¹Department of Endocrinology, Hvidovre University Hospital, Hvidovre, Denmark

²Department of Internal Medicine, Amager Hospital, Copenhagen, Denmark

³Department of Clinical Physiology and Nuclear Medicine, Hvidovre University Hospital, Hvidovre, Denmark

⁴Department of Clinical Physiology and Nuclear Medicine, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark

⁵Roskilde County Hospital, Roskilde, Denmark

Correspondence to: A Gotfredsen, Department of Internal Medicine, Amager Hospital, Italiensvej 1, DK-2300 København S, Denmark. E-mail: anders.gotfredsen@ah.hosp.dk

OBJECTIVE: To investigate the influence of the pancreas lipase inhibitor orlistat (OLS) on calcium metabolism, bone turnover, bone mass, bone density and body composition when given for obesity as adjuvant to an energy- and fat-restricted diet.

DESIGN: Randomized controlled double-blinded trial of treatment with OLS 120 mg three times daily or placebo for 1 y.

SUBJECTS: Thirty obese subjects with a mean body mass index (BMI) of 36.9±3.7 kg/m² and a mean age of 41±11 y. Sixteen patients were assigned to OLS and 14 to placebo.

MEASUREMENTS: Dual energy X-ray absorptiometry (DXA) measurements of bone mineral and body composition included total bone mineral content (TBMC), total bone mineral density (TBMD), lumbar spine BMC and BMD, forearm BMC and BMD, fat mass (FM), fat free-mass (FFM), percentage fat mass (FM%) as well as a DXA estimate of the body weight. Body composition (FM, FFM and FM%) was estimated by total body potassium (TBK). Indices of calcium metabolism and bone turnover included serum values of ionized calcium (Ca⁺⁺), iPTH (parathyroid hormone), alkaline phosphatase, 25(OH)-vitamin D, 1,25(OH)₂ vitamin D and osteocalcin as well as fasting urinary ratios of hydroxyproline/creatinine and Ca/creatinine (fU-OHpr/creat, fUCa/creat).

RESULTS: There were no significant differences between OLS and placebo groups as to any of the body composition variables (FFM, FM, FM%) at baseline or after 1 y treatment. Weight loss was of 11.2±7.5 kg in the OLS group and 8.1±7.5 kg in the placebo group (NS). The changes in FM and FM% were significant in both groups determined by DXA as well as by TBK, but the group differences between these changes were not significant. The composition of the weight loss was approximately 80% fat in both groups. FFM only changed significantly by DXA in the OLS group (-1.3 kg), but the difference from the placebo group was not significant. Forearm BMD in both groups, forearm BMC in the OLS group and TBMD in the placebo group fell discretely but significantly, but there were no significant group differences between the OLS and the placebo-treated group. All biochemical variables except s-osteocalcin changed significantly after 1 y in the OLS group, disclosing a pattern of an incipient negative vitamin D balance, a secondary increase in PTH-secretion, and an increase in bone turnover with the emphasis on an increase in resorption parameters (fU-OHpr/creat, fUCa/creat). In the placebo group, only s-25(OH)vitamin D and fU-OHpr/creat changed significantly, but the pattern was also that of a deteriorated vitamin D status and an increase in PTH levels and bone turnover. The only biochemical variable which was significantly different between OLS and placebo groups after one year was the fU-OHpr/creat ratio, which increased from 12.0 to 20.1 in the OLS group but only from 10.9 to 13.2 in the placebo group.

CONCLUSION: One year's treatment with OLS induces a lipid malabsorption which enhances a dietary weight loss without any significant deleterious effects on body composition. OLS induces a relative increase in bone turnover in favour of resorption, possibly due to malabsorption of vitamin D and/or calcium. However, no changes in bone mass or density are seen after 1 y of OLS treatment apart from those explained by the weight loss itself. Thus 1 y of OLS treatment seems safe from a 'bone preserving' point of view. A vitamin D and calcium supplement should be taken during the treatment.

International Journal of Obesity (2001) 25, 1154-1160

obesity treatment; orlistat; body composition; bone turnover; bone mass; bone mineral content; bone mineral density

Introduction

Orlistat (OLS) is a recently introduced pharmacological treatment of obesity.¹ It promotes weight loss by reducing fat uptake from the intestine through a partial pancreas lipase inactivation which causes a 30% malabsorption of dietary fat.

Obesity influences calcium metabolism and bone status. Obese subjects most frequently have increased bone mass and bone density,^2,3,4,5,6,7 which is known to fall towards the level of non-obese subjects after weight loss.^3,4,8,9 The pathophysiological mechanism for this is not known but may be related to some humoral mediator, eg changes in parathyroid function.^10,11,12 Morbid obesity was in one study associated with secondary hyperparathyroidism, which normalized with dietary weight loss.¹³ Also the excess amount of fat tissue may stimulate bone formation through the mechanical load on the skeleton.

It is known from various clinical conditions that fat malabsorption can be associated with a significant intestinal loss of dietary calcium and vitamin D which may lead to osteomalacia. Fat malabsorption as a treatment for obesity was first introduced through jejunoileal bypass operations, and bypass patients were reported to be at risk of vitamin D hypovitaminosis, hypocalcemia, secondary hyperparathyroidism and osteomalacia.^3,14,15,16

Fat malabsorption induced by OLS is gererally more limited than that induced by jejunoileal bypass surgery. As OLS in other aspects may prove safe for long-term use, it is of interest to investigate possible changes in calcium metabolism and bone mineral content during long-term treatment of obesity with OLS.

Treating obesity, the aim is to reduce fat mass selectively and to spare not only bone tissue but lean body mass as a whole. Thus, the composition of the weight lost is crucial and therefore it is of interest to determine changes in body composition in relation to new weight loss regimens and drugs.

On this background the present study compared calcium metabolic variables, bone mineral content and composition of the body mass lost during 1 y of treatment of obesity with either OLS or placebo as supplement to the recommended energy and fat restricted diet.

Patients

Patients included in the present study represented a subsample of the patients resently reported on by Sjöström et al.¹ Criteria for entry were age >18 y, body mass index (BMI) between 30 and 43 kg/m², and a motivation for weight loss. All the subjects had been obese for several years, but were otherwise healthy as assessed from history, physical examination, routine blood tests, abdominal ultrasound examination and ECG. None were taking any medication known to influence fat-free mass, fat mass or calcium metabolism. Women with a childbearing potential were included if they were using adequate non-hormonal contraception.

The Copenhagen center included a total of 48 patients in the main protocol. The present analysis comprises all patients (n=30, age 41±11 y, BMI=36.9±3.7 kg/m²) who completed the first year of the study. All patients were given a hypocaloric diet (600 kcal/day deficit) with 30% of the energy from fat. Sixteen patients were randomized to treatment with OLS 120 mg three times daily and 14 patients had placebo. Clinical data are shown in Table 1. Initially there were no significant differences between the OLS and placebo groups regarding age, height, weight and BMI. There was a non-significant tendency towards a greater height and weight in the OLS group, partly due to the overrepresentation of men (three in the OLS group, one in the placebo group). The different number of subjects in the two groups (16 in the OLS group, 14 in the placebo group) is due to the fact that this study is a substudy of a larger multicenter randomized controlled trial.

Methods

Bone mineral content (BMC) and density (BMD) were measured by dual energy X-ray absortiometry (DXA) using a Norland XR36 densiometer. BMC and BMD were measured in the total body (TBMC and TMBD) as well as in the distal forearm (BMC_fa and BMD_fa) and in the lumbar spine (BMC_ls and BMD_ls).^17,18,19

Body composition was measured by two methods: (1) dual energy X-ray absorptiometry (DXA) using the above-mentioned Norland XR36 densitometer. This apparatus determines body weight, fat-free mass (FFM) (which includes the bone mineral content), and fat mass (FM).^17,18,19,20 (2) Measurement of total body potassium (TBK) by counting the naturally occurring ⁴⁰K in a whole body counter.^20,21 TBK has a constant ratio to the FFM.

Body weight was measured to the nearest 0.1 kg using a precision hospital scales. The precision (CV%) for the DXA BMC and BMD measurements was 1-2%, and the accuracy (SEE%) 3-6%.^17,19,20 The precision (CV%) for DXA FFM, FM and FM% was 2-3%, and the accuracy (SEE%) was 4-15%.^18,19,20 The precision (CV%) from the TBK measurement was 2.7% for FFM, 5.0% for FM, and 1.73% for FM%.²⁰

All DXA measurements were performed before drug treatment and after 1 y of treatment, ie 1 month after run-in and after 11 months treatment with either OLS or placebo. Specimens for the biochemical analyses were obtained with the same interval.

Bone metabolism was evaluated from serum measurements of ionized calcium, total alkaline phosphatase and iPTH, as well as 1,25(OH)₂D₃ by RIA (CV=14%),²² 25(OH)D₂+D₃ by competitive proteine binding (CBP) (CV=10%),²³ osteocalcin (=BGP) by RIA (CV=6.5%),²³ fasting urinary hydroxyproline/creatinine ratio^24,26 and fasting urinary calcium/creatinine ratio.^26,27 All specimens were kept frozen at -18°C until analysis, which was performed blinded with specimens in random order.

Statistics

Data were calculated as mean values and their confidence intervals (CI). Differences between treatment groups were evaluated by the Student's t-test for unpaired data. Differences within treatment groups between start and end of the study were evaluated by the Student's t-test for paired data. Relationships between variables and between changes in variables were evaluated by standard linear correlation and regression analysis. The level for statistical significance was P<0.05.

Ethics

Before entering the study all participants were informed about the nature and the purpose of the study and all gave their informed consent. The study protocol was approved by the Ethical Committee for Copenhagen and it complied with Helsinki Declaration II of 1975 as revised in 1983. The patient data file was approved by the Danish Data Protection Agency.

Results

Table 2 shows the results of the body composition analysis by TBK and DXA at the beginning and at the end of the study for the two groups (OLS and placebo). Body weights given together with the TBK values are those obtained with hospital scales. Body weights given together with the DXA values are those obtained with the DXA machine itself. There were no significant group differences regarding any of the variables, either before or after the study. There were tendencies towards small differences between the weights by DXA and by scales (from 1.1 to 1.4 kg) which were, however, not statistically significantly different. Table 3 shows the changes in body composition variables by TBK and DXA in the two groups. Both groups experienced a significant weight loss, which was, however, not significantly different between the groups. There was a non-significant trend towards a more pronounced weight loss with OLS than with placebo treatment: OLS group 12.9 kg (DXA) and 11.2 kg (hospital scales); placebo group 10.0 kg (DXA) and 8.1 kg (hospital scales). The FM and the FM% fell significantly in both groups, but not significantly differently so between the groups. Changes in FFM were very small and only significant with DXA in the OLS-group (1.3 kg). This change was, however, not significantly different from that of the placebo group. With TBK, FFM fell insignificantly in the OLS group (1.3 kg) and rose insignificantly in the placebo group, giving rise to a significant group difference.

Bone mineral content and density measurements did not change significantly except for the forearm measurements (Table 4). All changes were, however, very small (including those of the forearm), and no significant group differences were seen for any of the bone measurements.

The biochemical variables (Table 5) showed several significant changes, probably due to dieting and weight loss. S-iPTH increased moderately but significantly in both groups (no significant group difference). S-25-OH vitamin D in both groups and 1,25(OH)₂ vitamin D in the OLS group fell significantly (no significant group difference). S-alkaline phosphatase, a marker of bone formation, increased moderately and significantly so in the OLS group (no significant group difference). S-osteocalcin, also a marker of bone formation, did not change significantly, either within or between the groups.

The only variable that changed significantly different in the OLS and the placebo group was the fasting urinary hydroxyproline/creatinine ratio (fU-OHpr/creat), a marker of bone resorption. It rose significantly in both groups, and significantly more so in the OLS group (Table 5 and Figure 1). Fasting urinary calcium/creatinine ratio, another marker of bone resorption, rose significantly in the OLS group, but not significantly in the placebo group (no significant group difference).

Linear regression analysis showed no relationships between the changes in any of the bone mineral variables and changes in any of the biochemical variables.

Discussion

The lack of significant differences between OLS and placebo-treated patients in all bone mineral measurements and in most biochemical variables during 1 y of approximately 10 kg weight loss shows that, at least with this duration of treatment, OLS is a safe drug in terms of preserving bone mineral content and density and maintaining calcium homeostasis. Partly due to the weight loss and partly due to the malabsorption induced by OLS, significant changes and group differences in more of the reported variables could be expected. There were some modest changes. The BMC and BMD values of the forearm declined significantly (but not significantly differently between the groups), and the fasting urinary hydroxyproline/creatinine ratio (marker of bone resorption) rose significantly more in the OLS treatment group than in the placebo group. Several of the other biochemical variables also changed, demonstrating a state indicative of increased malabsorption of vitamin D and calcium, increased bone turnover, increased bone resorption and a tendency towards a secondary hyperparathyroidism. As mentioned earlier, these changes were, however, not significantly different between OLS and placebo.

Thus, treatment with OLS seems to induce a moderate malabsorption, with a resulting limited influence on calcium metabolism. However, during 1 y of treatment, no effect was seen on bone mass or density other than the decline expected from weight loss per se.

In spite of more than 20 y of effort to investigate the matter, the question of the influence of obesity and weight loss on BMC and BMD is still controversial. Several studies have found a significant positive relationship between BMC or BMD in most anatomical sites and total body weight, lean body mass and fat mass.^{5,6,7,28,29,30,31,32} Generally, obese subjects have 2-20% higher BMC or BMD values than normal-weight subjects. Furthermore, in premenopausal women, there is a trend towards a more powerful relationship between lean body mass and bone variables, whereas in postmenopausal women the most powerful relationship is found between fat mass and bone variables.^29,32 Also, some studies have found significant relationships between bone mass and body weight, BMI or fat mass in women but not in men.^5,29,31

The pathophysiology behind the higher bone mass and density in obesity is unknown, but several mechanisms may be active. The increased 'loading' or 'strain' on the skeleton may have a positive influence in net bone formation in the bone remodeling process.¹¹ In postmenopausal women the aromatization of androstendion to estrone (a well known bone preserving hormone) in the fat tissue may be a partial explanation. Finally, an increased PTH-level has been found in obese subjects, and this slightly increased PTH-level may have an anabolic effect on bone.¹⁰

In most longitudinal studies performed, significant effects have been found of weight loss on BMC or BMD. In a number of older studies in obese patients in whom intestinal by-pass or ventricular stapling operations were performed, there were decreases in BMC or BMD ranging from 2 to 10% after a considerable (20-40 kg) weight loss.^3,4 In intestinal bypass patients malabsorption of vitamin D and calcium may partly have influenced bone metabolism.^3,4,15,16,33

A number of recent studies on diet-induced weight loss have found a significant decline in bone mass values ranging from 1-13%, partly dependent on the magnitude of the weight loss (3-22 kg), the duration of the study (10 weeks to 12 months), the composition of diet, the region for determination of bone mass, and determination of either BMC or BMD.^{8,9,34,35,36,37,38} The loss in bone mass variables in the placebo group of the present study ranged from no change in TBMC, TBMD, lumbar spine BMC and lumbar spine BMD to 2.3 and 2.0% in forearm BMC and BMD, respectively. This was perfectly in accordance with the literature, however, contributing to a concept of a relatively modest bone loss with weight loss of only approximately 8 kg.

This is a substudy of a larger OLS study involving 688 patients receiving OLS or placebo.¹ After 1 y the 284 patients in the main study who completed OLS treatment had lost 10.3 vs 6.1 kg lost by the 260 patients who completed 1 y of placebo treatment. In our substudy the corresponding losses were 11.2 and 8.1 kg respectively, ie a lesser difference between OLS and placebo. It is possible that more significant differences would had been seen if bone mass and calcium metabolic variables had been measured in the whole population.

It should be noted, that elevated parathyroid hormone levels in the context of obesity and obesity treatment may have at least three causes. First, a secondary hyperparathyroidism may be seen in morbid obesity due to change of behavioral pattern leading to decreased exposure to sunlight, and maybe a change in diet in favor of less vitamin D and calcium containing foods.^13,39 Second, obesity in itself may in some unknown way induce a slight overproduction of PTH, which may partly act to increase bone formation thereby contributing to the increase in bone mass seen in obesity.¹⁰ Third, the treatment, such as in this case OLS, may induce a vitamin D and calcium deficit and thereby a secondary hyperparathyroidism.

The final confounding aspect in this rather complex matter is the question of the validity of DXA to measure BMC and BMD in obesity and changes in these variables during weight loss. DXA has a well known, and extensively investigated, inborn tendency to erroneous measurements of BMC and BMD due to very thick or very thin layers of soft tissue covering the bones. This 'error' depends on rather simple radiophysical relationships, and it can therefore be predicted and corrected by theoretical or empirical computer algorithms and mechanical/technical filtering procedures. The problem is, however, that the correction procedures have been shown to have shortcommings due to the very wide variability and non-predictability of body soft tissue thicknesses and compositions. Therefore, even with the most sophisticated corrections, errors may be seen, resulting in over- or underestimation of BMC, bone area and BMD, and maybe even occational overcorrection of the errors.^{17,18,40,41,42,43} Furthermore, the different brands of DXA-scanners use quite different correction procedures which make comparisons between them a rather delicate matter.^17,18,40 The typical direction of the errors are such that the BMC (and maybe, but not always BMD) may be overestimated in extreme obesity and underestimated in extreme thinness. Therefore also the decrease in BMC and/or BMD during weight loss may be overestimated. Irregularities in the fat deposition (ie over bone vs next to bone) may also cause errors, and the direction and magnitude of the errors may vary between scanner brands, types and software versions. Furthermore, estimation of the fat percentage may also be influenced by errors (typically an overestimation of the fat percentage in extreme obesity), whereas the DXA-estimate of the total soft tisssue mass is always very accurate. The bottom line is that studies of the errors of accuracy of DXA must be performed with the same DXA-scanner brand, type and software as the clinical study in question.

There were in the present study small, non-significant differences between the weights measured by scales and estimated by DXA (Table 2; 1.1-1.4 kg differences). These differences are most likely to be explained by the precision error of the DXA weight estimate, which is approximately 2%.²⁰

We have made extensive studies on the accuracy errors of the Norland XR 36 also used in this study,¹⁹ and using phantom measurements and measurements of large thicknesses of lard and ox muscle placed on the abdomen and thighs of healthy subjects, and found this apparatus to be very accurate for both BMC, BMD and fat percentage. In particular, we did not find any significant dependence on the total body BMC or BMD of changing the thickness and composition of these tissue materials. We therefore feel confident that the results we reached in the present study as to BMC, BMD and soft tissue composition values are valid.

We note that the body compositional changes largely corresponded between DXA and the TBK method. In both groups, and with both methods, the weight loss was approximately 10 kg, of which approximately 80% was fat lost, resuIting in a decrease of approximately 5% in the FM%.

In conclusion, 1 y of treatment with OLS induces a lipid malabsorption which enhances a dietary weight loss without any significant deleterious effects on body composition. OLS induces a relative increase in bone turnover in favor of resorption, possibly due to malabsorption of vitamin D and/or calcium. However, no changes in bone mass or density are seen after 1 y of OLS treatment apart from those explained from the weight loss itself. Thus a one year OLS treatment seems safe from a 'bone preserving' point of view. A vitamin D and calcium supplement should be taken during the treatment.

Acknowledgements

We are grateful to The Foundation of Director Madsen and his wife Olga Madsen, The Danish Hospital Foundation for Medical Research, Region of Copenhagen, the Faroe Islands and Greenland, and to Hoffmann-La Roche for financial support. We cordially thank Lena Vind, Britta Skotte Hansen, Susanne Svalling for their skilled technical assistance during body composition measurements; Susanne Munch, Susanne Reimer, Vladimira Cvitanicic for laboratory assistance; and Helle Kanstrup for planning the examination program.

1 Sjøstrøm L, Rissanen A, Andersen T, Boldrin M, Golay A, Koppeschaar HPF, Krempf M. Randomised placebo-controlled trial of OLS for weight loss and prevention of weight regain in obese patients. Lancet 1998; 352: 167-173, Article MEDLINE

2 Dalén N, Halberg D, Lamke B. Bone mass in obese subjects. Acta Med Scand 1975; 197: 353-355, MEDLINE

3 Rickers H, Balslev I, Foltved H, Rødbro P. Bone mineral content before and after bypass operation in obese patients. Acta Med Scand 1981; 209: 203-207, MEDLINE

4 Krølner B, Ranløv PJ, Clemmesen T, Nielsen PS. Bone loss after gastroplasty for morbid obesity: side-effect or adaptive response to weight reduction? Lancet 1982; i: 956-957,

5 Kin K, Kushida K, Yamazaki K, Okamoto S, Inoue T. Bone mineral density of the spine in normal japanese subjects using dual-energy X-ray absorptiometry: effect of obesity and menopausal status. Calcif Tissue Int 1991; 49: 101-106, MEDLINE

6 Lindsay R, Cosman F, Herrington BS, Himmelstein S. Bone mass and body composition in normal women. J Bone Min Res 1992; 7: 55-63,

7 Compston JE, Bhambhani M, Laskey MA, Murphy S, Khaw KT. Body composition and bone mass in post-menopausal women. Clin Endocrinol 1992; 37: 426-431,

8 Compston JE, Laskey MA, Croucher PI, Coxon A, Kreitzman S. Effect of diet-induced weight loss on total bone mass. Clin Sci 1992; 82: 429-432, MEDLINE

9 Hyldstrup L, Andersen T, McNair P, Breum L, Transbøl I. Bone metabolism in obesity: changes related to severe overweight and dietary weight reduction. Acta Endocrinol 1993; 129: 393-398, MEDLINE

10 Rosen CJ, Donahue LR. Parathyroid hormone and osteoporosis. Curt Opin Endocrinol Diabetes 1996; 3: 532-539,

11 Frost HM. Obesity, and bone strength and 'mass': a tutorial based on insights from a new paradigm. Bone 1997; 21: 211-214, MEDLINE

12 Jørgensen NR. Cytokines and osteoporosis. Ugeskr Læger 1997; 160: 24-28,

13 Andersen T, McNair P, Hyldstrup L, Fogh-Andersen N, Nielsen TT, Astrup A, Transbøl I. Secondary hyperparathyroidism of morbid obesity regresses during weight reduction. Metabolism 1988; 37: 425-428, MEDLINE

14 Andersen T, Juhl E, Quaade F. Jejunoileal bypass for obesity¾what can we learn from a literature study? Am J Clin Nutr 1980; 33: 440-445, MEDLINE

15 Danø P, Christiansen C. Calcium absorption and bone mineral content following intestinal shunt operation in obesity. Scand J Gastroenterol 1974; 9: 775-779, MEDLINE

16 Hey H, Lund B, Sørensen OH, Lund BJ, Christensen MS. Impairment of vitamin D and bone metabolism in patients with bypass operation for obesity. Acta Med Scand 1979; 624: (Suppl) 73-78,

17 Tothill P, Avenell A, Reid DM. Precision and accuracy of measurements of whole-body bone mineral: comparison between Hologic, Lunar and Norland dual-energy x-ray absorptiometers. Br J Radiol 1994; 67: 1210-1217, MEDLINE

18 Tothill R, Avenell A, Love J, Reid DM. Comparisons between Hologic, Lunar and Norland dual-energy x-ray absorptiometers and other techniques used for whole-body soft tissue measurements. Eur J Clin Nutr 1994; 48: 781-794, MEDLINE

19 Gotfredsen A, Bæksgaard L, Hilsted J. Body composition analysis by DEXA by using dynamically changing samarium filtration. J Appl Physiol 1997; 82: 1200-1209, MEDLINE

20 Hendel HW, Gotfredsen A, Andersen T, Højgaard L, Hilsted J. Body composition during weight loss in obese patients estimated by dual energy x-ray absorptiometry and by total body potassium. Int J Obes Relat Disord 1996; 20: 1111-1119,

21 Forbes GB, Gallup J, Hursh JB. Estimation of total body fat from potassium-40 content. Science 1961; 133: 101-102,

22 Mosekilde L, Charles P, Lindegreen P. Determinants for serum 1,25-dihydroxycholecalciferol in primary hyperparathyroidism. Bone Miner 1989; 5: 279-290, MEDLINE

23 Lund B, Sorensen OH. Measurement of 25-hydroxyvitamin D in serum and its relation to sunshine, age and vitamin D intake in the Danish population. Scand J Clin Lab Invest 1979; 39: 23-30, MEDLINE

24 Hyldstrup L et al. Presented at the IXth International Conference on Calcium Regulating Hormones and Bone Metabolism. Nice, 1986,

25 Bergman I, Loxley R. The determination of hydroxyproline in urine hydrolysates. Clin Chim Acta 1970; 27: 347-349, MEDLINE

26 Bertels H, Bohmer M, Heierli C. Serum creatinine dertimination without protein precipitation. Clin Chim Acta 1972; 37: 193-197, MEDLINE

27 Schelkopf GM, Milne DB. Wet microwave digestion of diet and fecal samples for inductively coupled plasma analysis. Anal Chem 1988; 60: 2060-2062, MEDLINE

28 Harris SS, Dawson-Hughes B. Weight, body composition, and bone density in postmenopausal women. Calcif Tissue Int 1996; 59: 428-432, Article MEDLINE

29 Baumgartner RN, Stauber PM, Koehler KM, Romero L, Garry PJ. Associations of fat and muscle masses with bone mineral in elderly men and women. Am J Clin Nutr 1996; 63: 365-372, MEDLINE

30 Khosla S, Atkinson EJ, Riggs BL, Melton JL III. Relationship between body composition and bone mass in women. Bone Miner Res 1996; 11: 857-863,

31 Reid IR, Plank LD, Evans MC. Fat mass is an important determinant of whole body bone density in premenopausal women but not in men. J Clin Endocrinol Metabol 1991; 75: 779-782,

32 Douchi T, Oki T, Nakamura S, Ijuin H, Yamamoto S, Nagata Y. The effect of body composition on bone density in pre- and postmenopausal women. Maturitas 1997; 27: 55-60, Article MEDLINE

33 Compston JE, Vedi S, Gianetta E, Watson G, Civalleri D, Scopinaro N. Bone histomorphometry and vitamin D status after biliopancreatic bypass for obesity. Gastroenterology 1984; 87: 350-356, MEDLINE

34 Ramsdale SJ, Bassey EJ. Changes in bone mineral density associated with dietary-induced loss of body mass in young women. Clin Sci 1994; 87: 343-347, MEDLINE

35 Jensen LB, Quaade F, Sørensen OH. Bone loss accompanying voluntary weight loss in obese humans. J Bone Min Res 1994; 9: 459-463,

36 Pritchard JE, Nowson CA, Wark JD. Bone loss accompanying diet-induced or exercise-induced weight loss: a randomised controlled study. Int J Obes Relat Metab Disord 1996; 20: 513-520, MEDLINE

37 Andersen RE, Wadden TA, Herzog RJ. Changes in bone mineral content in obese dieting women. Metabolism 1997; 46: 857-861, MEDLINE

38 Van Loan MD, Johnson HL, Barbieri TF. Effect of weight loss on bone mineral content and bone mineral density in obese women. Am J Clin Nutr 1998; 67: 734-738, MEDLINE

39 Compston JE, Vedi S, Ledger JE, Webb A, Gazet J-C, Pilkington TRE. Vitamin D status and bone histomorphometry in gross obesity. Am J Clin Nutr 1981; 34: 2359-2363, MEDLINE

40 Tothill P, Avenell A. Errors in dual-energy x-ray absorptiometry of the lumbar spine owing to fat distribution and soft tissue thickness during weight change. Br J Radiol 1994; 67: 71-75, MEDLINE

41 Tothill P, Hannan WJ, Cowen S, Freeman CP. Anomalies in the measurement of changes in total-body bone mineral by dual-energy x-ray absorptiometry during weight change. J Bone Min Res 1997; 12: 1908-1921,

42 Milliken LA, Going SB, Lohman TG. Effects of variations in regional composition on soft tissue measurements by dual-energy x-ray absorptiometry. Int J Obes Relat Metab Disord 1996; 20: 677-682, MEDLINE

43 Madsen OR, Jensen J-EB, Sørensen OH. Validation of a dual energy x-ray absorptiometer: measurement of bone mass and soft tissue composition. Eur J Appl Physiol 1997; 75: 554-558,

Figure 1 Ratio of fasting urinary hydroxyproline to creatinine (µmol/ µmol) at baseline and after 1 y in 14 obese patients treated with orlistat (OLS) and the 16 obese patients given placebo. There were significant increases in both the OLS (P<0.001) and in the placebo group (P<0.05); and a significantly greater increase in the OLS group compared to the placebo group (P<0.05). OLS=orlistat.

Table 1 Clinical data; mean and s.d.

Table 2 Absolute values of body composition before and after 1 y of treatment with either OLS or placebo; mean and s.d.

Table 3 Changes in body weight and in body composition measured by DXA and by TBK respectively in patients treated with either OLS or placebo; mean, s.d. and confidence intervals are given

Table 4 BMC and BMD values measured by DXA before and after 1 y of treatment with either OLS or placebo; mean and s.d. are given

Table 5 Biochemical markers of calcium metabolism and bone turnover before and after 1 y of treatment with either OLS or placebo; mean and s.d. are given