in this cohort. The inclusion of the control group, closely matched for age, body size and body composition demonstrated that there was no difference in the ability of body composition prediction methods to distinguish between fat and fat-free mass (FFM) in controls and men with schizophrenia when both groups had similar body composition. However this study indicated that an anthropometry-based equation previously used in people with schizophrenia was a poor predictor of body composition in this cohort, as evidenced by wide limits of agreement (25%) and systematic variation of the bias. In comparison, the best predictor of percentage body fat (%BF) in this group was gained when impedance values were used to predict percentage body fat via the equation published by Lukaski et al (1986). Although percentage body fat was underpredicted using the Lukaski et al. (1986) equation, the mean magnitude was relatively small (1.3%), with the limits of agreement approximately 13%. Linear regression analysis revealed that %BF predicted using the Lukaski et al. (1986) equation explained 25% more of the variance in percentage body fat than BMI. Further, this study also indicated that BIA was more sensitive than BMI in distinguishing between overweight and obesity in this cohort of men with schizophrenia. Because of the almost exclusive use of BMI as an indicator of obesity in people with schizophrenia, the level of excess body fat may be in excess of that previously indicated. The second study extended the examination of body composition in men with schizophrenia. In this study, the thirty-one participants with schizophrenia (age, 34.2 ± 5.7 years; BMI, 30.2 ± 5.7 kg/m2) were individually matched with sedentary controls by age, weight and BMI. Deuterium dilution was used to distinguish between FFM and fat mass. The previous study had indicated that while BIA was a suitable group measure for obesity, on an individual level the technique lacked the precision required for investigating body composition in men with schizophrenia. Waist circumference was used as an indicator of body fat distribution. The findings of this study indicated that in comparison with healthy sedentary controls of similar body size and age, men with schizophrenia had higher levels of body fat which was more centrally distributed. Percentage body fat was on average 4% higher and waist circumference, on average 5 cm greater in men with schizophrenia than the sedentary controls of the same age and BMI. Further, this study indicates that the use of BMI to predict body fat in men with schizophrenia will result in greater bias than when it is used to predict body fat in other sedentary men. Commonly used regression equations to predict energy requirements at rest are based on the relationships between weight and resting energy expenditure (REE) and in such equations, weight acts as a surrogate measure of FFM. The objectives of study three were to measure REE in a small group of men with schizophrenia who were taking the antipsychotic medication clozapine and to determine whether REE can be predicted with sufficient accuracy to substitute for the measurement of REE in the clinical and/or research settings. Body composition was determined using deuterium dilution and REE was measured using a Deltatrac Metabolic Cart via a ventilated hood. The male participants, (aged 28.0 ± 6.7 yrs, BMI 29.8 ± 6.8 kg/m2) were weight stable at the time of the study and had been taking clozapine for 20.5 ± 12.8 months, with doses of 450 ± 140 mg/day. Of the six prediction equations evaluated, the equation of Mifflin et al. (1990) with no systematic bias, the lowest bias and the lowest limits of agreement proved to be the most suitable equation to predict REE in this cohort. The overestimation of REE can be corrected for by deducting 160 kcal/day from the predicted REE value when using the Mifflin et al. (1990) equations. However, the magnitude of the error associated with the prediction of REE for an individual is 370 kcal/day. The findings of this study indicate that REE cannot be predicted with sufficient individual accuracy in men with schizophrenia, therefore it was necessary to measure rather than predict REE in subsequent studies. In the fourth study, indirect calorimetry (Deltatrac Metabolic Cart via ventilated hood) and deuterium dilution were used to accurately determine REE, respiratory quotient (RQ) and FFM in 31 men with schizophrenia and healthy sedentary controls individually matched for age and BMI. Data from this study indicated that gross REE was lower in men with schizophrenia than in healthy sedentary controls of a similar age and body size. However, there was no difference between the groups in REE when REE was adjusted for FFM using the mathematically correct method (analysis of covariance with FFM as the covariate). There was however a statistically and clinically significant difference in resting, fasted RQ between men with schizophrenia and controls, suggesting that RQ rather than REE may be an important correlate worthy of further investigation in men with schizophrenia who take antipsychotic medications. Studies five and six involved the application of the doubly labelled water (DLW) technique to accurately determine total energy expenditure (TEE) and activity energy expenditure (AEE) in a small group of men with schizophrenia who had been taking the atypical antipsychotic medication clozapine. The participants were those who took part in study three. The purpose of these studies was to assess the validity of a commercially available tri-axial accelerometer (RT3) for predicting free-living AEE and to investigate TEE and AEE in men with schizophrenia. There was poor agreement between AEE measured using DLW and AEE predicted using the RT3. However, using the RT3 to measure inactivity explained over two-thirds of the variance in AEE. This study found that the relationship between current AEE per kilogram of body weight and change from baseline weight in men taking clozapine was strong although not significant. The sedentary nature of the group of participants in this study was reflected in physical activity levels, (PAL, 1.39 ± 0.27), AEE (435 ±352 kcal/day) and TEE (2511 ± 606 kcal/day) that fell well short of values recommended by WHO (2000) for optimal health and to prevent weight gain. Given the increasing recognition of the importance of sedentary behaviour to weight gain in the general community, further examination of the unique contributing factors such as medication side effects and symptoms of mental illness to activity levels in this clinical group is warranted. The final study used accelerometry (RT3) to objectively measure activity in a group of 31 men with schizophrenia who had been taking atypical antipsychotic medications for more than four months. The purpose of this study was to explore the relationships between psychiatric symptomatology, side-effects of medication and physical activity. Accelerometry output was analysed to provide a measure of inactivity and moderate intensity activity (MIA). The well-validated and reliable standardised clinical interview, the Positive and Negative Syndrome Scale (PANSS) was used as a measure of psychiatric symptoms. Perceived side-effects of medication were assessed using the Liverpool University Neuroleptic Rating Side-Effects Scale (LUNSER). Surprisingly, there was no relationship reported between any measures of negative symptoms and physical inactivity. However, self-reported measures of medication side-effects relating to fatigue, sleepiness during the day and extrapyramidal symptoms explained 40% of the variance in inactivity. This study found significant relationships between some negative symptoms and moderate intensity activity. Despite the expectation that as symptoms of mental illness reduce, inactivity may diminish and moderate intensity activity will increase, it may not be surprising that in practice this is an overly simplistic view. It may be that measures of social functioning and possibly therefore cognition may be better predictors of physical activity than psychiatric symptomatology per se.

Gentamicin is an aminoglycoside antibiotic that possesses bactericidal activity against many gram-positive and gram-negative organisms. Clinically, it is used most often to treat life-threatening infections due to Pseudomonas, Proteus, and the Klebsiella-Enterobacter group. A relationship between gentamicin serum concentrations and clinical response has been demonstrated. Toxicities, notably ototoxicity and nephrotoxicity, are also associated with serum concentrations. Gentamicin is given intermittently either intramuscularly or intravenously resulting in peak and trough concentrations. The therapeutic range is defined as peak concentrations between 4-15mg/L (depending in part on the site of infection and the susceptibility of the infecting organism), and trough concentrations less than 2mg/L (to minimize toxicity). Gentamicin distributes into a space similar to the extracellular fluid volume (ECFV). Pathophysiologic changes which alter the extracellular fluid compartment also alter gentamicin volume of distribution (Vd). One intrinsic factor known to alter gentamicin Vd is obesity. Leanness is also thought to alter gentamicin Vd but its effect has not been quantitated. The objectives of this study were to: 1) accurately describe a Vd in "normal" patients, that is, those with no factors known to alter gentamicin volume of distribution; 2) determine if there is a continuous linear relationship between gentamicin volume of distribution (L/kg) and percent body fat; 3) determine if that relationship is associated with changes in ECFV; and 4) develop a formula for predicting Vd in a similar patient population. Twenty patients with no extrinsic factors known to alter gentamicin Vd participated in the study. Five blood samples were drawn around one steady state dose of gentamicin. A one-compartment model was used to calculate Vd. Tritiated water and anthropometric measurements were conducted simultaneously to provide estimates of body composition. Together these values were used to examine the relationship between gentamicin Vd and body composition. We have described a Vd for gentamicin that is larger but no less variable than is currently used to determine initial dosage regimens. This volume may be larger either due to the selection of patients or method of serum gentamicin analysis. This larger volume should be used to calculate empiric dosage regimens for similarly selected patients to decrease the risk of treatment failure. We were not able to describe a linear relationship between percent body fat and gentamicin volume of distribution. We have postulated several reasons as to why this relationship could not be detected; 1) the sample size may not have been large enough, 2) the relationship is not important in patients who are not at extremes of weight, or 3) the variations caused by changes in body composition were not as significant as other factors that may cause fluid alterations in hospitalized patients. There was a strong correlation between gentamicin Vd and total body water noted. Having eliminated all patients in whom the relationship between total body water and ECFV could not be assumed to be normal and constant, we have indirectly demonstrated a strong relationship between ECFV and gentamicin Vd. This relationship still leaves variability in gentamicin's distribution characteristics to be explained. The predictive formula is based on measurements of height, weight, and a larger Vd [L/kg(ideal body weight)] than has previously been used. The predictive formula recommended for clinical use in adults is Vd=0.30L/kg (Dosing Weight). Dosing weight equals ideal body weight (IBW) when actual body weight (ABW) is ≤ IBW, or 0.4(ABW-IBW)+IBW, when ABW is > IBW. The consequences of estimating a larger Vd are that patients empirically would receive larger doses than are currently being administered, thus more patients should obtain therapeutic serum concentrations within the first 24 hours of therapy. This information will be useful in our attempts to optimize gentamicin therapy.
Pharmaceutical Sciences, Faculty of
Graduate