IAGS 39:192-296, 1991
The Metabolic Effects of Calcium Supplementation in the Elderly Gary Kochersberger,MD,* Ronald WestZund,MS,t Kenneth W.Lyles,MDS
Because the value of calcium supplementation in agerelated bone loss is controversial, we conducted a study to examine the effects of six months of calcium supplementation on parathyroid hormone (PTH) and other indices of bone metabolism. Calcium carbonate (1.2 grams elemental calciumlday) or placebo was administered for 6 months to a group of 42 healthy, ambulatory men and women with a mean age of 71. Fasting blood samples and 2-hour urine collections were performed at baseline and at 3 and 6 months. The decline in serum PTH levels from baseline in the calcium-supplemented group was significant in comparison with the placebo group changes (for the calcium group: 40.9 pg/mL at baseline to 34.0 pglmL at 1 month and 33.2 at 6 months). This drop in PTH was associated with a decline in 1,25 OH-Vitamin
D levels from a mean of 38.3 pglmL to 30.4 pglmL at 1 month and 32.9 at 6 months. During calcium supplementation, no changes were observed in serum bone Gla protein values. A decline in urinary hydroxyproline excretion was observed at 6 months, but this did not reach significant levels. The present study demonstrates suppression of PTH levels with calcium supplementation, an effect which lasts at least 6 months. This change was accompanied by a decrease in serum 1,25 OH-Vitamin D. The lack of significant changes in either serum bone gla protein or urine hydroxyproline excretion fails to support any significant change in bone turnover occurring with this decrease in PTH. J Am Geriatr SOC39:192-196, 1991
he differences observed in the epidemiology and suspected pathogenesis of osteoporosis has led to the subdivision of the disease into type I (post-menopausal) and type I1 (senile).' The cause of type I osteoporosis is not fully understood, but is felt to be related to a decline in osteoblast activity and an observed increase in circulating levels of parathyroid hormone (PTH).*This increase in PTH is felt to be secondary to an age related decline in the formation of 1,25-dihydroxyvitamin D and a decrease in intestinal calcium absorption in part caused by the reduced 1,25 dihydroxyvitamin D3.
The use of calcium supplements in the prevention of osteoporosis has been extensively studied in numerous clinical trial^.^'^' ' For the most part, however, these studies have only examined women in the immediate post-menopausal period, who are at risk for type I osteoporosis. We recently examined the effects of calcium supplementation in a group of elderly men and women at risk for type I1 osteopor~sis.~ In that study we demonstrated that 4 weeks of calcium supplementation was capable of significantly reducing serum levels of immunoreactive PTH, suggesting possible utility in the management of type I1 disease. A study was therefore conducted to expand upon this earlier work and to look at the effects of longer duration supplementation including biochemical parameters felt to be reflective of bone turnover.
T
From the * Department of Veterans Affairs Medical Centers, Batavia and Buffalo, New York and the Department of Medicine, Division of Geriatrics, University of Buffalo, Buffalo, New York; t the Center for the Study of Aging and Human Development, Duke University, Durham, North Carolina; and $ GRECC, Department of Veterans Affairs Medical Center, and the Department of Medicine, Divisions of Geriatrics and Endocrinology, Duke University, Durham, North Carolina. Research Supported by the American Federation for Aging Research, VA Medical Research Service, and NIA Academic Award A600367. Abstract was presented at the annual meeting of the American Geriatric Society, May 1989, Boston, Massachusetts. Address correspondence to Gary Kochersberger, MD, Department of Veterans Affairs Medical Center, 508 Fulton St., Durham NC 27705. 1991 by the American Geriatrics Society
METHODS Subject Selection Study subjects were recruited from a registry of veterans served by the Batavia, NY Department of Veterans Affairs medical center where the study was based. Letters were sent to these prospective subjects (individuals >60 years of age), and the first 50 who met entry critieria were enrolled. Individuals with histories of renal stones, hypercal10002-8614/91/$3.50
IAGS-FEBRUARY 1991-VOL. 39, NO. 2
EFFECTS OF CALCIUM SUPPLEMENTATION IN THE ELDERLY
cemia, or use of estrogens, thiazide diuretics, glucocorticoids, anabolic steroids, or vitamin D supplements were excluded from the study.
193
ng/mL. Total urinary hydroxyproline, measured on 2hour timed collections following an overnight fast, was assayed using a commercially available kit (Hypronosticon, Organon Teknika) employing the methodology of Kivirikko." Urine calcium and creatinine, serum creatinine, and phosphorus levels were measured by automated technique employing a colorimetric assay.
Study Design A randomized, double-blind, clinical trial served as the overall study design. Data were collected at baseline and at the end of 1 and 6 months. After collection of baseline data, subjects were randomized to receive either calcium or placebo. The calcium Statistical Analyses Independent sample f tests supplemented subjects received 1.2 grams of elemental and chi-square analyses were used to determine any calcium per day in the form of calcium carbonate (600 differences between the placebo and calcium supplemg twice each day). Subjects were instructed to take mented groups at baseline. Wilcoxon rank sum scores their calcium tablets with meals to enhance absorp- were utilized for between-group comparisons of the tion.* Pill counts and interviews were used to assess observed changes in the various biochemical paramecompliance. Dietary consumption of calcium and phos- ters, and paired f tests were performed to determine phorus were assessed through the use of 3-day diet any within-group changes effected. Pearson correlation diaries which were analyzed with a computerized nu- coefficients were calculated to determine the degree of trient analysis program with food data based upon association between continuoius variables. All values U.S.D.A. handbooks No. 8-1 through No. 8-16. Ques- given are the mean plus or minus the standard error tionnaires relating to gastrointestinal symptoms (nau- of the mean. The Bonferroni method was employed to sea, constipation, diarrhea) were completed by subjects establish appropriate significance levels when multiple at baseline and at the end of the 6-month study period. comparisons were performed; for variables measured The study took place during the months May to Octo- at 1and 6 months this meant establishing a significance ber. level of P = .025. Specimen Collection Patients were given instructions on a low collagen diet to be followed during the day prior to their labwork. Blood samples were drawn after on overnight fast. First morning urine samples were obtained from patients as well as a 2-hour timed urine collection. Serum was removed, and for analyses not performed immediately, was frozen at -80 degrees centigrade until assayed.
RESULTS
Twenty-six subjects were randomized to the calcium group and 24 to the placebo group. Eight subjects dropped out during the course of the study, four within the first 4 weeks of the study and the remainder following the data collection at the end of 1 month. Three of the dropouts were from the calcium group, and five were in the placebo group. One subject in the Biochemical Determinations Ionized calcium de- placebo group discontinued participation because of terminations were measured on serum samples (kept the development of arthritic symptoms he attributed refrigerated and anaerobic) using a Nova 2 ionized to the intervention. Another subject dropped out durcalcium analyzer (Nova Biomedical, Baltimore, MD) ing hospitalization for prostate surgery and elected not employing an ion-specific electrode. The assay has a to resume the study post-operatively. Two subjects coefficient of variation of 1%. Serum 25-hydroxyvi- moved away from the area precluding further followtamin D levels were determined by competitive protein up, and the remaining subjects (two in the calcium binding assays according to previously published group, and two in the placebo group) dropped out for methods.' The intra and interassay coefficient of vari- lack of interest in continuing participation. Baseline ation are 7.9% and 8.9%, and the reference range is characteristics did not differ significantly between the 15-80 nmol/L. Serum parathyroid hormone levels calcium and placebo groups (see Table l), nor were were determined with an intact PTH molecule assay there differences at baseline between study dropouts (Allegro Intact PTH, Nichols Institute Diagnostics, San and those who finished the study. Juan Capistrano, CA). The intra and interassay coeffiEffects on Biochemical Parameters (Tables 2 and cients of variation were 5.7% and 18%, respectively, and the reference range 13-64 pg/mL. Serum 1,25- 3) No significant difference between groups was noted dihydroxyvitamin D was measured utilizing a radio- for changes from baseline to 1 and 6 months for serum receptor assay on samples purified on a C-18 Sep-Pak ionized calcium, phosphorus, or creatinine levels. Uricartridge." The reference range is 15-60 pg/mL. Serum nary excretion of calcium did increase with supple13.1 mg/g Bone Gla protein was measured using a radioimmuno- mental calcium from a mean of 116.9 188.2 -t25.1 mg at 1 month and 173.9 k creatinine to assay (Osteocalcin, Immuno Nuclear Corp., Stillwater, MN). The intra and interassay coefficients of variation 13.8 mg at 6 months. No significant changes in urinary are 8.4% and 9.2%. The reference range is 1.8 to 6.6 calcium excretion were observed in the placebo group.
*
194
JAGS-FEBRUARY 1991-VOL. 39, NO. 2
KOCHERSBERGER ET AL
Serum PTH levels for the calcium-supplemented group decreased from a mean of 40.9 pg/mL at baseline to 34.0 pg/mL at 1 month, and 33.2 pg/mL at 6 months. For the placebo subjects, PTH levels were 44.9 at baseline, 43.8 at 1 month, and 45.3 at 6 months. 1,25 OH-Vitamin D levels also decreased significantly during calcium supplementation from a mean of 38.3 pg/ mL to 30.4 at 1 month and 31.9 at 6 months. 1,25 OHVitamin D in the placebo group was 34.9 at baseline, TABLE 1. SELECTED BASELINE CHARACTERISTICS OF STUDY GROUPS Variable Placebo ( n = 19) Calcium ( n = 23) Males/females Age (range) Dietary calcium (mg/24 hrs) 25 OH-Vit D (1580 ng/mL)
12/7 71.8 (66-81) 726 f 53
15/8 70.1 (66-83) 708 f 89
29.6 f 1.8
29.0 f 2.2
35.0 at 1 month, and 36.4 at 6 months. No differences for serum bone Gla protein levels were observed (P = .25 and .13 for changes at 1 and 6 months, respectively), but urinary excretion of hydroxyproline did decrease at 6 months (but not at 1 month) in the calcium supplemented subjects. This decrease, 25.6 to 21.2 mg hydroxyproline/g creatinine, differed from the placebo group (P = .036), but with adjustment for multiple comparisons it was not deemed statistically significant. There was no correlation observed between baseline PTH and 1,25 OH-Vitamin D levels, nor with the changes in serum levels which occurred with suppIementation. However, 1,25 OH-Vitamin D at baseline was positively correlated with hydroxyproline excretion ( r = .40, P = .008). This correlation did not persist during the supplementation period. Separate analysis of the variables of interest failed to reveal any gender effect associated with calcium
TABLE 2. BIOCHEMICAL VARIABLES FOR CALCIUM AND PLACEBO GROUPS* Six Month Baseline One Month Variable
Calcium
Placebo
Ionized calcium (4.44.81 C 0.04 4.83 C 0.04 5.3 mg/dL) Serum phosphorus 3.30 f 0.10 3.48 f 0.14 (2.6-4.9 mg/dL) Serum creatine (0.51.06 t 0.09 0.94 k 0.05 1.4 mg/dL) Urine calcium (mg/ 117 k 13 133 f 15 gram creatinine) PTH (13-64 pg/mL) 40.9 f 3.3 44.9 f 5.0 Bone Gla protein (1.82.64 t 0.42 2.29 f 0.29 6.6 ng/mL) 25.6 f 2.8 21.6 f 3.5 Urine hydroxyproline (mg/gram creatinine) 1,25 OH-Vit D (15-60 38.3 f 2.9 34.9 & 2.1 pg/m * Values given are means & SEM. ** P < ,001 Comparison with Baseline Values. *** P < ,025 Comparison with Baseline Values.
Calcium
Placebo
Calcium
Placebo
4.90 f 0.03
4.91 f 0.05
4.88 f 0.04
4.87 k 0.05
3.17 f 0.13
3.34 f 0.13
3.23 & 0.13
3.45 C 0.14
1.01 & 0.05
0.95 & 0.05
1.02 k 0.05
0.98 f 0.04
188 f 25**
133 f 14
174 f 14**
127 t 14
34.0 f 3.6*** 2.78 t 0.46
43.8 f 5.7 2.10 & 0.23
33.2 f 3.0** 2.49 t 0.39
45.3 k 4.9 2.52 t 0.28
27.2 f 4.1
20.8 f 3.6
21.2 f 4.0
21.5 f 2.5
30.4 f 2.1**
35.0 f 1.5
31.9 f 2.3***
36.4 f 2.0
TABLE 3. MEAN CHANGES FROM BASELINE VARIABLE VALUES AT ONE AND SIX MONTHS* Six Month One Month Variable Ionized Calcium Serum Phosphorus Serum Creatine Urine Calcium PTH Bone Gla Protein Urine Hydroxyproline 1,25 OH-Vit D
Calcium +0.08 -0.13 -0.06 +71.3 -6.97 +0.06 +1.61 -7.91
0.02 0.05 0.05 10.7** f 1.13 f 0.16 f 2.68 C 1.18**
C & & &
* Values given are mean differences from baseline & SEM. ** P < ,001 Comparison with Placebo Group.
Placebo +0.08 -0.15 +0.01 -0.33 +1.12 -0.19 -0.73 +0.11
f 0.02 f 0.05 f 0.01 f 8.2 f 1.94 f 0.15 f 2.09 f 1.13
Calcium f0.10 -0.07 -0.05 +57.1 -7.76 -0.05 -4.34 -6.43
f 0.04 & 0.05 f 0.05 f 8.8** & 0.96** f 0.12 f 2.44 f 1.49**
Placebo +0.05 -0.04 +0.04 -6.09 +0.38 +0.23 -0.03 +1.56
f 0.03 f 0.06 f 0.01 f 7.9 & 1.52 f 0.19 f 1.86 k 1.36
IAGS-FEBRUARY 1991-VOL. 39, NO. 2
EFFECTS OF CALCIUM SUPPLEMENTATION IN THE ELDERLY
,195
supplementation changes with the exception of urinary strate any statistically significant changes in markers calcium. The difference in mean change scores from of bone metabolism (serum bone Gla protein and uribaseline to 1 month for females given calcium was nary hydroxyproline) to support a beneficial effect greater than that for males (124 mg vs 51 mg). This from this PTH reduction. The slight reduction in uriwas due to 1female subject's tripling of urinary calcium nary hydroxyproline excretion we observed may reflect excretion, a change which largely disappeared by 6 a trend towards decreased bone resorption occurring months. When analyzed separately, calcium-supple- with calcium supplementation, and this may have bemented males still demonstrated a significant increase come more apparent with a larger sample size. Our observation that 1,25-dihydroxyvitamin D levels also in urinary calcium excretion at 1 and 6 months. Complianceand Adverse Effects Compliance was decreased with supplementation suggests a down-regcalculated by pill counts and averaged 80% for the ulation of its production in response to a more positive placebo group and 84% for the calcium-supplemented calcium balance, and it is presumed that this was group. Constipation was the most frequently reported mediated by the decline in PTH levels. Previous studies have shown variable effects of ingastrointestinal side effect. At baseline 25% of the creased calcium intake on seirum PTH levels.16r17r18 calcium-supplemented group reported experiencing Recently, Orwoll et all9 studied a group of men whose constipation at least 1 day of 7. This increased to 40% diets were supplemented with calcium and vitamin D. at the end of 6 months (Chi-square = 4.12, P = .04). For the placebo group, 31% reported symptoms of They observed a slight decline in serum PTH levels constipation at least one day of the week at baseline, without evidence of a decline in bone loss measured at peripheral and axial sites. Our study examined an older and at the end of 6 months, 36% (Chi square = .12, population (10 years and 15 years older for the calcium P = .73). For the other GI symptoms monitored (diar- and placebo groups, respectively), and we observed a rhea, nausea, "gas") there were no differences between greater reduction in serum PTH, but the lack of congroups. sistent and sizable changes in indices of bone turnover leads us to similar conclusions. It has been hypotheDISCUSSION sized that the increased serum PTH in the elderly may Calcium metabolism in humans is affected by several be necessary to ensure adequate production of 1,25changes which occur with normal aging. Studies have dihydroxyvitamin D by senescent kidneys. Silverberg demonstrated a reduction in the gastrointestinal abet a1," demonstrated that a group of eight osteoporotic sorption of calcium which is thought to be in part females had little increase in serum PTH in response related to age-related declines in the production of to phosphate-induced hypocalcemia and were inca1,25-dihydroxyvitamin D.12,13This decreased absorppable of maintaining serum 1,25-dihydroxyvitamin D tion is compounded by a tendency for the elderly to levels. Reduction in PTH may, therefore, not be a consume less calcium than younger individ~als.'~ PTH necessary or even an appropriate goal in the managelevels go up with normal aging, a trend thought to be ment of type I1 osteporosis. It would appear that whiIe compensatory for the decrease in calcium absorption. calcium supplementation in the elderly is well tolerated Deleterious effects of PTH have been speculated to and does effect some changes in mineral metabolism, play a role in the development of type I1 osteoporosis2, it remains to be demonstrated that supplemental caland it has recently been suggested that removing seacium positively influences the rate of bone loss in this sonal variation in PTH levels by maintaining an adeage group. quate vitamin D intake may prevent bone 1 0 ~ s . ' ~ Increasing dietary intake of calcium has long been REFERENCES advocated as preventive therapy for osteoporosis. 1. Riggs B, Melton L. Evidence for two distinct syndromes of Studies examining the effects of calcium supplemeninvolutional osteoporosis. Am J Med 1983;75:899-901. tation have almost exclusively been conducted in re2. Riggs B, Melton L. Involutional osteoporosis. N Engl J Med cently post-menopausal females, and their results have 1986;314:1676-1686. generally been d i ~ a p p o i n t i n g . ~The ,~,~ changes in cal- 3. Orwoll E, Meier D. Alterations in calcium, Vitamin D, and parathyroid hormone physiology in normal men with aging: cium metabolism that occur with advancing age sugRelationship to the development of senile osteoporosis. J Clin gest that these findings may not be generalizable to the Endocrinol Metab 1986;63:1262-1 269. 4. Recker R. Calcium absorption and achlorhydria. N Engl J Med elderly and that older individuals may actually be more 1985;313~70-73. likely to benefit from additional calcium. 5. Ettinger B, Genant H. Postmenopausal bone loss is prevented Our present study, as well as our previous by treatment with low dosage estrogen and calcium. Ann Intern Med 1987;106:40-45. have demonstrated that calcium supplementation is 6. Nilas L, Christiansen C, Rodbro P. Calcium supplementation well tolerated by the elderly, and does cause a reducand postmenopausal bone loss. Elr Med J 1984;289:1103-1106. tion of serum PTH levels, an effect which appears to 7. Kochersberger G, Bales C, Lobaugh B, Lyles K. Calcium supplementation lowers serum parathyroid hormone levels in elderly last at least 6 months. We failed, however, to demon-
196
KOCHERSBERGER ET AL
subjects. J Gerontol: Med Sci 1990;45:159-162. 8. Recker R. Calcium absorption and achlorhydria. N Engl J Med 1985;313:70-73. 9. Lyles KW, Drezner MK. Parathyroid hormone effects on serum 1,25 Dihydroxyvitamin D levels in patients with X-linked hypophosphatemic rickets: Evidence for abnormal 25-hydroxyvitamin D-1-hydroxylase activity. J Clin Endocrinol Metab 1982; 54~638-644. 10. Reinhardt T, Horst R, Orf J. A microassay for 1,25-dihyroxyvitamin D not requiring high performance liquid chromatography: application to clinical studies. J Clin Endocrinol Metab 1984;58: 91-98. 11. Kivirikko K, Laitinen 0, Prockop D. Modifications of a specific assay for hydroxyproline in urine. Anal Biochem 1967;19:249255. 12. Gallagher J, Riggs B. Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients: Effect of age and dietary calcium. J Clin Invest 1979;64: 729-736. 13. Aloia J, Vaswani A, Yeh J. Calcitriol in the treatment of postmenopausal osteoporosis. Am J Med 1988;84:401-408. 14. Abraham S, Johnson C. Dietary intake findings, United States
IAGS-FEBRUARY 1991-VOL. 39, NO. 2
15. 16.
17.
18. 19. 20.
1976-1980. Hyattsville, MD: National Center for Health Statistics, US Department of Human Services, 1982. Krall E, Sahyoun N, Tannenbaum S. Effect of vitamin D intake on seasonal variations in parathyroid hormone secretion in postmenopausal women. N Engl J Med 1989;321:1777-1783. Bikle D, Herman R, Halloran B. Adaptive response of humans to changes in dietary calcium: relationship between vitamin D regulated intestinal function and serum 1,25-&hydroxyvitamin D levels. Gastroenterology 1983;84:314-323. Baran D, Sorensen A, Roche J. Dietary modification with dairy products for preventing vertebral bone loss in premenopausal women: A three-year prospective study. J C l i Endocrinol Metab 1990;70:264-270. Chapuy M, Chapuy, ' I Mounier P. Calcium and vitamin D supplements effects on calcium metabolism in elderly people. Amer J Clin Nutr 1987;46:324-328. Orwoll E, Oviatt S, McClung M. The rate of bone mineral loss in normal men and the effects of calcium and cholecalciferol supplementation. Ann Intern Med 1990;112:29-34. Silverberg S, Shane E, Bilezikian J. Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D formation in women with osteoporosis. N Engl J Med 1989;320:277-281.
The Metabolic Effects of Calcium Supplementation in the Elderly Gary Kochersberger,MD,* Ronald WestZund,MS,t Kenneth W.Lyles,MDS
Because the value of calcium supplementation in agerelated bone loss is controversial, we conducted a study to examine the effects of six months of calcium supplementation on parathyroid hormone (PTH) and other indices of bone metabolism. Calcium carbonate (1.2 grams elemental calciumlday) or placebo was administered for 6 months to a group of 42 healthy, ambulatory men and women with a mean age of 71. Fasting blood samples and 2-hour urine collections were performed at baseline and at 3 and 6 months. The decline in serum PTH levels from baseline in the calcium-supplemented group was significant in comparison with the placebo group changes (for the calcium group: 40.9 pg/mL at baseline to 34.0 pglmL at 1 month and 33.2 at 6 months). This drop in PTH was associated with a decline in 1,25 OH-Vitamin
D levels from a mean of 38.3 pglmL to 30.4 pglmL at 1 month and 32.9 at 6 months. During calcium supplementation, no changes were observed in serum bone Gla protein values. A decline in urinary hydroxyproline excretion was observed at 6 months, but this did not reach significant levels. The present study demonstrates suppression of PTH levels with calcium supplementation, an effect which lasts at least 6 months. This change was accompanied by a decrease in serum 1,25 OH-Vitamin D. The lack of significant changes in either serum bone gla protein or urine hydroxyproline excretion fails to support any significant change in bone turnover occurring with this decrease in PTH. J Am Geriatr SOC39:192-196, 1991
he differences observed in the epidemiology and suspected pathogenesis of osteoporosis has led to the subdivision of the disease into type I (post-menopausal) and type I1 (senile).' The cause of type I osteoporosis is not fully understood, but is felt to be related to a decline in osteoblast activity and an observed increase in circulating levels of parathyroid hormone (PTH).*This increase in PTH is felt to be secondary to an age related decline in the formation of 1,25-dihydroxyvitamin D and a decrease in intestinal calcium absorption in part caused by the reduced 1,25 dihydroxyvitamin D3.
The use of calcium supplements in the prevention of osteoporosis has been extensively studied in numerous clinical trial^.^'^' ' For the most part, however, these studies have only examined women in the immediate post-menopausal period, who are at risk for type I osteoporosis. We recently examined the effects of calcium supplementation in a group of elderly men and women at risk for type I1 osteopor~sis.~ In that study we demonstrated that 4 weeks of calcium supplementation was capable of significantly reducing serum levels of immunoreactive PTH, suggesting possible utility in the management of type I1 disease. A study was therefore conducted to expand upon this earlier work and to look at the effects of longer duration supplementation including biochemical parameters felt to be reflective of bone turnover.
T
From the * Department of Veterans Affairs Medical Centers, Batavia and Buffalo, New York and the Department of Medicine, Division of Geriatrics, University of Buffalo, Buffalo, New York; t the Center for the Study of Aging and Human Development, Duke University, Durham, North Carolina; and $ GRECC, Department of Veterans Affairs Medical Center, and the Department of Medicine, Divisions of Geriatrics and Endocrinology, Duke University, Durham, North Carolina. Research Supported by the American Federation for Aging Research, VA Medical Research Service, and NIA Academic Award A600367. Abstract was presented at the annual meeting of the American Geriatric Society, May 1989, Boston, Massachusetts. Address correspondence to Gary Kochersberger, MD, Department of Veterans Affairs Medical Center, 508 Fulton St., Durham NC 27705. 1991 by the American Geriatrics Society
METHODS Subject Selection Study subjects were recruited from a registry of veterans served by the Batavia, NY Department of Veterans Affairs medical center where the study was based. Letters were sent to these prospective subjects (individuals >60 years of age), and the first 50 who met entry critieria were enrolled. Individuals with histories of renal stones, hypercal10002-8614/91/$3.50
IAGS-FEBRUARY 1991-VOL. 39, NO. 2
EFFECTS OF CALCIUM SUPPLEMENTATION IN THE ELDERLY
cemia, or use of estrogens, thiazide diuretics, glucocorticoids, anabolic steroids, or vitamin D supplements were excluded from the study.
193
ng/mL. Total urinary hydroxyproline, measured on 2hour timed collections following an overnight fast, was assayed using a commercially available kit (Hypronosticon, Organon Teknika) employing the methodology of Kivirikko." Urine calcium and creatinine, serum creatinine, and phosphorus levels were measured by automated technique employing a colorimetric assay.
Study Design A randomized, double-blind, clinical trial served as the overall study design. Data were collected at baseline and at the end of 1 and 6 months. After collection of baseline data, subjects were randomized to receive either calcium or placebo. The calcium Statistical Analyses Independent sample f tests supplemented subjects received 1.2 grams of elemental and chi-square analyses were used to determine any calcium per day in the form of calcium carbonate (600 differences between the placebo and calcium supplemg twice each day). Subjects were instructed to take mented groups at baseline. Wilcoxon rank sum scores their calcium tablets with meals to enhance absorp- were utilized for between-group comparisons of the tion.* Pill counts and interviews were used to assess observed changes in the various biochemical paramecompliance. Dietary consumption of calcium and phos- ters, and paired f tests were performed to determine phorus were assessed through the use of 3-day diet any within-group changes effected. Pearson correlation diaries which were analyzed with a computerized nu- coefficients were calculated to determine the degree of trient analysis program with food data based upon association between continuoius variables. All values U.S.D.A. handbooks No. 8-1 through No. 8-16. Ques- given are the mean plus or minus the standard error tionnaires relating to gastrointestinal symptoms (nau- of the mean. The Bonferroni method was employed to sea, constipation, diarrhea) were completed by subjects establish appropriate significance levels when multiple at baseline and at the end of the 6-month study period. comparisons were performed; for variables measured The study took place during the months May to Octo- at 1and 6 months this meant establishing a significance ber. level of P = .025. Specimen Collection Patients were given instructions on a low collagen diet to be followed during the day prior to their labwork. Blood samples were drawn after on overnight fast. First morning urine samples were obtained from patients as well as a 2-hour timed urine collection. Serum was removed, and for analyses not performed immediately, was frozen at -80 degrees centigrade until assayed.
RESULTS
Twenty-six subjects were randomized to the calcium group and 24 to the placebo group. Eight subjects dropped out during the course of the study, four within the first 4 weeks of the study and the remainder following the data collection at the end of 1 month. Three of the dropouts were from the calcium group, and five were in the placebo group. One subject in the Biochemical Determinations Ionized calcium de- placebo group discontinued participation because of terminations were measured on serum samples (kept the development of arthritic symptoms he attributed refrigerated and anaerobic) using a Nova 2 ionized to the intervention. Another subject dropped out durcalcium analyzer (Nova Biomedical, Baltimore, MD) ing hospitalization for prostate surgery and elected not employing an ion-specific electrode. The assay has a to resume the study post-operatively. Two subjects coefficient of variation of 1%. Serum 25-hydroxyvi- moved away from the area precluding further followtamin D levels were determined by competitive protein up, and the remaining subjects (two in the calcium binding assays according to previously published group, and two in the placebo group) dropped out for methods.' The intra and interassay coefficient of vari- lack of interest in continuing participation. Baseline ation are 7.9% and 8.9%, and the reference range is characteristics did not differ significantly between the 15-80 nmol/L. Serum parathyroid hormone levels calcium and placebo groups (see Table l), nor were were determined with an intact PTH molecule assay there differences at baseline between study dropouts (Allegro Intact PTH, Nichols Institute Diagnostics, San and those who finished the study. Juan Capistrano, CA). The intra and interassay coeffiEffects on Biochemical Parameters (Tables 2 and cients of variation were 5.7% and 18%, respectively, and the reference range 13-64 pg/mL. Serum 1,25- 3) No significant difference between groups was noted dihydroxyvitamin D was measured utilizing a radio- for changes from baseline to 1 and 6 months for serum receptor assay on samples purified on a C-18 Sep-Pak ionized calcium, phosphorus, or creatinine levels. Uricartridge." The reference range is 15-60 pg/mL. Serum nary excretion of calcium did increase with supple13.1 mg/g Bone Gla protein was measured using a radioimmuno- mental calcium from a mean of 116.9 188.2 -t25.1 mg at 1 month and 173.9 k creatinine to assay (Osteocalcin, Immuno Nuclear Corp., Stillwater, MN). The intra and interassay coefficients of variation 13.8 mg at 6 months. No significant changes in urinary are 8.4% and 9.2%. The reference range is 1.8 to 6.6 calcium excretion were observed in the placebo group.
*
194
JAGS-FEBRUARY 1991-VOL. 39, NO. 2
KOCHERSBERGER ET AL
Serum PTH levels for the calcium-supplemented group decreased from a mean of 40.9 pg/mL at baseline to 34.0 pg/mL at 1 month, and 33.2 pg/mL at 6 months. For the placebo subjects, PTH levels were 44.9 at baseline, 43.8 at 1 month, and 45.3 at 6 months. 1,25 OH-Vitamin D levels also decreased significantly during calcium supplementation from a mean of 38.3 pg/ mL to 30.4 at 1 month and 31.9 at 6 months. 1,25 OHVitamin D in the placebo group was 34.9 at baseline, TABLE 1. SELECTED BASELINE CHARACTERISTICS OF STUDY GROUPS Variable Placebo ( n = 19) Calcium ( n = 23) Males/females Age (range) Dietary calcium (mg/24 hrs) 25 OH-Vit D (1580 ng/mL)
12/7 71.8 (66-81) 726 f 53
15/8 70.1 (66-83) 708 f 89
29.6 f 1.8
29.0 f 2.2
35.0 at 1 month, and 36.4 at 6 months. No differences for serum bone Gla protein levels were observed (P = .25 and .13 for changes at 1 and 6 months, respectively), but urinary excretion of hydroxyproline did decrease at 6 months (but not at 1 month) in the calcium supplemented subjects. This decrease, 25.6 to 21.2 mg hydroxyproline/g creatinine, differed from the placebo group (P = .036), but with adjustment for multiple comparisons it was not deemed statistically significant. There was no correlation observed between baseline PTH and 1,25 OH-Vitamin D levels, nor with the changes in serum levels which occurred with suppIementation. However, 1,25 OH-Vitamin D at baseline was positively correlated with hydroxyproline excretion ( r = .40, P = .008). This correlation did not persist during the supplementation period. Separate analysis of the variables of interest failed to reveal any gender effect associated with calcium
TABLE 2. BIOCHEMICAL VARIABLES FOR CALCIUM AND PLACEBO GROUPS* Six Month Baseline One Month Variable
Calcium
Placebo
Ionized calcium (4.44.81 C 0.04 4.83 C 0.04 5.3 mg/dL) Serum phosphorus 3.30 f 0.10 3.48 f 0.14 (2.6-4.9 mg/dL) Serum creatine (0.51.06 t 0.09 0.94 k 0.05 1.4 mg/dL) Urine calcium (mg/ 117 k 13 133 f 15 gram creatinine) PTH (13-64 pg/mL) 40.9 f 3.3 44.9 f 5.0 Bone Gla protein (1.82.64 t 0.42 2.29 f 0.29 6.6 ng/mL) 25.6 f 2.8 21.6 f 3.5 Urine hydroxyproline (mg/gram creatinine) 1,25 OH-Vit D (15-60 38.3 f 2.9 34.9 & 2.1 pg/m * Values given are means & SEM. ** P < ,001 Comparison with Baseline Values. *** P < ,025 Comparison with Baseline Values.
Calcium
Placebo
Calcium
Placebo
4.90 f 0.03
4.91 f 0.05
4.88 f 0.04
4.87 k 0.05
3.17 f 0.13
3.34 f 0.13
3.23 & 0.13
3.45 C 0.14
1.01 & 0.05
0.95 & 0.05
1.02 k 0.05
0.98 f 0.04
188 f 25**
133 f 14
174 f 14**
127 t 14
34.0 f 3.6*** 2.78 t 0.46
43.8 f 5.7 2.10 & 0.23
33.2 f 3.0** 2.49 t 0.39
45.3 k 4.9 2.52 t 0.28
27.2 f 4.1
20.8 f 3.6
21.2 f 4.0
21.5 f 2.5
30.4 f 2.1**
35.0 f 1.5
31.9 f 2.3***
36.4 f 2.0
TABLE 3. MEAN CHANGES FROM BASELINE VARIABLE VALUES AT ONE AND SIX MONTHS* Six Month One Month Variable Ionized Calcium Serum Phosphorus Serum Creatine Urine Calcium PTH Bone Gla Protein Urine Hydroxyproline 1,25 OH-Vit D
Calcium +0.08 -0.13 -0.06 +71.3 -6.97 +0.06 +1.61 -7.91
0.02 0.05 0.05 10.7** f 1.13 f 0.16 f 2.68 C 1.18**
C & & &
* Values given are mean differences from baseline & SEM. ** P < ,001 Comparison with Placebo Group.
Placebo +0.08 -0.15 +0.01 -0.33 +1.12 -0.19 -0.73 +0.11
f 0.02 f 0.05 f 0.01 f 8.2 f 1.94 f 0.15 f 2.09 f 1.13
Calcium f0.10 -0.07 -0.05 +57.1 -7.76 -0.05 -4.34 -6.43
f 0.04 & 0.05 f 0.05 f 8.8** & 0.96** f 0.12 f 2.44 f 1.49**
Placebo +0.05 -0.04 +0.04 -6.09 +0.38 +0.23 -0.03 +1.56
f 0.03 f 0.06 f 0.01 f 7.9 & 1.52 f 0.19 f 1.86 k 1.36
IAGS-FEBRUARY 1991-VOL. 39, NO. 2
EFFECTS OF CALCIUM SUPPLEMENTATION IN THE ELDERLY
,195
supplementation changes with the exception of urinary strate any statistically significant changes in markers calcium. The difference in mean change scores from of bone metabolism (serum bone Gla protein and uribaseline to 1 month for females given calcium was nary hydroxyproline) to support a beneficial effect greater than that for males (124 mg vs 51 mg). This from this PTH reduction. The slight reduction in uriwas due to 1female subject's tripling of urinary calcium nary hydroxyproline excretion we observed may reflect excretion, a change which largely disappeared by 6 a trend towards decreased bone resorption occurring months. When analyzed separately, calcium-supple- with calcium supplementation, and this may have bemented males still demonstrated a significant increase come more apparent with a larger sample size. Our observation that 1,25-dihydroxyvitamin D levels also in urinary calcium excretion at 1 and 6 months. Complianceand Adverse Effects Compliance was decreased with supplementation suggests a down-regcalculated by pill counts and averaged 80% for the ulation of its production in response to a more positive placebo group and 84% for the calcium-supplemented calcium balance, and it is presumed that this was group. Constipation was the most frequently reported mediated by the decline in PTH levels. Previous studies have shown variable effects of ingastrointestinal side effect. At baseline 25% of the creased calcium intake on seirum PTH levels.16r17r18 calcium-supplemented group reported experiencing Recently, Orwoll et all9 studied a group of men whose constipation at least 1 day of 7. This increased to 40% diets were supplemented with calcium and vitamin D. at the end of 6 months (Chi-square = 4.12, P = .04). For the placebo group, 31% reported symptoms of They observed a slight decline in serum PTH levels constipation at least one day of the week at baseline, without evidence of a decline in bone loss measured at peripheral and axial sites. Our study examined an older and at the end of 6 months, 36% (Chi square = .12, population (10 years and 15 years older for the calcium P = .73). For the other GI symptoms monitored (diar- and placebo groups, respectively), and we observed a rhea, nausea, "gas") there were no differences between greater reduction in serum PTH, but the lack of congroups. sistent and sizable changes in indices of bone turnover leads us to similar conclusions. It has been hypotheDISCUSSION sized that the increased serum PTH in the elderly may Calcium metabolism in humans is affected by several be necessary to ensure adequate production of 1,25changes which occur with normal aging. Studies have dihydroxyvitamin D by senescent kidneys. Silverberg demonstrated a reduction in the gastrointestinal abet a1," demonstrated that a group of eight osteoporotic sorption of calcium which is thought to be in part females had little increase in serum PTH in response related to age-related declines in the production of to phosphate-induced hypocalcemia and were inca1,25-dihydroxyvitamin D.12,13This decreased absorppable of maintaining serum 1,25-dihydroxyvitamin D tion is compounded by a tendency for the elderly to levels. Reduction in PTH may, therefore, not be a consume less calcium than younger individ~als.'~ PTH necessary or even an appropriate goal in the managelevels go up with normal aging, a trend thought to be ment of type I1 osteporosis. It would appear that whiIe compensatory for the decrease in calcium absorption. calcium supplementation in the elderly is well tolerated Deleterious effects of PTH have been speculated to and does effect some changes in mineral metabolism, play a role in the development of type I1 osteoporosis2, it remains to be demonstrated that supplemental caland it has recently been suggested that removing seacium positively influences the rate of bone loss in this sonal variation in PTH levels by maintaining an adeage group. quate vitamin D intake may prevent bone 1 0 ~ s . ' ~ Increasing dietary intake of calcium has long been REFERENCES advocated as preventive therapy for osteoporosis. 1. Riggs B, Melton L. Evidence for two distinct syndromes of Studies examining the effects of calcium supplemeninvolutional osteoporosis. Am J Med 1983;75:899-901. tation have almost exclusively been conducted in re2. Riggs B, Melton L. Involutional osteoporosis. N Engl J Med cently post-menopausal females, and their results have 1986;314:1676-1686. generally been d i ~ a p p o i n t i n g . ~The ,~,~ changes in cal- 3. Orwoll E, Meier D. Alterations in calcium, Vitamin D, and parathyroid hormone physiology in normal men with aging: cium metabolism that occur with advancing age sugRelationship to the development of senile osteoporosis. J Clin gest that these findings may not be generalizable to the Endocrinol Metab 1986;63:1262-1 269. 4. Recker R. Calcium absorption and achlorhydria. N Engl J Med elderly and that older individuals may actually be more 1985;313~70-73. likely to benefit from additional calcium. 5. Ettinger B, Genant H. Postmenopausal bone loss is prevented Our present study, as well as our previous by treatment with low dosage estrogen and calcium. Ann Intern Med 1987;106:40-45. have demonstrated that calcium supplementation is 6. Nilas L, Christiansen C, Rodbro P. Calcium supplementation well tolerated by the elderly, and does cause a reducand postmenopausal bone loss. Elr Med J 1984;289:1103-1106. tion of serum PTH levels, an effect which appears to 7. Kochersberger G, Bales C, Lobaugh B, Lyles K. Calcium supplementation lowers serum parathyroid hormone levels in elderly last at least 6 months. We failed, however, to demon-
196
KOCHERSBERGER ET AL
subjects. J Gerontol: Med Sci 1990;45:159-162. 8. Recker R. Calcium absorption and achlorhydria. N Engl J Med 1985;313:70-73. 9. Lyles KW, Drezner MK. Parathyroid hormone effects on serum 1,25 Dihydroxyvitamin D levels in patients with X-linked hypophosphatemic rickets: Evidence for abnormal 25-hydroxyvitamin D-1-hydroxylase activity. J Clin Endocrinol Metab 1982; 54~638-644. 10. Reinhardt T, Horst R, Orf J. A microassay for 1,25-dihyroxyvitamin D not requiring high performance liquid chromatography: application to clinical studies. J Clin Endocrinol Metab 1984;58: 91-98. 11. Kivirikko K, Laitinen 0, Prockop D. Modifications of a specific assay for hydroxyproline in urine. Anal Biochem 1967;19:249255. 12. Gallagher J, Riggs B. Intestinal calcium absorption and serum vitamin D metabolites in normal subjects and osteoporotic patients: Effect of age and dietary calcium. J Clin Invest 1979;64: 729-736. 13. Aloia J, Vaswani A, Yeh J. Calcitriol in the treatment of postmenopausal osteoporosis. Am J Med 1988;84:401-408. 14. Abraham S, Johnson C. Dietary intake findings, United States
IAGS-FEBRUARY 1991-VOL. 39, NO. 2
15. 16.
17.
18. 19. 20.
1976-1980. Hyattsville, MD: National Center for Health Statistics, US Department of Human Services, 1982. Krall E, Sahyoun N, Tannenbaum S. Effect of vitamin D intake on seasonal variations in parathyroid hormone secretion in postmenopausal women. N Engl J Med 1989;321:1777-1783. Bikle D, Herman R, Halloran B. Adaptive response of humans to changes in dietary calcium: relationship between vitamin D regulated intestinal function and serum 1,25-&hydroxyvitamin D levels. Gastroenterology 1983;84:314-323. Baran D, Sorensen A, Roche J. Dietary modification with dairy products for preventing vertebral bone loss in premenopausal women: A three-year prospective study. J C l i Endocrinol Metab 1990;70:264-270. Chapuy M, Chapuy, ' I Mounier P. Calcium and vitamin D supplements effects on calcium metabolism in elderly people. Amer J Clin Nutr 1987;46:324-328. Orwoll E, Oviatt S, McClung M. The rate of bone mineral loss in normal men and the effects of calcium and cholecalciferol supplementation. Ann Intern Med 1990;112:29-34. Silverberg S, Shane E, Bilezikian J. Abnormalities in parathyroid hormone secretion and 1,25-dihydroxyvitamin D formation in women with osteoporosis. N Engl J Med 1989;320:277-281.
