Bone Mineral Changes in Obese Women During a Moderate Weight Loss With and Without Calcium Supplementation



A significant relationship between body weight (BW) and bone mass (BM) has been established previously. A diet-induced weight loss is accompanied by a significant decrease in bone mineral density (BMD) and total body bone mineral (TBBM), but the underlying mechanisms are not clarified. Sixty-two obese women were included in the study. Dual-energy X-ray absorptiometry (DXA) and measurements of a series of calcium-regulating hormones and biochemical markers of bone turnover were performed at baseline and after 1 month and 3 months on a low calorie diet. Thirty of the women were randomized to a daily supplement of 1 g of calcium. After an additional 3 months without dietary prescriptions or calcium supplements, a subgroup of 48 subjects (24 from each group) were scanned again using DXA. There was a significant decrease in TBBM after 1 month and 3 months. A similar pattern was observed in the bone mineral content (BMC) of the lumbar spine in the patients who did not receive a calcium supplement, whereas no changes occurred in the supplemented group. The initial calcium supplementation seemed to protect against bone loss in the lumbar spine but not in the TBBM. In the nonsupplemented group, a statistically significant inverse correlation was found between the calcium/creatinine ratio in the morning urine and the changes in BMC of the lumbar spine. Such a relationship was not seen in the calcium-supplemented group. In the nonsupplemented group, no significant biochemical changes were observed, whereas a significant decrease in serum parathyroid hormone (PTH) was seen in the calcium-supplemented group. This might explain some of the protective effects of calcium supplementation on trabecular bone mass. We conclude that a diet-induced weight loss is accompanied by a generalized bone loss, which probably is explained mainly by a reduced mechanical strain on the skeleton. This loss can be partly inhibited by a high calcium intake. Therefore, a calcium supplementation should be recommended during weight loss, even if the diet contains the officially recommended amounts of calcium.


ASIGNIFICANT relationship between body weight (BW) and bone mass (BM) and between total body bone mineral (TBBM) and body fat mass (FM) has been established in pre- and postmenopausal women.(1–3) The positive correlation of BM to BW indicates that a readjustment of BM must take place subsequently to alterations in BW.

The three body components—TBBM, FM, and lean body mass (LBM)—can be measured separately by dual-energy X-ray absorptiometry (DXA), a method whereby the bone mineral measurements, only in extreme situations, are influenced by changes in adjacent fat tissue, and therefore it is convenient for monitoring the skeleton in subjects who are losing weight.(4–8) Both regional bone mineral and TBBM can be evaluated with high precision and accuracy.(4,9,10)

A diet-induced weight loss is accompanied by a significant decrease in bone mineral density (BMD) and TBBM,(8,11) and one study indicates that a 1 g/day calcium substitution can suppress bone turnover in energy-restricted subjects.(12) The underlying mechanisms for the bone changes are not clear. Suggested explanations are: (1) weight relief on the bones causes an adjustment of the skeleton probably mediated by humoral factors; (2) secondary hyperparathyroidism has been described in obese patients but, during weight loss, an increase in Ca2+ is seen followed by a decrease in parathyroid hormone (PTH),(13,14) which causes increased renal calcium loss as a result of increased ultrafiltration and decreased reabsorption of calcium; (3) diminished extraovarial estrogen synthesis; and (4) though the calcium supplied by weight-reducing diets generally may have been sufficient by current recommendations, obese patients losing weight may have a higher calcium demand, either because of the increased loss of calcium mentioned previously or because of reduced intestinal absorption.

The purpose of the present investigation was: (1) to measure changes in the skeleton in relation to changes in the body composition in obese patients undergoing a moderate weight loss on a low calorie regimen; (2) to relate a number of biochemical markers of bone metabolism and calcium regulating hormones to the changes in the BM; and (3) to investigate the influence of a calcium supplement on bone loss in obese persons during a low calorie diet.


Sixty-two obese women (body mass index [BMI], 25.2–51.6 kg/m2) with a motivation for weight loss were included in the study and randomized at baseline to no treatment or to 1 g (26 mmol) of calcium supplement (Calcium-Sandoz; Sandoz, Tietgensgade, Copenhagen, Denmark) in the first 3 months.(12,15) The result of the randomization was open and no placebo was used. All women attended an outpatient clinic and underwent initial physical examination and history. Seventeen of the women were postmenopausal. Eight of these were in the treatment group. Estrogen substitution was an exclusion criterion as well as severe somatic diseases and metabolic calcium disturbances.

A formula diet (NUPO; Oluf Mørk Ltd., Glostrup, Denmark) yielding 1.9 MJ with a qualitatively free supplement of food and drink up to 4.2 MJ(16) was prescribed for the first 3 months. The daily amount of formula diet contained 58 g of protein, 800 mg of calcium, 800 mg of phosphate, and 200 IU of vitamin D. After the first 3 months the patients continued on a normal diet but were advised to reduce the calorie intake.

Of the initial 62 women, 4 women were excluded from the study during the first month because of lack of compliance and no weight loss, 2 women failed to take the prescribed calcium supplement, and 4 women left the study for personal reasons. Of the remaining 52 women, 14 were postmenopausal and 8 of these were in the no-treatment group.

Whole body measurements of the bone mineral content (BMC), FM, and fat free mass (FFM) (CV 1.1, 1.6, and 1.6%, respectively (8,17)) and regional measurements of the lumbar and femoral BMC (CV 1% and 1.5% respectively(18)) were performed by a Hologic QDR-2000 DXA scanner (Hologic Inc., Westham, MA, USA) using the pencil-beam mode, at baseline after 1 month and after 3 months in 52 women. Twenty-four women (6 postmenopausal) from each group were rescanned 6 months from baseline. To obtain the best possible comparability at follow-ups, the baseline regional scans were restored in the scanner and the compare facility was applied, so that corresponding areas were analyzed at baseline and follow-ups. Fasting blood samples were drawn in the morning. Serum was stored at −20°C. All samples were analyzed in duplicate in the same assays. Urine was collected as the second void fasting morning urine. Serum concentrations of osteocalcin were measured by a bovine ELISA assay (DAKO, Glostrup, Denmark), and serum 25-hydroxyvitamin D [25(OH)D] and 1,25-dihydroxyvitamin D [1,25(OH)2D] were measured by competitive protein binding assays using rachitic rat kidney cytosol and thymus receptor, respectively. Intact serum PTH was measured by an immunoradiometric assay (Allegro Intact PTH; Nichols Institute, Nichols Institute Diagnostics, Los Angeles, CA, U.S.A.). Urinary calcium and creatinine were analyzed by standard methods and pyridinoline and urine-deoxypyridinoline were analyzed by high-performance liquid chromatography and fluorescence detection after hydrolysis of the urine as previously described.(19)

The study fulfilled the Helsinki declaration and was approved by the local ethics committee. All participants gave informed consent to participate in the study.

Statistical analysis

The multivariate analysis of variance (MANOVA) repeated measures design, Tukey's post hoc test, Student's t-tests, and Pearson's linear regression analysis were used where appropriate. Statistically significant changes over time are marked with an asterisk at the end of the tables. Statistically significant differences between groups are marked with an asterisk in the right margins, except in cases in which the precise p values are tabulated. When statistically significant bone changes were found, correlations of these to weight changes and to bone markers in blood and urine were calculated. The values of p < 0.05 were considered statistically significant. The statistical package “Systat for the Macintosh” (Systat, Inc., Evanston, IL, U.S.A.) was used for all statistical analysis.



The differences between groups were not statistically significant at baseline. The mean weight losses in the first month were 3.3 kg and 3.1 kg in the untreated group and in the calcium-supplemented group, respectively. The corresponding weight losses in the following 2 months were 2.9 kg and 2.3 kg. At the 6-month follow-up the untreated group had regained 1.4 kg while the calcium-supplemented group had no overall weight change. The changes in BMI followed the same pattern (Table 1).

Table Table 1.. Weight, BMI, and Bone Mineral Changes
original image


The differences between groups were not statistically significant at baseline. The mean changes in the BMC of the lumbar spine, right hip, and whole body are shown in Table 1. There were no mean bone changes in the first month in the lumbar spine or in the hip, but the untreated patients lost on average 1.5% of TBBM, and the supplemented group lost 2.0% of TBBM in that period. There were significant bone losses (2.0% and 2.2%) in the lumbar spine after 3 months and 6 months in the untreated group but not in the supplemented group. The unadjusted difference between groups was borderline statistically significant (p = 0.055), but adjustment for the initial weight loss decreased the value of p < 0.01.

The same tendency was found in the right hip with a 4.0%, though not statistically significant, bone loss after 6 months in the untreated group only.

The decline in TBBM was statistically significant already after 1 month in both groups (Table 1).

Weight-loss and BMC in the lumbar spine

The relationships between the changes in the BMC in the lumbar spine and the weight losses after 1, 3, and 6 months are shown in Table 2. The change in BMC in the lumbar spine after 6 months was significantly correlated to the weight changes after 1 month and 3 months but not to the weight change after 6 months and only in the untreated group. The weight loss after 1 month did not correlate to the BMC change after 1 month but accounted for 25% of the variation in the BMC change after 6 months.

Table Table 2.. Changes in the BMC in the Lumbar Spine Related to Changes in the BW
original image

Biochemical parameters

The serum levels of osteocalcin, 25(OH)D and 1,25(OH)2D, and PTH are shown in Table 3. There was an overall tendency to an increase in the osteocalcin, especially in the first months and especially in the untreated group. The increase was statistically significant after 1 month in the untreated group. No significant changes occurred in 25(OH)D and 1,25(OH)2D. The PTH declined in the treated group with 11.5% after 1 month, 15.5% after 3 months, and 19.6% after 6 months (p < 0.05%, p < 0.01%, and p < 0.05%, respectively). The changes in PTH and the bone changes did not correlate.

Table Table 3.. Osteocalcin and Calciotropic Hormones in Serum
original image

The calcium/creatinine ratio and the concentrations of pyridinoline and deoxypyridinoline in the spot urine are shown in Table 4.

Table Table 4.. Urinary Calcium and Pyridinolines
original image

The calcium/creatinine ratio increased after 1 month and remained significantly elevated after 3 months in the treated group. In the untreated group the 3-month value showed an increasing tendency. Urinary pyridinoline and deoxypyridinoline did not change.

Changes in the calcium/creatinine ratio in spot urine and the BMC in the lumbar vertebrae

The relationships between the changes in the BMC in the lumbar spine and the changes in urine calcium/creatinine ratios after 1 month and 3 months are shown in Table 5.

Table Table 5.. Changes in the BMC in the Lumbar Spine Related to Changes in the Urine Calcium/Creatinine Ratio
original image

There was a uniform pattern with a negative correlation between the changes in the urine calcium/creatinine ratios and the changes in the lumbar BMC in the untreated group.


In agreement with earlier investigations,(8,11) a close correlation was observed in the present study between weight loss and TBBM loss. We found similar changes in the lumbar spine and in the proximal femur regions that are especially exposed to osteoporotic fractures. No mineral was lost in the lumbar spine when a 1 g calcium supplement was prescribed, in addition to the formula diet already yielding sufficient calcium. This is a new observation that might have practical implications for future treatment of obesity and may suggest some explanation for the bone changes accompanying weight loss.

The urinary calcium excretion analyses confirm the bone changes and to some degree the theory of impaired calcium supply. When no calcium supplement was given, urinary calcium and BMC changes in the lumbar spine were correlated in the expected manner, with increasing loss of calcium in the urine accompanying loss of BMC. PTH decreased in the calcium-supplemented group, an expected effect of calcium supplementation. Suppression of PTH will result in a reduced bone resorption and thus a reduced loss of bone.

The changes in bone mass probably are induced by local factors and by changes in the mechanostat described by Frost.(20) According to Frost, the mechanostat comprises one or more mechanisms that orchestrate the biological activities that maintain the skeleton's architecture and strength. It is not known which cells participate in the regulation, but according to Frost et al.(21) it is neither the osteoblasts nor the osteoclasts. In recent years there has been increasing experimental support that the osteocytes, through their extensive network, are able to detect changes in bone strain and thereby to control bone modeling and remodeling.(22) The molecular mechanisms by which the cells perceive changes in mechanical loading are not exactly known, but one attracting theory is the involvement of integrins, which are receptors through which cells attach the matrix where the strain is produced.(23) Local effects might be mediated through prostaglandins, cytokines, and growth factors.

The hormone leptin, which seems to have a central appetite regulating action, is greatly influenced by weight loss and recently has been investigated in the light of bone metabolism.(24–27) We did not measure serum leptin in our patients, but other studies point in the direction that covariation between bone mass and leptin is caused by the strong correlation between leptin and BW and between BW and bone mass, and that leptin has only marginal, if any, direct effect on the bone mass in the mature skeleton.(26,27) The situation seems to be more complex if the skeleton is still growing, and correlation analyses suggest a positive effect of leptin on BMD in children.(25)

The present study includes measurements after 1 month and at the end of the diet. Therefore, it is possible to evaluate the timing of the bone changes in relation to the weight changes. Although half of the weight loss took place in the first month, it took 3 months before there was any detectable bone loss in the lumbar spine, and the lumbar bone status was much better predicted from the weight status after 1 month than from the weight status after 6 months. It is not surprising that any effect that initiates a change in bone needs this time for the full effect to show, and the bone changes measured initially may be reflections of the increased number of reabsorption lacunae caused by bone remodeling. An extrapolation of this would be that the weight at 6 months would predict the bone status at 12 months, and thus any regained weight would result in regained bone at that time. We have no measurements at 12 months to support this, but a previous study of TBBM included 9-month measurements, and we found that weight regain resulted in a corresponding regain of TBBM.(8) The delay between weight loss and its corresponding bone loss is a strong argument against the explanation that the observed bone changes should be systematical measurement errors, caused by alterations in the soft tissues.

The most convincing effect of the calcium substitution was in the lumbar spine. The loss of TBBM on the other hand, was significant also in the treatment group, but the tendency to a lesser bone loss, when extra calcium was given, was the same in all regions measured, and some factors may have tended to underestimate the between group differences. Primarily, the result of the randomization was known to the participants. This may have accomplished self-treatment in the placebo group, which would level out some of the difference, and uncontrolled confounders such as tobacco smoking and exercise could increase variation and make differences harder to show. The estimated SE on the TBBM being 75 g demands the difference between groups to be at least 147 g in order to be statistically significant. A larger study or a more substantial weight loss may more clearly show a difference in all regions. Overestimation of the TBBM and of the BMC in the lumbar spine in the treatment group, because unabsorbed calcium in the gut occasionally may overlay bone pixels, cannot explain the preservation of bone in the treatment group. Only 1 g of calcium was given daily, and in our opinion only a small fraction of this is likely to be mistaken as bone, because only calcium in bone pixels is measured as such. This line of argument is supported by the fact that the preservation of bone also lasts through the final 3 months, when no calcium was given.

It has been suggested that bone loss in weight losers is an unwanted effect of very low calorie diets, and that it can be avoided by using regimes with a higher energy content.(28) Our regime was more liberal with a prescribed energy intake up to 4.2 MJ and moderate weight losses, and we still found significant losses in TBBM and lumbar BMC in the absence of calcium substitution. We conclude that a bone loss is an inevitable result of weight loss and that it is not avoided by a higher calorie regimen as long as weight is lost. It can be diminished by additional dietary calcium. In dieting subjects, the usual recommended calcium intake is insufficient and a calcium supplement is advisable.