ELSEVIER

Bone Mineral Metabolism is Normal in Non-Insulin-Dependent Diabetes Mellitus Manuel Sosa, Miri Dominguez, May C. Navarro, Mary C. Segarra, D. Hernhndez, P. de Pablos, and Pedro Betancor

ABSTRACT

Because of the previous controversial findings in non-insulin-dependent diabetes mellitus (NIDDM), we measured bone-mineral density (BMD) by two different methods, studied biochemical markers of bone remodeling and calciotropic hormones (parathyroid hormone and calcitonin) in women with NIDDM, and compared the results with age-matched controls. Forty-seven women with NIDDM and 252 healthy nondiabetic women as controls were recruited for this study. BMD was measured by dual X-ray absorptiometry (DEXA) and by quantitative computed tomography (QCT). Biochemical markers of bone remodeling included plasma alkaline phosphatase (AP),

INTRODUCTION

one disease is a well-known complication of dia-

B betes mellitus, particulary in children.‘f2 Al- bright and Reifenstein reported the coexistence of diabetes and osteoporosis,3 but there is con-

siderable current controversy.’ In insulin-dependent dia- betes mellitus (IDDM), bone mass is probably decreased

4-8 compared to controls, despite some normal bone mass values.’ In non-insulin-dependent diabetes mellitus (NIDDM), the reports are more contradictory, and bone mass may be increased, normal, or decreased.5,‘0,” Reports of metabolic studies are likewise contradictory.‘z-‘5

University of Las I’almas de Gran Canaria, Faculty of Medicine, Department of Clinical Sciences, Bone Metabolic Unit, Hospital Insu- lar, Las Palmas, Canary Islands, Spain

Reprint requests to be sent to: Dr. Manuel Sosa, University of Las I’almas, Faculty of Medicine, Department of Clinical Sciences, Bone Metabolic Unit, Box 550, 35080 Las Palmas, Canary Islands, Spain.

Journal of Diabetes and Its Complicntions 1996; ZO:ZOZ-205 0 Elsevier Science Inc., 1996 655 Avenue of the Americas, New York, NY 10010

osteocalcin (BGP), tartrate-resistant acid phosphatase (TRAP), parathyroid hormone (PTH), calcitonin (CT), and 24-h urine calcium, hydroxyproline. Diabetic patients were more obese with a higher body-mass index (BMI) than controls. Bone mass was normal in NIDDM, both by DEXA and by QCT. Biochemical markers of bone remodeling, PTH and CT were also normal. There was no statistical correlation between bone mass and any of the other measurements studied. There is no evidence that NIDDM produces any change in bone metabolism or mass. (Journal of Diabetes and Its Complications, 10;4:201-205, 1996.)

We have measured bone mass by dual X-ray absorp- tiometry (DEXA) and quantitative computed tomogra- phy (QCT), and biochemical bone remodeling markers and bone-mineral hormones in 47 NIDDM women to investigate any problem further.

METHODS

Subjects. We evaluated 47 Canarian female patients with NIDDM who were outpatients at the Bone Meta- bolic Unit. NIDDM was diagnosed on the basis of clas- sic symptoms, laboratory findings, and abnormal glu- cose tolerance test, according to National Diabetes Data Group criteria. l6 Most were obese and had a family history of NIDDM, 23 were receiving a sulphonylurea, 10 were receiving a sulfonylurea and a biguanide, and 11 were controlled with diet. None of them received insulin during the last 6 months.

A questionnaire about risk factor for osteoporosis was completed, and a physical examination was per- formed. None was taking medications or supplements

1056-8727/96/$15.00 SSDI 1056-8727(95)00062-7

known to affect with bone metabolism such as calcium supplements, Vitamin D preparations, etidronate, cal- citonin, or hormonal replacement therapy (HRT). None had a history of hepatic or renal disorders unrelated to diabetes or prolonged immobilization. Forty-four were postmenopausal women, and the remaining three were perimenopausal. Only women with natural meno- pause were accepted, and those with prior oophorec- tomy were excluded.

tion at 1500 g for 10 min, serum was aliquoted and stored within 1 h at -82°C for tartrate resistent-acid phosphatase (TRAP) and -20°C for the remainder assays; 24-h urine was also collected and stored frozen at -20°C until assays were performed.

The control population comprised 252 normal women who had been examined in a study of bone-mineral metabolism parameters and bone mass both with DEXAi7 and QCT. The methods used were the same as in the patient group. Written consent was obtained in every case.

Serum immunoreactive parathyroid hormone (II’TH) was measured by the Allegro intact PTH immunoas- say, which measures the biologically intact 84 amino- acid chain of PTH. The intra- and interassay variations were less than 3.5% and less than 6%. Sensitivity has been estimated in 1 pg/mL.

Procedure. Height and weight were measured to ob- tain the body-mass index (BMI) of each subject by the following equation: BMI = body weight/height! (kg/m2). Bone-mass density (BMD) was determined by DEXA and QCT. DEXA was performed with a densitometer model QDR-1000 (Hologic Esptia). BMD was mea- sured in the first, second, third, and fourth lumbar vertebrae (Ll-L4), the neck, trochanter, intertrochan- ter, and Ward’s triangle of the left femur.

Serum calcitonin (CT) was measured by Nichols In- stitute Diagnostics radioimmunoassay (IA) kit, based on principles first elucidated by Berson et al.‘” The intraassay variations was estimated in 11.6%, while inter- assay variations were less than 8.3%. Detection limit was established in 3 pg/mL. Serum osteocalcin (BGP) was also determined by Nichols RIA, and intraassay variation was 14.8% and interassay variation was 9.2%.

Results are presented as the mean of measurements of Ll-L4 and femoral neck. The precision is high; the coefficient of variation for absorptiometry is 0.6% for lumbar spine and 3.6% for femoral neck.

Calcium, alkaline phosphatase, phosphorus, urea, creatinin, and total proteins were measured by Kodak Ektachem Clinical Chemistry Slides. TRAP was deter- mined by spectrophotometric assay after to inactivate with sodium L (+) tartrate. The hydroxyproline in urine was measured by ion exchange spectrophotometry.

QCT bone density was measured in a Toshiba scan- ner model 600 HQ. A computed radiograph (lateral scout view) for localization of the vertebrae was first performed, and later a lo-mm-thick section was ob- tained at L3. An oval region of interest, centered in the midvertebral body was used to determine cancellous bone mineral content (mg/cc).18 A standardized ver- sion of Cann-Genant calibration phantom” is also used. DEXA and QCT measurements were performed in every patient and controls.

Statistical Analysis. Statistical studies were per- formed with SAS program (Statistical SAS Institute Inc, Cary, NC, USA). Results are expressed as mean t SD. Nonparametric test of Kolmogorov-Smirnoff was used to assess the normal distribution of each parameter. In those who had a normal distribution, unpaired t test was used to compare pairs of independent means. in those parameters that did not have a normal distribu- tion, Mann-Whitney test was used to compare means. Two-factor analysis of variance (ANOVA) was used to examine the effect of diabetes and weight in bone mass. Correlation between variables were also studied. The level of significance was defined as OL of 0.05.

Sample Collections and Assay. Serum and urine speci- mens were obtained after an overnight fast and 24 h of a gelatin-free diet. Blood was collected in vacutainers without additive between 8 and 9 a.m. After centrifuga-

Others. Serum calcium values were corrected by Par- fitt formula: corrected calcium = previous calcium/ (0.55 + total proteins/lb). Body surface area was calcu- lated by standard nomogram. Only three patients smoked.

TABLE 1. PATIENT CHARACTERISTICS

Control Group NIDDM

Number 252 47 Height (cm) 156.8 t 6 156.0 ? 5.8 Weight (kg) 69.3 ‘_ 10.9 73.1 ? 11 Surface area (m2) 1.5 ” 0.1 1.5 -+ 0.1 BMI (kg/m*) 28.2 I! 4.2 30.0 -e 4.2 Age (years) 58.8 i 8.5 61.3 ‘- 7

-- NIDDM, non -insulin-dependent diabetes mellitus; BMI, body-mass index.

__.. p value

NS < 0.05

NS < 0.01

NS

J Diab Comp 1996; 10:201-205 NORMAL BONE MASS IN NIDDM 203

TABLE 2. RENAL FUNCTION AND SOME SERUM BIOCHEMICAL VALUES

Creatinine (mg/dL) Ccr (mL/m) Calcium (mg/dL) Phosphorus (mg/dL) Total proteins (g/dL) Corrected calcium (mg/dL) Fasting glucose (mg/dL)

Control Group NIDDM p value

(Mean f SD) 1 -c 0.2 1 2 0.2 NS

88.1 ? 24.3 84.6 t 33.7 NS 9.4 k 0.5 9.4 -c 0.5 NS 3.7 5 0.5 3.8 t 0.6 NS 7.3 t 0.5 7.5 ? 0.4 p < 0.05 9.3 -e 0.5 9.2 2 0.4 NS

102.3 ? 18.3 154.1 ? 47.6 p < 0.001

NIDDM, non-insulin-dependent diabetes mellitus; Ccr, creatinine clearance; corrected calcium, calcium corrected with total proteins.

RESULTS

Patient characteristics are shown in Table 1. Diabetic patients are heavier than controls and have a higher BMI. Renal function evaluated by plasma creatinin and other biochemical data are shown in Table 2. Serum creatinine values and creatinine clearance were similar in both groups. Biochemical markers of bone remod- elling and bone homostasis hormones are shown in Table 3. There were no statistical differences between patients and controls with any of these markers. Levels of PTH and CT were also similar in both groups. Bone mass levels measured by DEXA and by QCT are shown in Table 4. There were no differences in bone mass between diabetics and controls either by DEXA in lum- bar spine or femoral neck or by QCT.

Analysis of variance (ANOVA) was used to examine the effects of diabetes and weight in bone mass. There were no statistical differences.

DISCUSSION

The aim of this work was to investigate if NIDDM produces alterations in bone mineral metabolism and bone mass. As osteoporosis is far more frequent in postmenopausal women, we chose 47 NIDDM women, 44 of whom had a natural menopause. We eliminated other known factors that could interfere in bone metab-

olism; no one was receiving HRT, had a history of immobilization, or evidence of renal or hepatic disease. The age of the two groups was similar.

NIDDM patients were more obese as shown in Table 1, their weight and BMI were higher than controls (p < 0.05 and p < O.Ol), which is a well-known feature. Calcium, phosphorus, and calcium corrected with total proteins were similar in both groups, without statistical differences (Table 2). Lower levels of calcium and phosphorus compared to controls have been previously described in IDDM, although no correlation was ob- served between these minerals and PTH and 1,25 DHCC.12 In our study, serum levels of calcium, phos- phorus, and corrected calcium were normal in both groups, which agrees with the results found by other investigators in NIDDM patients.9J’J2 We have found that NIDDM patients have higher total protein levels than controls, and we do not have an explanation for this fact. Nevertheless, in both cases, total protein levels were between standard normal values.2”

Biochemical markers of bone remodeling were nor- mal in NIDDM. Both parameters of osteoblast activity [osteocalcin (BGP) and alkaline phosphatase (Al’)] and osteoclast activity (serum TRAP and urine calcium and hydroxyproline) had no differences in diabetics com- pared to controls. Other papers described hypercalci- uria and hyperphosphaturia in IDDM reversible with

TABLE 3. BIOCHEMICAL MARKERS OF BONE REMODELING AND BONE HOMEOSTASIS HORMONES

Control group NIDDM p value

AI’ (U/L) BGP (ng/mL) TRAP (U/L) Hydroxyproline (mg/24 h) Urine calcium (mg/24 h) PTH (pg/mL) Calcitonin (pg/mL)

(Mean -+ SD) 85.7 5 31.4 94.7 +- 38.3 NS 8.3 2 6.5 9.5 t 6.5 NS 3.2 2 1.1 3.3 ? 1.1 NS

38.6 k 26.5 40.7 5 26.8 NS 179.7 t 110.8 185.1 If: 93 NS 36.5 t 14.5 34.7 ? 14.6 NS

7.6 t 5.2 8 2 6.8 NS

NIDDM, non-insulin-dependent diabetes mellitus; AP, alkaline phosphatase; BGP, osteocalcin; TRAP, tartrate-resistant-acid-phosphatase; PTH, parathyroid hormone.

204 SOSA ET AL

TABLE 4. BONE MASS LEVELS BY DEXA AND QCT

Control group NIDDM

DEXA (Ll-L4) (g/cm*) 0.892 ? 0.138 0.898 -: 0.137 DEXA (femoral neck) (g/cm’) 0.737 5 0.115 0.756 = 0.146

QCT (mg/cc) 126.5 k 41.2 111.8 x 47.9

NIDL1M, non-insulin-dependent diabetes mellifus; DEXA, dual X-ray absorptiometry; QCT, quantifatiue computed tomopzpohy.

p value

NS NS NS

continuous or conventional subcutaneous insulin treat- ment,24,25 which sometimes can be due to an extremely poor metabolic control of diabetes.24 In other papers, hypercalciuria is not observed neither in IDDM nor in NIDDM9 as happened with our patients. Al’ has been described increased in diabetic patients with neuropa- thyz6 and Nishikawa et a1.27 quantifying Al’ by electro- phoresis, found the bone isoenzyme increased only in diabetic patients with osteopenia. Some papers suggest that in diabetics exist an alteration in osteoblast func- tion or a decrease in their number or both, documented it by low serum levels of BGP2* but in our patients, serum BGP levels were similar to controls.

In our study, PTH and basal CT were also normal. Although some papers described an increase in PTH levels in rats, when better rat PRH assays were used, low or undetectable PTH levels were observed.2 Low PTH levels were described in poorly controlled IDDMF2 Indeed, no signs of secondary hyperparathyroidism were observed in bone histology. We also studied pos- sible correlations between bone mass, measured either by QDR in lumbar spine or neck femoral or QCT, and biochemical and hormonal values, and found no correlation at all between them.

Bone mass measured both by DEXA and by QCT showed similar values to controls. This point is a matter of controversy. Some authors have found decreased bone mass levels in NIDDM,“,29-31 but the techniques used vary widely; radiogrammetry, single (SPA) and dual photon absorptiometry (DPA), total-body neutron activation, and resonant frequency of the ulna have been used.*Jj Barrett-Connor and Holbrook, using both DPA and DEXA, found normal bone mass values in women with NIDDM, and even increased values in some when compared to controls,32 as happened with our patients. Some other authors have found normal bone mass values in NIDDM as we11.9-11 Comparing the results obtained by Wakasugi et al.“” to ours using the same method (Hologic QDR-lOOO), the results were very similar: In lumbar spine: 0.87 g/cm2 versus 0.89 g/cm2.

We have measured bone mass by two different meth- ods, DEXA and QCT, and in different sites, lumbar spine and femoral neck; thus we can conclude, in view of the results, that bone mass in NIDDM is normal.

ACKNOWLEDGMENT

We would like to thank to Dr. Roger Smith from the Nuffield Orthopaedic Center, Headington, Oxford, United Kingdom, for his comments and review of this paper, and to Ciba- Geigy Laboratories, who provided us the bibliography with their service “CG-On line.”

C. Dominguez received a grant from the University Foun- dation of Las Palmas. Hollogic densitometer was supplied by Rhone-Poulenc-Rorer Laboratories.

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

REFERENCES

Morrison LB, Bogan IK: Bone development in diabetic children: A roentgen study. Am J Med Sci 174:313-319, 1927.

Bouillon R: Diabetic bone disease. Cal@ Tissue lnt 49: 155-160, 1991.

Albright F, Reifenstein EC: The Parathyroid Glands and Metabolic Bone Disease: Selected Studies. Baltimore, Williams and Wilkins, 1948.

Ringe JD, Kuhlencordt F, Kruse HP: Bone mineral deter- minations on long-term diabetics. Am J Roentgenol 126: 1300-1301,1976.

Levin ME, Boisseau VC, Avioli LV: Effects of diabetes mellitus on bone mass in juvenile and adult-onset diabe- tes. N Engl J Med 294:241-245, 1976.

McNair P, Madsbad S, Christensen MS, et al.: Bone mineral loss in insulin-treated diabetes mellitus: Studies on pathogenesis. Acta Endocrinol 90:463-472, 1979.

Rosenbloom AL, Lezotte DC, Weber FT, et al.: Diminu- tion of bone mass in childhood diabetes. Diabetes 26: 1052-1055, 1977.

Wiske S, Wentworth SM, Norton JA Jr, Epstein S, John- ston CC Jr: Evaluation of bone mass and growth in young diabetes. Metabolism 311848-854, 1982.

Giacca A, Fassina A, Caviezel F, Cattaneo AG, Caldirola G, Pozza G: Bone mineral density in diabetes mellitus. Bone 9:29-36, 1988.

Meema EF, Meema S: The relationship of diabetes melli- tus and body weight to osteoporosis in elderly females. Can Med Assoc J 96:132-139,1967.

De Leeuw I, Abs R: Bone mass and bone density in maturity-type diabetes measured by the ‘=I photon- absorption technique. Diabetes 26:1130-x135, 1977.

Auwerx J, Dequecker J, Bouillon R, Geusens P, Nijs J: Mineral metabolism and bone mass at peripheral and axial skeleton in diabetes mellitus. Dkbetes 37%12,1988.

Frazer TE, White NH, Hough S, et al.: Alterations in circulating vitamin D metabolites in the young insulin-

T Diab Comp 1996; 10:201-205 NORMAL BONE MASS IN NIDDM 205

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

dependent diabetic. f Clin Endocrinol Metab 53:1154- 1159,1981.

Witt MF, White NH, Santiago JV, Seino Y, Avioli LV: Use of oral calcium test loading to characterize the hyp- ercalciuria of young insulin-dependent diabetes. I Clin Endocvinol Metab 57:94-100, 1983.

Heath H III, Melton LJ III, Chu C-P: Diabetes mellitus and the risk of skeletal fracture. N Engl J Med 303:567- 570, 1983.

National Diabetes Data Group: Classification and diag- nosis of diabetes mellitus and other categories of glucose intolerance. Diabetes 63:843-847, 1977.

Proyecto Multicentrico de Investigation de Osteoporo- sis: Bone mineral density in the Spanish population measured by dual x-ray absorptiometry (DEXA). XILh International Conference on Calcium Regulating Hor- mones. Bone Miner 17(suppl):133, 1992.

Lang I’, Steiger I’, Faulkner K, Gltier C, Genant H: Osteo- porosis: Current techniques and recent developments in quantitative bone densitometry. Radio1 Clin North Am 29:49-76, 1991.

Cann CE, Genant HK: Single versus dual-energy CT for vertebral mineral quantification. J Cornput Assist Tomogr 7:551-560, 1983.

Berson SA, Yallow RS, Bauman A, Rothschild MA, New- erly K. I Clin Invest 28:170-175, 1957.

Heath H III, Lambert PW, Service FJ, Amaud SB: Cal- cium homeostasis in diabetes mellitus. J Clin Endocrinol Metab 491462466, 1979.

Fialip J, Moinade S, Thieblot I’, et al.: Study of phospho- rus and calcium metabolism in varied groups of diabet- ics: Insulin dependent and non-insulin dependent, well and poorly controlled. Diabetes Metab 11:283-288, 1985.

Elin RJ: Reference intervals and laboratory values of clinical importance, in Wyngaardeen JB, Smith LIH, Ce- cil Textbook of Medicine (eds). WB Saunders, Philadel- phia. 1988: pp. 2394-2404.

24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

Gertner JM, Tamborlane WT, Horst RL, Sherwin RS, Felig I’, Gene1 M: Mineral metabolism in diabetes melli- tus: changes accompanying treatment with a portable subcutaneous infusion system. 1 Clin Endocrinol Metab 49:462466, 1979.

Raskin I’, Stevenson RM, Barilla DE, Pak CYC: The hypercalciuria of diabetes mellitus: Its amelioration with insulin. Clin Endocrinol (Oxfl 9:329-335, 1978.

Cundy TF, Edmonds ME, Watkins PJ: Osteopenia and metatarsal fractures in diabetic neuropathy. Diabetic Med 2:461464, 1985.

Nishikawa Y, Kanda T, Yoshihara H, Fukumoto K, Uet- masu I: Cellulose acetate electrophoretic determination of bone alkaline phosphatase activity in healthy subjects and diabetic patients with and without osteopenia. Clin Ckim Acta 210:13-22, 1992.

Rico H, Hemandez ER, Cabranes JA, Gomez-Castresana F: Suggestion of a deficit osteoblastic function in diabetes mellitus: The possible cause of osteopenia in diabetics. Calcif Tissue lnt 45:71-73, 1989.

Shao AH, Wang FG, Hu YF, Zhang LM: Calcium metab- olism and osteopathy in diabetes mellitus. Contrib Nepkrol 90:212-216, 1991.

Isaia G, Bodrato L, Carlevatto V, Mussetta M, Salamano G, Molinatti GM: Osteoporosis in type II diabetes. Acta Diabetol 24:305-310, 1987.

Ishida H, Seino Y, Matsukura S, et al.: Diabetic osteope- nia and circulating levels of vitamin D metabolites in type I (noninsulin-dependent) diabetes. Metabolism 34: 797-801, 1985.

Barrett-Connor E, Holbrook TL: Sex differences in osteo- porosis in older adults with non-insulin-dependent dia- betes mellitus. JAMA 268:3333-3337, 1992.

Wakasugi W, Wakao R, Tawata M, Gan N, Koizumi K, Onaya T: Bone mineral density measured by dual x-ray absorptiometry in patients with non-insulin-dependent diabetes mellitus. Bone 14:29-33, 1993.