Volume 115 Number 7
1 October 1991
Annals of Internal Medicine
Effect of Vitamin D Supplementation on Wintertime and Overall Bone Loss in Healthy Postmenopausal Women Bess Dawson-Hughes, MD; Gerard E. Dallal, PhD; Elizabeth A. Krall, PhD; Susan Harris, MS; Lori J. Sokoll, MCC; and Gladys Falconer, MS
• Objectives: To determine whether relative vitamin D deficiency during the winter months contributes to age-related bone loss and whether rates of change in hard- and soft-tissue mass vary during the year. • Design: Double-blind, placebo-controlled, 1-year trial in 249 women in which equal numbers of women were randomized to either placebo or 400 IU of vitamin D daily. All women received 377 mg/d of supplemental calcium largely as calcium citrate malate. • Patients: Healthy, ambulatory postmenopausal women with usual intakes of vitamin D of 100 lU/d. • Measurements: Duplicate spine and whole-body scans were done by dual energy x-ray absorptiometry at 6-month intervals that were timed to periods when 25hydroxyvitamin D levels were highest and lowest. Period 1 was June-July to December-January and period 2 was December-January to the next June-July. Serum parathyroid hormone and plasma 25-hydroxyvitamin D levels were measured during periods 1 and 2. • Main Results: In the placebo group, spinal bone mineral density increased in period 1, decreased in period 2, and sustained no net change. Women treated with vitamin D had a similar spinal increase in period 1 (1.46% compared with 1.40% in placebo), less loss in period 2 ( - 0.54% compared with - 1.22%, CI for the difference, 0.05% to 1.31 %, P = 0.032) and a significant overall benefit (0.85% compared with 0.15%, CI for the difference, 0.03% to 1.37%, F = 0.04). In period 2, 25-hydroxyvitamin D levels were lower and parathyroid hormone levels were higher in the placebo than in the vitamin D group. Whole-body lean and fat tissue and bone mineral density varied during the year but did not change overall. • Conclusions: At latitude 42 degrees, healthy postmenopausal women with vitamin D intakes of 100 IU daily can significantly reduce late wintertime bone loss and improve net bone density of the spine over one year by increasing their intake of vitamin D to 500 IU daily. A long-term benefit of preventing vitamin D insufficiency in the winter seems likely although it remains to be shown. Observed changes in bone as well as in fat and lean tissue appear to be related to season. Annuls of Internal Medicine . 1991;115:505-512. 1 October 1991 •
From the U.S. Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston, Massachusetts. For current author Addresses, see end of text.
In 1964, Smith and colleagues (1) reported that serum antirachitic activity varied with season in Michigan but not in Puerto Rico. More recently, it has been recognized that serum parathyroid hormone and plasma 25hydroxyvitamin D (25[OH]D) levels vary inversely with one another and that the concentration of each changes with season in the temperate zone (2-5). In a crosssectional study of 333 healthy ambulatory postmenopausal women in Massachusetts (6), serum parathyroid hormone was lowest in those women tested between August and October and highest in those tested between March and May. When examined in relation to vitamin D intake, however, a higher parathyroid hormone level between March and May was present only in those women who had vitamin D intakes of less than 220 IU daily. Because of the relatively small number of women with intakes over this level, we were unable to be certain that the vitamin D intake associated with no wintertime increase in parathyroid hormone was not higher than 220 IU (6). There is indirect evidence that seasonal increases in parathyroid hormone may affect bone mass adversely. Patients with hip fractures have been shown to have wintertime increases in parathyroid hormone and lower levels of 25(OH)D than controls (7). Calcium balance in Scandinavian prisoners was more negative in winter than in summer (8, 9). Hard- and soft-tissue composition changes with season. Seasonal variation in bone mineral content of the metacarpals (10) and forearm (11) and in density of the spine (12, 13) has been reported in women residing in northern parts of Europe and the United States. Vitamin D deficiency and decreased levels of physical activity were cited as possible explanations for decreases in metacarpal bone mineral content from winter to summer (10). However, others (14) have found no influence of season on bone mass. Seasonal variation in softtissue composition has been documented in small ani• Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
505
mals (15, 16) and also in some (17), but not in other (18, 19), studies in humans. In the past, the methods used to measure body fat and lean tissue have been unsuitable for detecting small changes and for studying large populations because they were labor intensive, imprecise, and sometimes associated with significant radiation exposure. The newly developed dual-energy x-ray absorptiometry method overcomes these obstacles and provides an accurate and reproducible measurement of soft-tissue composition when compared with established methods (20). Our randomized, placebo-controlled, double-blind 1-year study was done to determine the effect of increased vitamin D intake on rates of bone loss during the 6-month periods when parathyroid hormone levels are highest and lowest (6). A second objective was to compare rates of change in fat and lean tissue during these two intervals. The first interval was between June or July and December or January (parathyroid hormone at its peak) and the second was between December or January and the following June or July (parathyroid hormone at its lowest). Because of the documented benefit of increasing calcium intake to 800 mg daily (21), all women received calcium supplements. Calcium citrate malate was used because it is well absorbed (22) and is effective in reducing bone loss in this population (21). Methods The 276 postmenopausal women enrolled in this study met the following entry criteria: white race, good general health, normal ambulation, and normal results on screening laboratory tests. Twenty-four women had menopause within the previous 5 years (mean, 3.9 ± 0.13 [SE] years), and 252 women had menopause 6 or more years previously. All had completed a 2-year calcium supplement trial at the center (21). For the previous trial, selection criteria included calcium intake under 650 mg and absence of compression fracture on thoracic and lumbar-spine radiography. Exclusion criteria for this study included a history of nontraumatic fracture of any bone; renal, hepatic, or gastrointestinal disorders associated with abnormal calcium or bone metabolism; use of estrogen, glucocorticoids, anticonvulsants, or other medications known to affect calcium or bone metabolism; or spine bone-mineral-density values of 2 SD or more below the age-matched reference mean. The pro-
tocol was approved by the Human Investigation Review Committee of Tufts University, and written informed consent was obtained from each woman. Study Design and Supplements In this 12-month, double-blind, placebo-controlled trial, women were randomly assigned to treatment with 400 IU of vitamin D or placebo. Both the vitamin D and placebo tablets contained 127 mg of elemental calcium as calcium phosphate. In addition, all subjects received 250 mg of elemental calcium as calcium citrate malate daily so that each had a total of 377 mg of supplemental calcium per day. Women were randomized to treatment after being stratified by dietary calcium intake (less than 400 mg/d or 400 to 650 mg/d), treatment group in the previous trial (placebo or 500 mg of elemental calcium as either calcium citrate malate or calcium carbonate), and previous category of years since menopause (greater or fewer than 6 years). The mean interval between the previous and current trials was 0.42 ± 0.31 years, and it did not vary by treatment group. In this interval, women were counseled to ingest 800 mg of calcium daily. During the study, the women were instructed to take the tablets at bedtime, to maintain their usual diets, and to avoid taking calcium and vitamin D supplements other than those provided. Each woman came to the center three times at 6-month intervals (first in June or July, next in December or January, and finally in June or July) for measurements of spine and whole-body bone-mineral density, whole-body soft-tissue composition, height, and weight. The medical history was assessed on the first and last visits. In the periods of August through November and February through May, each woman was assessed for calcium and vitamin D intake, physical activity, and quadriceps strength; and had blood and urine measurements. During the last four visits, pill compliance was recorded. Calcium citrate malate was prepared as previously described (22) with a calcium:citrate:malate molar ratio of 6:2:3. This calcium source is commercially available only in fortified citrus juices (Citrus Hill Plus Calcium, The Procter and Gamble Company, Cincinnati, Ohio). Measurements Dietary calcium and vitamin D intakes were estimated by means of a food-frequency questionnaire that was administered by a trained technician (6). Food models and household measuring tools were used to help subjects estimate portion sizes. Alcohol use and cigarette smoking were also assessed by questionnaire. Travel to lower latitudes was documented on a questionnaire and a travel index calculated as days traveled divided by latitude x 100. Current physical activity was assessed by use of a questionnaire adapted from Kriska and colleagues (23) and
Table 1. Clinical Characteristics at Enrollment of Women Who Completed the Trial* Characteristic
Treatment Group> Placebo
Number Age, y Time since menopause^, y < 6 y since menopause, % Weight,^ Smokers, % Alcohol consumption, glwk Bone mineral density, g/ctn (A?)§ Spine Whole body
125 61.9 13.4 18.4 67.4 7.2 44.8
±0.5 ± 0.6 ± 1.0 ± 6.6
1.03 ± 0.01 (110) 1.05 ± 0.01 (125)
P Valuet Vitamin D
124 61.4 13.5 9.7 68.5 8.1 32.1
± 0.5 ± 0.6 ± 1.1 ± 4.8
1.05 ± 0.02(110) 1.07 ± 0 . 0 1 (121)
> 0.2 >0.2 0.07 > 0.2 >0.2 0.12 > 0.2 0.08
* Plus-minus values are means ± SE. t For two-sample Mest (or Pearson chi-square test in the case of percentage of women less than 6 years since menopause and of percentage of smokers). X Year of menopause was unknown for 11 women (vitamin D supplement group, 5; and placebo group, 6) and was assumed to have occurred at age 50. § Number of women with complete data is given in parentheses. 506
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
Table 2. Mean Change in Bone Mineral Density during Treatment with Vitamin D or Placebo* Site and Treatment
Idumber of Women
Period 1 (June-July through December-January) % Change/6 Months
Period 2 (DecemberJanuary through June-July) % Change/6 Months
< Spine Placebo Vitamin D Difference Whole body Placebo Vitamin D Difference
Differencet
Overall % Change/Year
%(95%CI)
>
110 110
1.40 (0.97 to 1.83) 1.46 (0.99 to 1.93) 0.06 (- 0.58 to 0.70)
- 1.22 (- 1.69 to - 0.75) - 0.54 (- 0.95 to - 0.13) 0.68 (0.05 to 1.31)*
- 2.63 (- 3.38 to - 1.88) - 2.00 (- 2.75 to - 1.25)
0.15 (- 0.34 to 0.64) 0.85 (0.40 to 1.30) 0.70 (0.03 to 1.37)§
125 121
0.61 (0.37 to 0.85) 0.62 (0.40 to 0.84) 0.01 (- 0.31 to 0.33)
- 0.66 (- 0.86 to - 0.46) - 0.56 (- 0.78 to - 0.34) 0.10 (- 0.19 to 0.39)
- 1.27 (- 1.64 to - 0.90) - 1.18 (- 1.54 to - 0.82)
- 0.08 (- 0.30 to 0.14) 0.03 (- 0.21 to 0.27) 0.11 (- 0.21 to 0.43)
* Values are means with 95% confidence intervals (CI). t Rate of change in period 2 differs from rate in period 1 in all instances, P < 0.001. X Treatment groups differ at P = 0.03. § Treatment groups differ at P = 0.04.
reported as kilocalories per day. Quadriceps muscle strength of the dominant leg was assessed with use of a Thigh-Knee Dynamic (Universal, Cedar Rapids, Iowa). Each woman, in the sitting position, pushed a padded bar forward with the lower leg (at the level of the distal tibia). The maximum weight pushed by each leg was recorded. Bone-mineral density of the spine (L-2 to L-4) and whole skeleton and whole-body soft-tissue composition were measured with a model DPX dual-energy x-ray absorptiometer (Lunar Radiation, Madison, Wisconsin). Coefficients of variation were 1% for the spine and 0.6% for the whole-skeleton bone density measurements, 2% for fat, and 1% for lean body mass, as described previously (24). The scans were analyzed with Lunar Radiation Corporation software version 3.10. On each of three visits, women had two spine and whole-body scans, with repositioning between the two scans. The mean of the two bone density and body composition values at each time point was used. An aluminum phantom was scanned biweekly in a mixture of oil:water (30:70) at total thicknesses ranging from 15.2 cm to 27.9 cm (24) over the last 9 months of the study. During this time, the bone-mineral density of the phantom was stable. In addition, bone mineral density of the phantom scanned in 15.2 to 25.4 cm of water was stable throughout the study (24). Throughout the physiologic thickness range of 15.2 to 27.9 cm, phantom bone mineral density was a mean of 0.6% lower when the phantom was scanned in oil/water (30:70) than in water alone (24). Intact parathyroid hormone in serum was measured with Allegro Intact PTH kits obtained from Nichols Institute (San Juan Capistrano, California) with intra-assay and interassay coefficients of variation of 5.6% and 6.6%, respectively. Plasma levels of 1,25-dihydroxyvitamin D (l,25(OH)2D) were measured by the competitive protein-binding method of Reinhardt and colleagues (25) with intra-assay and interassay coefficients of variation of 4.9% and 7.7%, respectively. Plasma levels of 25(OH)D were measured by the method of Preece and colleagues (26) with intra-assay and interassay coefficients of variation of 5.0% and 7.3%, respectively. The normal range for adults, as determined in 75 healthy persons from 18 to 65 years of age, was 20 to 138 nmol/L. Serum ionized calcium levels were measured with a Nova 7 analyzer (Nova Biomedical, Waltham, Massachusetts); serum alkaline phosphatase, phosphorus, and creatinine levels were measured by colorimetry with a Cobas Mira; and urine creatinine levels were with a Cobas Fara centrifugal analyzer (Roche Instruments, Belleville, New Jersey). Creatinine clearance was computed and adjusted for body-surface area (27). Urinary calcium was measured by direct-current plasma emission spectroscopy with a Spectrascan 6 (Beckman Instruments, Palo Alto, California). In women receiving thyroid hormone therapy, levels of thyroid-stimulating hormone were measured with Allegro HS-TSH kits obtained from Nichols Institute. Analyses were done in batches as the samples were col-
lected, except for serum levels of thyroid-stimulating hormone, which were measured only near the end of the study. Statistical Analysis Sample sizes were estimated for an 0.05 level test to have a chance of at least 80% of detecting a significant difference if the loss for the group treated with vitamin D was 0.005 g/cm2 per year less in the lumbar spine and 0.004 g/cm2 per year less in the whole body than for the controls. Estimates of variability were obtained from a previous calcium supplement study (21). The mean rates of bone, fat, and lean-tissue change for the two periods and for the whole study were standardized to 6-month and 1-year intervals, respectively. Paired /-tests were used to assess differences in rates of change of bone mineral density and soft-tissue composition between periods within each treatment group. Treatment effects were assessed by unpaired t-tests. To determine which treatment groups lost or gained bone-mineral density during the study, we tested for non-zero annualized rates of change within each group. Bone density changes in the two treatment groups of women (those who had menopause within the previous 6 years and those who had menopause more than 6 years previously) were similar in each period and, therefore, all women in each treatment group were analyzed together. All P values are two-tailed. One woman taking vitamin D had changes in spinal bone mineral density that were very different from those of other vitamin D-supplemented women ( - 8.2% compared with a range of - 4.7% to 7.3%). The subject appears to have been compliant. Her data were set aside, however, because the change in spine density was an outlier in a normal probability plot and was rejected by the Grubbs test (P < 0.05) (28). The Mann-Whitney U test (28) for differences in change in spine density (vitamin D compared with placebo) is significant whether the outlier is included (P = 0.04) or not (P = 0.03). Results Of the 276 subjects enrolled, 27 (vitamin D, 15; placebo, 12) did not complete the study or were excluded from analyses. Twelve women (4.7%) dropped out of the study: six (vitamin D, four; placebo, two) for personal reasons, five (vitamin D, one; placebo, four) because of serious illness, and one (vitamin D) to start estrogen therapy. One woman (in the vitamin D group) was asked to withdraw from the study because of a 24-hour urine calcium to creatinine ratio of greater than 1129 mmol/mol. An additional 14 were excluded: three (vitamin D, two; placebo, one) because they were found
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
507
Figure 1. Effect of vitamin D on annualized rates of change in bone mineral density during the year. Period 1 is June-July through December-January, and period 2 is December-January through the following June-July. Within each panel, bars labeled with asterisks (*) differed from zero (P < 0.01); daggers (f) indicate a difference between treatment groups (P < 0.05). The standard error is also shown.
to be taking vitamin D supplements, two (vitamin D, one; placebo, one) for glucocorticoid use during the study, one (placebo) for renal failure, seven (vitamin D, four; placebo, three) because they were found to be taking inappropriate doses of thyroid hormone for hypothyroidism when serum concentrations of thyroid hormone in their initial blood samples were measured with a sensitive assay near the end of the study, and one (vitamin D) whose data were inconsistent (see Statistical Analysis). None had hypercalcemia or kidney stones; however, 11 (vitamin D, six; placebo, five) had 24-hour urine calcium to creatinine ratios (Ca:Cr) of greater than 847 mmol/mol at one point during the study. All medical exclusions were made by the principal investigator without knowledge of treatment assign508
ment. Compliance, measured by the percentage of tablets taken, was 99% during the study. The clinical characteristics of the remaining 249 women are given in Table 1. There were no statistically significant differences at enrollment in these characteristics of the two treatment groups. Because they had radiographic abnormalities in the field of the spine scan (noted on radiographs taken during screening for the earlier calcium trial), 23 women were excluded from analysis of the spinal data. Of the 249 women contributing data in this study, 220 had complete spinal data and 246 had complete whole-body data. Bone mineral density increased from June and July through December and January (period 1) and decreased from December and January through June and July (period 2, Table 2, Figure 1). During period 1, the increases in spinal bone mineral density were similar in the two treatment groups (Table 2). In period 2, however, bone loss was greater in the placebo than in the vitamin D group (P = 0.032). Overall, the placebo group had no change (0.15% [CI, - 0.34% to 0.64%]) whereas the vitamin D-treated women had a modest increase of 0.85% (CI, 0.40% to 1.30%, P < 0.001) in spinal bone mineral density. The net change in spinal bone density was greater in the vitamin D group than in the placebo group (P = 0.04). Increases during period 1 and decreases during period 2 in bone mineral density of the whole body were similar in the two treatment groups, and there was no net change in whole-body bone mineral density during the study in either treatment group. Whole-body lean-tissue mass changed in periods 1 and 2 (Table 3, Figure 2), and these changes were similar in direction and amplitude to those of spinal bone mineral density. In contrast, fat tissue tended to decrease in period 1 and increased in period 2. There was no net change and no effect of vitamin D treatment on either whole-body fat or lean-tissue mass during the study. Fat tissue in the truncal region changed similarly in the treatment groups (mean - 0.65% [CI, - 1.81% to 0.51%] for period 1 and 4.28% [CI, 3.14% to 5.42%] for period 2 for all women). Physical activity levels were similar in the two treatment groups and declined from period 1 to period 2 (Table 4). There was no correlation (or other apparent relationship) between physical activity and change in bone density or in soft-tissue composition within either period, and none between change in activity (from period 1 to period 2) and change in bone density or softtissue composition during period 2. Mean changes in weight during the study were similar in the placebo and vitamin D treatment groups during each period and overall (net change, 0.43 ± 0.29 [SE] kg and 0.58 ± 0.25 kg, respectively). Total daily intake of calcium was approximately 10% higher in the vitamin D group than in the placebo group (Table 4). There was no correlation, however, between calcium intake and change in bone mineral density of the spine or whole body (or in body composition) in either group or period. Intake of vitamin D from diet was also greater in the vitamin D group than in the placebo group (Table 4). The travel index was similar in the placebo and vitamin D groups
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
and tended to be higher in period 2 (33 ± 7 [SE] and 34 ± 7) than in period 1 (12 ± 3 and 6 ± 2), respectively. For both treatment groups, the mean intervals between the first and second dual-energy scans (0.50 ± 0.01 [SE] years) and between the second and third scans (0.50 ± 0.01 years) were the same. There were no differences between the vitamin D and placebo groups in muscle strength, creatinine clearance, alcohol or cigarette use, or distribution of treatment assignment from their previous study (calcium citrate malate, calcium carbonate, or placebo within intake categories of under 400 mg and 400 to 650 mg of calcium daily). The lack of a net change in spinal bone mineral density noted in the placebo group was independent of treatment assignment during the earlier calcium trial. Several biochemical measures including plasma 25(OH)D, serum parathyroid hormone and phosphorus, and 24-hour urine Ca:Cr ratio differed in periods 1 and 2 (Table 5). Supplementation with vitamin D greatly attenuated the change in 25(OH)D, prevented significant variation in serum parathyroid hormone and in 24-hour urine Ca:Cr ratio, and had no effect on changes in serum phosphorus concentration. Discussion In healthy postmenopausal women residing in Massachusetts and taking calcium supplements to approximately the Recommended Dietary Allowance (RDA) (29), increasing vitamin D intake from 100 to 500 IU daily produced an overall benefit in spinal bone mineral density. Spinal bone loss between December and June was less in the vitamin D-treated group than in the placebo group. The finding that treatment affected bone change at the spine (estimated to contain 35% to 65% trabecular bone) more than the whole skeleton (80% cortical bone) suggests that trabecular bone is more sensitive than cortical bone to vitamin D insufficiency. The positive bone effect of the vitamin cannot reasonably be ascribed to classical bone remodeling transients (30) because the observed vitamin D effect of reducing spinal bone loss took place only during the second 6 months of vitamin D supplementation rather than at the outset or throughout. The vitamin D group had somewhat higher calcium and vitamin D intakes. A difference
in calcium intake of 75 mg per day between groups receiving over 700 mg per day, however, is unlikely to have a statistically significant effect on bone change (21), and the modest difference in vitamin D intake of the two treatment groups does not confound the interpretation of the study. This 1-year study does not establish the value of long-term supplementation with vitamin D. A study period of more than 1 year is needed to define the long-term effects of vitamin D supplementation on the spine and another site of interest, the hip. Because wintertime changes in 25(OH)D and parathyroid hormone recurr each year in those with low vitamin D intakes, however, it is very likely that supplemental vitamin D in the winter will provide ongoing benefit. Winter decreases in 25(OH)D (2-5) and increases in parathyroid hormone (4, 5) are well documented. We postulated earlier that these hormonal changes in the winter could have deleterious effects on the skeleton because vitamin D is required to maintain calcium absorption and parathyroid hormone is a potent stimulator of bone resorption (6). This study was not designed to define the effects of individual hormones on the skeleton but it does show that the wintertime increase in parathyroid hormone can be prevented by increasing vitamin D intake. The mean dietary intake of vitamin D of the women in this study, about 100 IU daily, is half of the amount recommended (29) but is in the range commonly consumed both in this country (3, 31, 32) and in Europe (7, 33, 34). In contrast, because of supplementation, the mean calcium intake (800 mg daily) of the women in this study was higher than that of most postmenopausal women in the United States (35). It is notable that with calcium supplementation alone, the placebo group maintained their spinal and whole-body bone mineral density during the study. Vitamin D insufficiency, along with a lower calcium intake, would likely have resulted in even greater wintertime bone loss. Small changes in soft and hard tissue could be detected in this study because of the large sample size and the precision of the measurements. The timing, direction, and magnitude of the spinal changes in periods 1 and 2 are similar to seasonal changes observed during a
Table 3. Mean Changes in Lean- and Fat-Tissue Mass during Treatment with Vitamin D or Placebo* Site and Treatment
Number of Women
Period 1 (June-July through December-January) % Changes/6 Months
Differencet
Overall % Change/Year
Of., {Q^cr/. r^]\ /c \yj /O C/y
(
Lean Placebo Vitamin D Diiference Fat Placebo Vitamin D Difference
Period 2 (December-January through June-July) % Change/6 Months
125 121
1.54(1.03 to 2.05) 1.35 (0.82 to 1.88) - 0.19 ( - 0.92 to 0.54)
125 121
- 1.25 ( - 2.58 to 0.08) - 1.29 ( - 2.66 to 0.08) - 0.04 ( - 1.95 to 1.87)
- 1.83 ( - 2.26 to - 1.40) - 3.37 ( - 4.17 to - 2.55) - 0.35 ( - 0.84 to 0.14) - 1.37 ( - 1.90 to - 0.84) - 2.72 ( - 3.61 to - 1.83) - 0.07 ( - 0.64 to 0.50) - 0.28 ( - 0.47 to 1.03) 0.46 ( - 0.22 to 1.14) 2.64 (1.46 to 3.82) 2.79 (1.65 to 3.93) 0.15 ( - 1.49 to 1.79)
3.89 (1.90 to 5.88) 4.08 (2.14 to 6.02)
1.22 ( - 0.33 to 2.77) 1.43 ( - 0.14 to 3.00) 0.21 ( - 1.99 to 2.41)
* Values are means with 95% confidence intervals (CI). t Rate of change in period 2 differs from rate in period 1 in all instances, P < 0.001.
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
509
Figure 2. Effect of vitamin D on annualized rates of change in lean- and fat-tissue weight during the year. Period 1 is June-July through December-January, and period 2 is December-January through the following June-July. Within each panel, bars labeled with asterisks (*) differed from zero (P < 0.001). The standard error is also shown.
2-year period by Bergstralh and associates (13). They are also consistent with the seasonal findings of most (10-12), but not all, other investigators (14). Soft-tissue changes in this study are in line with well-documented seasonal changes in body composition related to photoperiod in small animals (15, 16). Evidence for seasonal soft-tissue changes in humans is limited and has been noted by some (17), but not by other (18, 19), investigators. In four adult men who underwent serial underwater weighings, body fat was reported to decrease by 16.5% in summer and to increase by 17.8% in winter, whereas lean-body mass did not vary significantly (17). The parallel patterns of change in whole-body lean tissue and bone density observed in this study are consistent with the linkage between lean- and bone-tissue mass that has been described (36). The observed hardand soft-tissue changes are probably related to season. We cannot exclude the possibility, however, that the bone changes in period 1 resulted, at least in part, from a reduction in the rate of bone remodeling associated with the use of supplemental calcium. In addition, increases in fat can cause artifactual decreases in apparent bone mineral density (24). However, the magnitude 510
of this artifact (0.6% decrease in spine phantom bone mineral density for a 30% increase in soft tissue equivalent fat) (24) is small when compared with the observed period 2 mean bone mineral density changes in women who had concomitant truncal and whole-body fat increases of only 5% and 2.5%, respectively. Understanding the scope and basis for hard- and softtissue changes with season is important for several reasons. First, seasonal changes need to be considered in the design and interpretation of clinical studies in which bone density and soft-tissue composition are measured. In addition, considerable research effort is underway to understand the mechanisms of age-related bone loss. An understanding of the causes of the late summertime increase in bone density could lead to more effective intervention. Because of the similarity in patterns of change in bone and lean tissue observed in this study, factors that influence bone are likely to influence soft tissue as well. Physical activity, for example, affects fatand lean-tissue mass, and weight-bearing exercise can have a positive effect on bone (37-39). We were unable to identify any association between change in activity and change in fat or lean tissue or in bone density, possibly because our physical activity assessment was not sufficiently sensitive. On the other hand, variation in activity within the moderate range shown by this study group may not account for the observed tissue changes. Several hormones including gonadal, thyroidal, and adrenal hormones and growth hormone influence the metabolism of soft tissue and bone. These hormones, which are under the regulation of the hypothalamic-pituitary axis, have been reported to have circannual as well as circadian rhythms in young (40-42) and older (43) healthy men and women. Vitamin D insufficiency contributes to spinal bone loss in the winter in healthy postmenopausal women consuming 800 mg of calcium and 100 IU of vitamin D per day. Increasing vitamin D intake to 500 IU daily reduces wintertime bone loss and improves overall den-
Table 4. Characteristics by Season of the 124 Women Treated with Vitamin D and 125 Women Treated with Placebo* Characteristic and Treatment Physical activity, kcalld Placebo Vitamin D Muscle strength, kg Placebo Vitamin D Calcium intake, mgldt Placebo Vitamin D Dietary vitamin D, ILJId Placebo Vitamin D
Within Period 1 (August to November)
Within Period 2 (February to May)
310 ± 31 337 ± 36
200 ± 21t 186 ± 17t
7.9 ± 0.3 8.1 ± 0.2
7.7 ± 0.3 7.9 ± 0.2
721 ± 17§ 810 ± 22
734 ± 17§ 809 ± 20
87 ± 5|| 106 ± 7
90 ± 6|| 110 ± 7
* Values are means ± SE. t Period 2 differs from period 1, P < 0.001. t Includes supplements. § Treatment groups differ at P < 0.005. || Treatment groups differ at P < 0.05.
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
Table 5. Laboratory Values by Season in Women Treated with Vitamin D and Placebo* Index and Treatment Group
Number of Women
Within Period 1 (August to November)
Within Period 2 (February to May) C/„ (Q'sO/r, Cl\ /o \ yj /o \^i)
^ Plasma 25(OH)D, nmollL Placebo Vitamin D Difference Plasma l,25(OH) 2 D, pmollL Placebo Vitamin D Difference Serum alkaline phosphatase, pukatlL Placebo Vitamin D Difference Serum parathyroid hormone, ngIL Placebo Vitamin D Difference Serum ionized calcium, mmollL Placebo Vitamin D Difference Serum phosphorus, mmollL Placebo Vitamin D Difference 24-Hour urinary Ca:Cr ratio, mmollmol Placebo Vitamin D Difference
124 123
Difference
81.3 (76.9 to 85.7) 97.0 (92.8 to 101.2) 15.7 (9.6 to 21.8)§
60.6 (55.6 to 65.6) 92.1 (87.9 to 96.3) 31.5(25.1 to38.0)§
- 20.7 ( - 24.3 to - 17.2)t - 4.9 ( - 7.9 to - 1.8)*
124 123
75.4 (72.0 to 78.9) 73.5 (69.9 to 77.1) - 1.9 ( - 6.9 to 3.1)
77.2 (74.0 to 80.4) 73.3 (69.8 to 76.8) - 3.9 ( - 8.7 to 0.9)
1.8 ( - 1.6 to 5.2) - 0.2 ( - 4.4 to 3.9)
124 123
1.06 (1.00 to 1.12) 1.02 (0.97 to 1.07) - 0.04 ( - 0.11 to 0.03)
1.03 (0.98 to 1.08) 1.00 (0.94 to 1.06) - 0.03 ( - 0 . 1 1 to 0.05)
- 0.03 ( - 0.06 to 0.00) - 0.02 ( - 0.06 to 0.02)
124 123
28.6 (26.8 to 30.4) 28.1 (26.5 to 29.7) - 0.5 ( - 2.9 to 1.9)
32.1 (29.8 to 34.4) 29.2 (27.4 to 31.0) - 2.9 ( - 5.8 to 0.0)||
3.5 (2.0 to 5.0)t 1.1 ( - 0.2 to 2.5)
115 117
1.28 (1.27 to 1.29) 1.28 (1.27 to 1.29) 0.00 ( - 0.01 to 0.01)
1.28 (1.27 to 1.29) 1.28 (1.27 to 1.29) - 0.01 ( - 0.01 to 0.01)
0.00 ( - 0.01 to 0.00) 0.01 (0.00 to 0.01)
123 122
1.15 (1.13 to 1.17) 1.17(1.15 to 1.19) 0.02 ( - 0.01 to 0.05)
1.22 (1.20 to 1.24) 1.20 (1.17 to 1.23) - 0.02 ( - 0.06 to 0.02)
0.07 (0.04 to 0.08)t 0.03 (0.01 to 0.05)11
125 124
529 (486 to 572) 535 (495 to 575) 6 ( - 53 to 65)
486 (450 to 522) 518 (479 to 557) 32 ( - 21 to 85)
- 43 ( - 74 to - 12)** - 17 ( - 50 to 17)
* Values are means with 95% confidence intervals (CI). t Period 2 differs from period 1, P < 0.001. $ Period 2 differs from period 1, P < 0.005. 8 Treatment groups differ, P < 0.001. || Treatment groups differ, P = 0.05. 11 Period 2 differs from period 1, P < 0.05. ** Period 12 differs from period 1, P < 0.01.
sity of the spine. In addition to changes in bone, fatand lean-tissue mass change during the year. These changes are large enough to affect the interpretation of clinical studies in which bone density and body composition are measured. The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. government. Acknowledgments: The authors thank Loran Salamone, Jennifer Johnson, Katherine Deady, Judy King, Sherry Allen, and the staffs of the Metabolic Research Unit, Nutrition Evaluation Laboratory, and Division of Scientific Computing for their contributions. Grant Support: By the U.S.D.A. Human Nutrition Research Center on Aging at Tufts University (contract no. 53-3K06-5-10) and the Procter and Gamble Company. Requests for Reprints: Bess Dawson-Hughes, MD, Calcium and Bone Metabolism Laboratory, U.S.D.A. Human Nutrition Research Center on Aging at Tufts University, 711 Washington Street, Boston, MA 02111. Current Author Addresses: Drs. Dawson-Hughes, Dallal, and Krall and Ms. Harris, Ms. Sokoll, and Ms. Falconer: U.S.D.A. Human Nutrition Research Center on Aging at Tufts University, 711 Washington Street, Boston, MA 02111.
References 1. Smith RW, Rizek J, Frame B. Determinants of serum antirachitic activity. Special reference to involutional osteoporosis. Am J Clin Nutr. 1964;14:98-108.
Stamp TC, Round JM. Seasonal changes in human plasma levels of 25-hydroxyvitamin D. Nature. 1974;247:563-5. Omdahl JL, Garry PJ, Hunsaker LA, Hunt WC, Goodwin JS. Nutritional status in a healthy elderly population: vitamin D. Am J Clin Nutr. 1982;36:1225-33. Lips P, Hackeng WH, Jongen MJ, van Ginkel FC, Netelenbos JC. Seasonal variation in serum concentrations of parathyroid hormone in elderly people. J Clin Endocrinol Metab. 1983;57:204-6. Meller Y, Kestenbaum RS, Galinsky D, Shany S. Seasonal variation in serum levels of vitamin D metabolites and parathormone in geriatric patients with fractures in Southern Israel. Isr J Med Sci. 1986;22:8-11. Krall EA, Sahyoun N, Tannenbaum S, Dallal GE, Dawson-Hughes B. Effect of vitamin D intake on seasonal variations in parathyroid hormone secretion in postmenopausal women. N Engl J Med. 1989; 321:1777-83. Lips P, van Ginkel FC, Jongen MJ, Rubertus F, van der Vijgh WJ, Netelenbos JC. Determinants of vitamin D status in patients with hip fracture and in elderly control subjects. Am J Clin Nutr. 1987;46: 1005-10. Malm OJ. Calcium requirement and adaptation in adult men. Scand J Clin Lab Invest. 1958;10(Supp 36): 1-290. Tothill P, Kennedy NS, Nicoll J, Smith MA, Reid DM, Nuki G. The seasonal variation of total body calcium. Clin Phys Physiol Meas. 1986;7:361-7. Aitken JM, Anderson JB, Horton PW. Seasonal variation in bone mineral content after the menopause. Nature. 1973;241:59-60. Hyldstrup L, McNair P, Jensen GF, Transbol I. Seasonal variations in indices of bone formation precede appropriate bone mineral changes in normal men. Bone. 1986;7:167-70. Krolner B. Seasonal variation of lumbar spine bone mineral content in normal women. Calcif Tissue Int. 1983;35:145-7. Bergstralh EJ, Sinaki M, Offord KP, Wahner HW, Melton LJ 3d. Effect of season on physical activity score, back extensor muscle strength, and lumbar bone mineral density. J Bone Min Res. 1990; 5:371-7.
1 October 1991 • Annals
of Internal
Medicine
• V o l u m e 115 • N u m b e r 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
511
14. Overgaard K, Nilas L, Johansen JS, Christiansen C. Lack of seasonal variation in bone mass and biochemical estimates of bone turnover. Bone. 1988;9:285-8. 15. Dark J, Zucker I, Wade GN. Photoperiodic regulation of body mass, food intake, and reproduction in meadow voles. Am J Physiol. 1983;245:R334-8. 16. Wade GN, Gray JM. Gonadal effects on food intake and adiposity: a metabolic hypothesis. Physiol Behav. 1979;22:583-93. 17. Tashiro Y. Studies on the relation between body composition and function. No. 3: Relation between body composition and basal metabolism with special reference to its seasonal fluctuation. Taishitsu Igaku Kenkyusho Houkoku. 1961;12:589-97. 18. Acheson KJ, Campbell IT, Edholm OG, Miller DS, Stock MJ. A longitudinal study of body weight and body fat changes in Antarctica. Am J Clin Nutr. 1980;33:972-7. 19. Sato H, Watanuki S, Iwanaga K, Baba N. Seasonal variation in body composition in women. Ann Physiol Anthropol. 1984;3:152-6. 20. Heymsfield SB, Wang J, Heshka S, Kehayias JJ, Pierson RN. Dualphoton absorptiometry: comparison of bone mineral and soft tissue mass measurements in vivo with established methods. Am J Clin Nutr. 1989;49:1283-9. 21. Dawson-Hughes B, Dallal GE, Krall EA, Sadowski L, Sahyoun N, Tannenbaum S. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med. 1990;323:878-83. 22. Smith KT, Heaney RP, Flora L, Hinders SM. Calcium absorption from a new calcium delivery system (CCM). Calcif Tissue Int. 1987;41:351-2. 23. Kriska AM, Sandler RB, Cauley JA, LaPorte RE, Horn DL, Pambianco G. The assessment of historical physical activity and its relation to adult bone parameters. Am J Epidemiol. 1988;127:1053-63. 24. Johnson J, Dawson-Hughes B. Precision and stability of dual-energy X-ray absorptiometry measurements. Calcif Tissue Int. 1991 ;49: 174-8. 25. Reinhardt TA, Horst RL, Orf JW, Hollis BW. A microassay for 1,25-dihydroxyvitamin D not requiring high performance liquid chromatography: application to clinical studies. J Clin Endocrinol Metab. 1984;58:91-8. 26. Preece MA, O'Riordan JL, Lawson DE*, Kodicek E. A competitive protein-binding assay for 25-hydroxycholecalciferol and 25-hydroxyergocalciferol in serum. Clin Chem Acta. 1974;54:235-42. 27. Kaplan A, Szabo LL. Clinical Chemistry: Interpretation and Techniques. 2d ed. Philadelphia: Lea & Febiger; 1983. 28. Snedecor GW, Cochran WG. Statistical Methods. 7th ed. Ames, Iowa: The Iowa State University Press; 1980. 29. National Research Council. Subcommittee on the Tenth Edition of the RDAs. Recommended dietary allowances: Subcommittee on the Tenth Edition of the RDAs, Food and Nutrition Board, Commission on Life Sciences, National Research Council. 10th ed. Washington, DC: National Academy Press; 1989.
512
30. Frost HM. The origin and nature of transients in human bone remodeling dynamics. In: Frame B, Parfitt AM, Duncan H; eds. Clinical Aspects of Metabolic Bone Disease. International Conference Series no. 270. Amsterdam: Excerpta Medica; 1973:124-37. 31. Lukert BP, Carey M, McCarty B, Tiemann S, Goodnight L, Helm M, et al. Influence of nutritional factors on calcium-regulating hormones and bone loss. Calcif Tissue Int. 1987;40:119-25. 32. Delvin EE, Imbach A, Copti M. Vitamin D nutritional status and related biochemical indices in an autonomous elderly population. Am J Clin Nutr. 1988;48:373-8. 33. Vir SC, Love AH. Nutritional status of institutionalized and noninstitutionalized aged in Belfast, Northern Ireland. Am J Clin Nutr. 1979;32:1934-47. 34. Lamberg-Allardt C. Vitamin D intake, sunlight exposure and 25hydroxyvitamin D levels in the elderly during one year. Ann Nutr Metab. 1984;28:144-50. 35. Abraham S, Carroll MD, Dresser CM, Johnston CL. Dietary Intake Findings, United States 1976-1980. Hyattsville, Maryland: Department of Health and Human Services; 1983. (DHHS publication no. (PHS) 83-1681.) 36. Cohn SH, Vaswani A, Zanzi I, Aloia JF, Roginsky MS, Ellis KJ. Changes in body chemical composition with age measured by totalbody neutron activation. Metabolism. 1976;25:85-95. 37. Krolner B, Toft B, Pors Nielsen S, Tondevold E. Physical exercise as prophylaxis against involutional vertebral bone loss: a controlled trial. Clin Sci. 1983;64:541-6. 38. Pocock NA, Eisman JA, Yeates MG, Sambrook PN, Eberl S. Physical fitness is a major determinant of femoral neck and lumbar spine bone mineral density. J Clin Invest. 1986;78:618-21. 39. Dalsky GP, Stocke KS, Ehsani AA, Slatopolsky E, Lee WC, Birge SJ Jr. Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann Intern Med. 1988;108:824-8. 40. Touitou Y, Sulon J, Bogdan A, Reinberg A, Sodoyez JC, DemeyPonsart E. Adrenocortical hormones, ageing and mental condition: seasonal and circadian rhythms of plasma 18-hydroxy-l 1-deoxycorticosterone, total and free Cortisol and urinary corticosteroids. J Endocrinol. 1983;96:53-64. 41. Del Ponte A, Guagnano MT, Sensi S. Time-related behavior of endocrine secretion: circannual variations of FT 3 , Cortisol, HGH and serum basal insulin in healthy subjects. Chronobiol Int. 1984; 1:297-300. 42. Kennaway DJ, Royles P. Circadian rhythms of 6-sulphatoxy melatonin, Cortisol and electrolyte excretion of the summer and winter solstices in normal men and women. Acta Endocrinol (Copenh). 1986;13:450-6. 43. Nicolau GY, Lakatua D, Sackett-Lundeen L, Haus E. Circadian and circannual rhythms of hormonal variables in elderly men and women. Chronobiol Int. 1984;1:301-19.
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
1 October 1991
Annals of Internal Medicine
Effect of Vitamin D Supplementation on Wintertime and Overall Bone Loss in Healthy Postmenopausal Women Bess Dawson-Hughes, MD; Gerard E. Dallal, PhD; Elizabeth A. Krall, PhD; Susan Harris, MS; Lori J. Sokoll, MCC; and Gladys Falconer, MS
• Objectives: To determine whether relative vitamin D deficiency during the winter months contributes to age-related bone loss and whether rates of change in hard- and soft-tissue mass vary during the year. • Design: Double-blind, placebo-controlled, 1-year trial in 249 women in which equal numbers of women were randomized to either placebo or 400 IU of vitamin D daily. All women received 377 mg/d of supplemental calcium largely as calcium citrate malate. • Patients: Healthy, ambulatory postmenopausal women with usual intakes of vitamin D of 100 lU/d. • Measurements: Duplicate spine and whole-body scans were done by dual energy x-ray absorptiometry at 6-month intervals that were timed to periods when 25hydroxyvitamin D levels were highest and lowest. Period 1 was June-July to December-January and period 2 was December-January to the next June-July. Serum parathyroid hormone and plasma 25-hydroxyvitamin D levels were measured during periods 1 and 2. • Main Results: In the placebo group, spinal bone mineral density increased in period 1, decreased in period 2, and sustained no net change. Women treated with vitamin D had a similar spinal increase in period 1 (1.46% compared with 1.40% in placebo), less loss in period 2 ( - 0.54% compared with - 1.22%, CI for the difference, 0.05% to 1.31 %, P = 0.032) and a significant overall benefit (0.85% compared with 0.15%, CI for the difference, 0.03% to 1.37%, F = 0.04). In period 2, 25-hydroxyvitamin D levels were lower and parathyroid hormone levels were higher in the placebo than in the vitamin D group. Whole-body lean and fat tissue and bone mineral density varied during the year but did not change overall. • Conclusions: At latitude 42 degrees, healthy postmenopausal women with vitamin D intakes of 100 IU daily can significantly reduce late wintertime bone loss and improve net bone density of the spine over one year by increasing their intake of vitamin D to 500 IU daily. A long-term benefit of preventing vitamin D insufficiency in the winter seems likely although it remains to be shown. Observed changes in bone as well as in fat and lean tissue appear to be related to season. Annuls of Internal Medicine . 1991;115:505-512. 1 October 1991 •
From the U.S. Department of Agriculture Human Nutrition Research Center on Aging at Tufts University, Boston, Massachusetts. For current author Addresses, see end of text.
In 1964, Smith and colleagues (1) reported that serum antirachitic activity varied with season in Michigan but not in Puerto Rico. More recently, it has been recognized that serum parathyroid hormone and plasma 25hydroxyvitamin D (25[OH]D) levels vary inversely with one another and that the concentration of each changes with season in the temperate zone (2-5). In a crosssectional study of 333 healthy ambulatory postmenopausal women in Massachusetts (6), serum parathyroid hormone was lowest in those women tested between August and October and highest in those tested between March and May. When examined in relation to vitamin D intake, however, a higher parathyroid hormone level between March and May was present only in those women who had vitamin D intakes of less than 220 IU daily. Because of the relatively small number of women with intakes over this level, we were unable to be certain that the vitamin D intake associated with no wintertime increase in parathyroid hormone was not higher than 220 IU (6). There is indirect evidence that seasonal increases in parathyroid hormone may affect bone mass adversely. Patients with hip fractures have been shown to have wintertime increases in parathyroid hormone and lower levels of 25(OH)D than controls (7). Calcium balance in Scandinavian prisoners was more negative in winter than in summer (8, 9). Hard- and soft-tissue composition changes with season. Seasonal variation in bone mineral content of the metacarpals (10) and forearm (11) and in density of the spine (12, 13) has been reported in women residing in northern parts of Europe and the United States. Vitamin D deficiency and decreased levels of physical activity were cited as possible explanations for decreases in metacarpal bone mineral content from winter to summer (10). However, others (14) have found no influence of season on bone mass. Seasonal variation in softtissue composition has been documented in small ani• Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
505
mals (15, 16) and also in some (17), but not in other (18, 19), studies in humans. In the past, the methods used to measure body fat and lean tissue have been unsuitable for detecting small changes and for studying large populations because they were labor intensive, imprecise, and sometimes associated with significant radiation exposure. The newly developed dual-energy x-ray absorptiometry method overcomes these obstacles and provides an accurate and reproducible measurement of soft-tissue composition when compared with established methods (20). Our randomized, placebo-controlled, double-blind 1-year study was done to determine the effect of increased vitamin D intake on rates of bone loss during the 6-month periods when parathyroid hormone levels are highest and lowest (6). A second objective was to compare rates of change in fat and lean tissue during these two intervals. The first interval was between June or July and December or January (parathyroid hormone at its peak) and the second was between December or January and the following June or July (parathyroid hormone at its lowest). Because of the documented benefit of increasing calcium intake to 800 mg daily (21), all women received calcium supplements. Calcium citrate malate was used because it is well absorbed (22) and is effective in reducing bone loss in this population (21). Methods The 276 postmenopausal women enrolled in this study met the following entry criteria: white race, good general health, normal ambulation, and normal results on screening laboratory tests. Twenty-four women had menopause within the previous 5 years (mean, 3.9 ± 0.13 [SE] years), and 252 women had menopause 6 or more years previously. All had completed a 2-year calcium supplement trial at the center (21). For the previous trial, selection criteria included calcium intake under 650 mg and absence of compression fracture on thoracic and lumbar-spine radiography. Exclusion criteria for this study included a history of nontraumatic fracture of any bone; renal, hepatic, or gastrointestinal disorders associated with abnormal calcium or bone metabolism; use of estrogen, glucocorticoids, anticonvulsants, or other medications known to affect calcium or bone metabolism; or spine bone-mineral-density values of 2 SD or more below the age-matched reference mean. The pro-
tocol was approved by the Human Investigation Review Committee of Tufts University, and written informed consent was obtained from each woman. Study Design and Supplements In this 12-month, double-blind, placebo-controlled trial, women were randomly assigned to treatment with 400 IU of vitamin D or placebo. Both the vitamin D and placebo tablets contained 127 mg of elemental calcium as calcium phosphate. In addition, all subjects received 250 mg of elemental calcium as calcium citrate malate daily so that each had a total of 377 mg of supplemental calcium per day. Women were randomized to treatment after being stratified by dietary calcium intake (less than 400 mg/d or 400 to 650 mg/d), treatment group in the previous trial (placebo or 500 mg of elemental calcium as either calcium citrate malate or calcium carbonate), and previous category of years since menopause (greater or fewer than 6 years). The mean interval between the previous and current trials was 0.42 ± 0.31 years, and it did not vary by treatment group. In this interval, women were counseled to ingest 800 mg of calcium daily. During the study, the women were instructed to take the tablets at bedtime, to maintain their usual diets, and to avoid taking calcium and vitamin D supplements other than those provided. Each woman came to the center three times at 6-month intervals (first in June or July, next in December or January, and finally in June or July) for measurements of spine and whole-body bone-mineral density, whole-body soft-tissue composition, height, and weight. The medical history was assessed on the first and last visits. In the periods of August through November and February through May, each woman was assessed for calcium and vitamin D intake, physical activity, and quadriceps strength; and had blood and urine measurements. During the last four visits, pill compliance was recorded. Calcium citrate malate was prepared as previously described (22) with a calcium:citrate:malate molar ratio of 6:2:3. This calcium source is commercially available only in fortified citrus juices (Citrus Hill Plus Calcium, The Procter and Gamble Company, Cincinnati, Ohio). Measurements Dietary calcium and vitamin D intakes were estimated by means of a food-frequency questionnaire that was administered by a trained technician (6). Food models and household measuring tools were used to help subjects estimate portion sizes. Alcohol use and cigarette smoking were also assessed by questionnaire. Travel to lower latitudes was documented on a questionnaire and a travel index calculated as days traveled divided by latitude x 100. Current physical activity was assessed by use of a questionnaire adapted from Kriska and colleagues (23) and
Table 1. Clinical Characteristics at Enrollment of Women Who Completed the Trial* Characteristic
Treatment Group> Placebo
Number Age, y Time since menopause^, y < 6 y since menopause, % Weight,^ Smokers, % Alcohol consumption, glwk Bone mineral density, g/ctn (A?)§ Spine Whole body
125 61.9 13.4 18.4 67.4 7.2 44.8
±0.5 ± 0.6 ± 1.0 ± 6.6
1.03 ± 0.01 (110) 1.05 ± 0.01 (125)
P Valuet Vitamin D
124 61.4 13.5 9.7 68.5 8.1 32.1
± 0.5 ± 0.6 ± 1.1 ± 4.8
1.05 ± 0.02(110) 1.07 ± 0 . 0 1 (121)
> 0.2 >0.2 0.07 > 0.2 >0.2 0.12 > 0.2 0.08
* Plus-minus values are means ± SE. t For two-sample Mest (or Pearson chi-square test in the case of percentage of women less than 6 years since menopause and of percentage of smokers). X Year of menopause was unknown for 11 women (vitamin D supplement group, 5; and placebo group, 6) and was assumed to have occurred at age 50. § Number of women with complete data is given in parentheses. 506
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
Table 2. Mean Change in Bone Mineral Density during Treatment with Vitamin D or Placebo* Site and Treatment
Idumber of Women
Period 1 (June-July through December-January) % Change/6 Months
Period 2 (DecemberJanuary through June-July) % Change/6 Months
< Spine Placebo Vitamin D Difference Whole body Placebo Vitamin D Difference
Differencet
Overall % Change/Year
%(95%CI)
>
110 110
1.40 (0.97 to 1.83) 1.46 (0.99 to 1.93) 0.06 (- 0.58 to 0.70)
- 1.22 (- 1.69 to - 0.75) - 0.54 (- 0.95 to - 0.13) 0.68 (0.05 to 1.31)*
- 2.63 (- 3.38 to - 1.88) - 2.00 (- 2.75 to - 1.25)
0.15 (- 0.34 to 0.64) 0.85 (0.40 to 1.30) 0.70 (0.03 to 1.37)§
125 121
0.61 (0.37 to 0.85) 0.62 (0.40 to 0.84) 0.01 (- 0.31 to 0.33)
- 0.66 (- 0.86 to - 0.46) - 0.56 (- 0.78 to - 0.34) 0.10 (- 0.19 to 0.39)
- 1.27 (- 1.64 to - 0.90) - 1.18 (- 1.54 to - 0.82)
- 0.08 (- 0.30 to 0.14) 0.03 (- 0.21 to 0.27) 0.11 (- 0.21 to 0.43)
* Values are means with 95% confidence intervals (CI). t Rate of change in period 2 differs from rate in period 1 in all instances, P < 0.001. X Treatment groups differ at P = 0.03. § Treatment groups differ at P = 0.04.
reported as kilocalories per day. Quadriceps muscle strength of the dominant leg was assessed with use of a Thigh-Knee Dynamic (Universal, Cedar Rapids, Iowa). Each woman, in the sitting position, pushed a padded bar forward with the lower leg (at the level of the distal tibia). The maximum weight pushed by each leg was recorded. Bone-mineral density of the spine (L-2 to L-4) and whole skeleton and whole-body soft-tissue composition were measured with a model DPX dual-energy x-ray absorptiometer (Lunar Radiation, Madison, Wisconsin). Coefficients of variation were 1% for the spine and 0.6% for the whole-skeleton bone density measurements, 2% for fat, and 1% for lean body mass, as described previously (24). The scans were analyzed with Lunar Radiation Corporation software version 3.10. On each of three visits, women had two spine and whole-body scans, with repositioning between the two scans. The mean of the two bone density and body composition values at each time point was used. An aluminum phantom was scanned biweekly in a mixture of oil:water (30:70) at total thicknesses ranging from 15.2 cm to 27.9 cm (24) over the last 9 months of the study. During this time, the bone-mineral density of the phantom was stable. In addition, bone mineral density of the phantom scanned in 15.2 to 25.4 cm of water was stable throughout the study (24). Throughout the physiologic thickness range of 15.2 to 27.9 cm, phantom bone mineral density was a mean of 0.6% lower when the phantom was scanned in oil/water (30:70) than in water alone (24). Intact parathyroid hormone in serum was measured with Allegro Intact PTH kits obtained from Nichols Institute (San Juan Capistrano, California) with intra-assay and interassay coefficients of variation of 5.6% and 6.6%, respectively. Plasma levels of 1,25-dihydroxyvitamin D (l,25(OH)2D) were measured by the competitive protein-binding method of Reinhardt and colleagues (25) with intra-assay and interassay coefficients of variation of 4.9% and 7.7%, respectively. Plasma levels of 25(OH)D were measured by the method of Preece and colleagues (26) with intra-assay and interassay coefficients of variation of 5.0% and 7.3%, respectively. The normal range for adults, as determined in 75 healthy persons from 18 to 65 years of age, was 20 to 138 nmol/L. Serum ionized calcium levels were measured with a Nova 7 analyzer (Nova Biomedical, Waltham, Massachusetts); serum alkaline phosphatase, phosphorus, and creatinine levels were measured by colorimetry with a Cobas Mira; and urine creatinine levels were with a Cobas Fara centrifugal analyzer (Roche Instruments, Belleville, New Jersey). Creatinine clearance was computed and adjusted for body-surface area (27). Urinary calcium was measured by direct-current plasma emission spectroscopy with a Spectrascan 6 (Beckman Instruments, Palo Alto, California). In women receiving thyroid hormone therapy, levels of thyroid-stimulating hormone were measured with Allegro HS-TSH kits obtained from Nichols Institute. Analyses were done in batches as the samples were col-
lected, except for serum levels of thyroid-stimulating hormone, which were measured only near the end of the study. Statistical Analysis Sample sizes were estimated for an 0.05 level test to have a chance of at least 80% of detecting a significant difference if the loss for the group treated with vitamin D was 0.005 g/cm2 per year less in the lumbar spine and 0.004 g/cm2 per year less in the whole body than for the controls. Estimates of variability were obtained from a previous calcium supplement study (21). The mean rates of bone, fat, and lean-tissue change for the two periods and for the whole study were standardized to 6-month and 1-year intervals, respectively. Paired /-tests were used to assess differences in rates of change of bone mineral density and soft-tissue composition between periods within each treatment group. Treatment effects were assessed by unpaired t-tests. To determine which treatment groups lost or gained bone-mineral density during the study, we tested for non-zero annualized rates of change within each group. Bone density changes in the two treatment groups of women (those who had menopause within the previous 6 years and those who had menopause more than 6 years previously) were similar in each period and, therefore, all women in each treatment group were analyzed together. All P values are two-tailed. One woman taking vitamin D had changes in spinal bone mineral density that were very different from those of other vitamin D-supplemented women ( - 8.2% compared with a range of - 4.7% to 7.3%). The subject appears to have been compliant. Her data were set aside, however, because the change in spine density was an outlier in a normal probability plot and was rejected by the Grubbs test (P < 0.05) (28). The Mann-Whitney U test (28) for differences in change in spine density (vitamin D compared with placebo) is significant whether the outlier is included (P = 0.04) or not (P = 0.03). Results Of the 276 subjects enrolled, 27 (vitamin D, 15; placebo, 12) did not complete the study or were excluded from analyses. Twelve women (4.7%) dropped out of the study: six (vitamin D, four; placebo, two) for personal reasons, five (vitamin D, one; placebo, four) because of serious illness, and one (vitamin D) to start estrogen therapy. One woman (in the vitamin D group) was asked to withdraw from the study because of a 24-hour urine calcium to creatinine ratio of greater than 1129 mmol/mol. An additional 14 were excluded: three (vitamin D, two; placebo, one) because they were found
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
507
Figure 1. Effect of vitamin D on annualized rates of change in bone mineral density during the year. Period 1 is June-July through December-January, and period 2 is December-January through the following June-July. Within each panel, bars labeled with asterisks (*) differed from zero (P < 0.01); daggers (f) indicate a difference between treatment groups (P < 0.05). The standard error is also shown.
to be taking vitamin D supplements, two (vitamin D, one; placebo, one) for glucocorticoid use during the study, one (placebo) for renal failure, seven (vitamin D, four; placebo, three) because they were found to be taking inappropriate doses of thyroid hormone for hypothyroidism when serum concentrations of thyroid hormone in their initial blood samples were measured with a sensitive assay near the end of the study, and one (vitamin D) whose data were inconsistent (see Statistical Analysis). None had hypercalcemia or kidney stones; however, 11 (vitamin D, six; placebo, five) had 24-hour urine calcium to creatinine ratios (Ca:Cr) of greater than 847 mmol/mol at one point during the study. All medical exclusions were made by the principal investigator without knowledge of treatment assign508
ment. Compliance, measured by the percentage of tablets taken, was 99% during the study. The clinical characteristics of the remaining 249 women are given in Table 1. There were no statistically significant differences at enrollment in these characteristics of the two treatment groups. Because they had radiographic abnormalities in the field of the spine scan (noted on radiographs taken during screening for the earlier calcium trial), 23 women were excluded from analysis of the spinal data. Of the 249 women contributing data in this study, 220 had complete spinal data and 246 had complete whole-body data. Bone mineral density increased from June and July through December and January (period 1) and decreased from December and January through June and July (period 2, Table 2, Figure 1). During period 1, the increases in spinal bone mineral density were similar in the two treatment groups (Table 2). In period 2, however, bone loss was greater in the placebo than in the vitamin D group (P = 0.032). Overall, the placebo group had no change (0.15% [CI, - 0.34% to 0.64%]) whereas the vitamin D-treated women had a modest increase of 0.85% (CI, 0.40% to 1.30%, P < 0.001) in spinal bone mineral density. The net change in spinal bone density was greater in the vitamin D group than in the placebo group (P = 0.04). Increases during period 1 and decreases during period 2 in bone mineral density of the whole body were similar in the two treatment groups, and there was no net change in whole-body bone mineral density during the study in either treatment group. Whole-body lean-tissue mass changed in periods 1 and 2 (Table 3, Figure 2), and these changes were similar in direction and amplitude to those of spinal bone mineral density. In contrast, fat tissue tended to decrease in period 1 and increased in period 2. There was no net change and no effect of vitamin D treatment on either whole-body fat or lean-tissue mass during the study. Fat tissue in the truncal region changed similarly in the treatment groups (mean - 0.65% [CI, - 1.81% to 0.51%] for period 1 and 4.28% [CI, 3.14% to 5.42%] for period 2 for all women). Physical activity levels were similar in the two treatment groups and declined from period 1 to period 2 (Table 4). There was no correlation (or other apparent relationship) between physical activity and change in bone density or in soft-tissue composition within either period, and none between change in activity (from period 1 to period 2) and change in bone density or softtissue composition during period 2. Mean changes in weight during the study were similar in the placebo and vitamin D treatment groups during each period and overall (net change, 0.43 ± 0.29 [SE] kg and 0.58 ± 0.25 kg, respectively). Total daily intake of calcium was approximately 10% higher in the vitamin D group than in the placebo group (Table 4). There was no correlation, however, between calcium intake and change in bone mineral density of the spine or whole body (or in body composition) in either group or period. Intake of vitamin D from diet was also greater in the vitamin D group than in the placebo group (Table 4). The travel index was similar in the placebo and vitamin D groups
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
and tended to be higher in period 2 (33 ± 7 [SE] and 34 ± 7) than in period 1 (12 ± 3 and 6 ± 2), respectively. For both treatment groups, the mean intervals between the first and second dual-energy scans (0.50 ± 0.01 [SE] years) and between the second and third scans (0.50 ± 0.01 years) were the same. There were no differences between the vitamin D and placebo groups in muscle strength, creatinine clearance, alcohol or cigarette use, or distribution of treatment assignment from their previous study (calcium citrate malate, calcium carbonate, or placebo within intake categories of under 400 mg and 400 to 650 mg of calcium daily). The lack of a net change in spinal bone mineral density noted in the placebo group was independent of treatment assignment during the earlier calcium trial. Several biochemical measures including plasma 25(OH)D, serum parathyroid hormone and phosphorus, and 24-hour urine Ca:Cr ratio differed in periods 1 and 2 (Table 5). Supplementation with vitamin D greatly attenuated the change in 25(OH)D, prevented significant variation in serum parathyroid hormone and in 24-hour urine Ca:Cr ratio, and had no effect on changes in serum phosphorus concentration. Discussion In healthy postmenopausal women residing in Massachusetts and taking calcium supplements to approximately the Recommended Dietary Allowance (RDA) (29), increasing vitamin D intake from 100 to 500 IU daily produced an overall benefit in spinal bone mineral density. Spinal bone loss between December and June was less in the vitamin D-treated group than in the placebo group. The finding that treatment affected bone change at the spine (estimated to contain 35% to 65% trabecular bone) more than the whole skeleton (80% cortical bone) suggests that trabecular bone is more sensitive than cortical bone to vitamin D insufficiency. The positive bone effect of the vitamin cannot reasonably be ascribed to classical bone remodeling transients (30) because the observed vitamin D effect of reducing spinal bone loss took place only during the second 6 months of vitamin D supplementation rather than at the outset or throughout. The vitamin D group had somewhat higher calcium and vitamin D intakes. A difference
in calcium intake of 75 mg per day between groups receiving over 700 mg per day, however, is unlikely to have a statistically significant effect on bone change (21), and the modest difference in vitamin D intake of the two treatment groups does not confound the interpretation of the study. This 1-year study does not establish the value of long-term supplementation with vitamin D. A study period of more than 1 year is needed to define the long-term effects of vitamin D supplementation on the spine and another site of interest, the hip. Because wintertime changes in 25(OH)D and parathyroid hormone recurr each year in those with low vitamin D intakes, however, it is very likely that supplemental vitamin D in the winter will provide ongoing benefit. Winter decreases in 25(OH)D (2-5) and increases in parathyroid hormone (4, 5) are well documented. We postulated earlier that these hormonal changes in the winter could have deleterious effects on the skeleton because vitamin D is required to maintain calcium absorption and parathyroid hormone is a potent stimulator of bone resorption (6). This study was not designed to define the effects of individual hormones on the skeleton but it does show that the wintertime increase in parathyroid hormone can be prevented by increasing vitamin D intake. The mean dietary intake of vitamin D of the women in this study, about 100 IU daily, is half of the amount recommended (29) but is in the range commonly consumed both in this country (3, 31, 32) and in Europe (7, 33, 34). In contrast, because of supplementation, the mean calcium intake (800 mg daily) of the women in this study was higher than that of most postmenopausal women in the United States (35). It is notable that with calcium supplementation alone, the placebo group maintained their spinal and whole-body bone mineral density during the study. Vitamin D insufficiency, along with a lower calcium intake, would likely have resulted in even greater wintertime bone loss. Small changes in soft and hard tissue could be detected in this study because of the large sample size and the precision of the measurements. The timing, direction, and magnitude of the spinal changes in periods 1 and 2 are similar to seasonal changes observed during a
Table 3. Mean Changes in Lean- and Fat-Tissue Mass during Treatment with Vitamin D or Placebo* Site and Treatment
Number of Women
Period 1 (June-July through December-January) % Changes/6 Months
Differencet
Overall % Change/Year
Of., {Q^cr/. r^]\ /c \yj /O C/y
(
Lean Placebo Vitamin D Diiference Fat Placebo Vitamin D Difference
Period 2 (December-January through June-July) % Change/6 Months
125 121
1.54(1.03 to 2.05) 1.35 (0.82 to 1.88) - 0.19 ( - 0.92 to 0.54)
125 121
- 1.25 ( - 2.58 to 0.08) - 1.29 ( - 2.66 to 0.08) - 0.04 ( - 1.95 to 1.87)
- 1.83 ( - 2.26 to - 1.40) - 3.37 ( - 4.17 to - 2.55) - 0.35 ( - 0.84 to 0.14) - 1.37 ( - 1.90 to - 0.84) - 2.72 ( - 3.61 to - 1.83) - 0.07 ( - 0.64 to 0.50) - 0.28 ( - 0.47 to 1.03) 0.46 ( - 0.22 to 1.14) 2.64 (1.46 to 3.82) 2.79 (1.65 to 3.93) 0.15 ( - 1.49 to 1.79)
3.89 (1.90 to 5.88) 4.08 (2.14 to 6.02)
1.22 ( - 0.33 to 2.77) 1.43 ( - 0.14 to 3.00) 0.21 ( - 1.99 to 2.41)
* Values are means with 95% confidence intervals (CI). t Rate of change in period 2 differs from rate in period 1 in all instances, P < 0.001.
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
509
Figure 2. Effect of vitamin D on annualized rates of change in lean- and fat-tissue weight during the year. Period 1 is June-July through December-January, and period 2 is December-January through the following June-July. Within each panel, bars labeled with asterisks (*) differed from zero (P < 0.001). The standard error is also shown.
2-year period by Bergstralh and associates (13). They are also consistent with the seasonal findings of most (10-12), but not all, other investigators (14). Soft-tissue changes in this study are in line with well-documented seasonal changes in body composition related to photoperiod in small animals (15, 16). Evidence for seasonal soft-tissue changes in humans is limited and has been noted by some (17), but not by other (18, 19), investigators. In four adult men who underwent serial underwater weighings, body fat was reported to decrease by 16.5% in summer and to increase by 17.8% in winter, whereas lean-body mass did not vary significantly (17). The parallel patterns of change in whole-body lean tissue and bone density observed in this study are consistent with the linkage between lean- and bone-tissue mass that has been described (36). The observed hardand soft-tissue changes are probably related to season. We cannot exclude the possibility, however, that the bone changes in period 1 resulted, at least in part, from a reduction in the rate of bone remodeling associated with the use of supplemental calcium. In addition, increases in fat can cause artifactual decreases in apparent bone mineral density (24). However, the magnitude 510
of this artifact (0.6% decrease in spine phantom bone mineral density for a 30% increase in soft tissue equivalent fat) (24) is small when compared with the observed period 2 mean bone mineral density changes in women who had concomitant truncal and whole-body fat increases of only 5% and 2.5%, respectively. Understanding the scope and basis for hard- and softtissue changes with season is important for several reasons. First, seasonal changes need to be considered in the design and interpretation of clinical studies in which bone density and soft-tissue composition are measured. In addition, considerable research effort is underway to understand the mechanisms of age-related bone loss. An understanding of the causes of the late summertime increase in bone density could lead to more effective intervention. Because of the similarity in patterns of change in bone and lean tissue observed in this study, factors that influence bone are likely to influence soft tissue as well. Physical activity, for example, affects fatand lean-tissue mass, and weight-bearing exercise can have a positive effect on bone (37-39). We were unable to identify any association between change in activity and change in fat or lean tissue or in bone density, possibly because our physical activity assessment was not sufficiently sensitive. On the other hand, variation in activity within the moderate range shown by this study group may not account for the observed tissue changes. Several hormones including gonadal, thyroidal, and adrenal hormones and growth hormone influence the metabolism of soft tissue and bone. These hormones, which are under the regulation of the hypothalamic-pituitary axis, have been reported to have circannual as well as circadian rhythms in young (40-42) and older (43) healthy men and women. Vitamin D insufficiency contributes to spinal bone loss in the winter in healthy postmenopausal women consuming 800 mg of calcium and 100 IU of vitamin D per day. Increasing vitamin D intake to 500 IU daily reduces wintertime bone loss and improves overall den-
Table 4. Characteristics by Season of the 124 Women Treated with Vitamin D and 125 Women Treated with Placebo* Characteristic and Treatment Physical activity, kcalld Placebo Vitamin D Muscle strength, kg Placebo Vitamin D Calcium intake, mgldt Placebo Vitamin D Dietary vitamin D, ILJId Placebo Vitamin D
Within Period 1 (August to November)
Within Period 2 (February to May)
310 ± 31 337 ± 36
200 ± 21t 186 ± 17t
7.9 ± 0.3 8.1 ± 0.2
7.7 ± 0.3 7.9 ± 0.2
721 ± 17§ 810 ± 22
734 ± 17§ 809 ± 20
87 ± 5|| 106 ± 7
90 ± 6|| 110 ± 7
* Values are means ± SE. t Period 2 differs from period 1, P < 0.001. t Includes supplements. § Treatment groups differ at P < 0.005. || Treatment groups differ at P < 0.05.
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
Table 5. Laboratory Values by Season in Women Treated with Vitamin D and Placebo* Index and Treatment Group
Number of Women
Within Period 1 (August to November)
Within Period 2 (February to May) C/„ (Q'sO/r, Cl\ /o \ yj /o \^i)
^ Plasma 25(OH)D, nmollL Placebo Vitamin D Difference Plasma l,25(OH) 2 D, pmollL Placebo Vitamin D Difference Serum alkaline phosphatase, pukatlL Placebo Vitamin D Difference Serum parathyroid hormone, ngIL Placebo Vitamin D Difference Serum ionized calcium, mmollL Placebo Vitamin D Difference Serum phosphorus, mmollL Placebo Vitamin D Difference 24-Hour urinary Ca:Cr ratio, mmollmol Placebo Vitamin D Difference
124 123
Difference
81.3 (76.9 to 85.7) 97.0 (92.8 to 101.2) 15.7 (9.6 to 21.8)§
60.6 (55.6 to 65.6) 92.1 (87.9 to 96.3) 31.5(25.1 to38.0)§
- 20.7 ( - 24.3 to - 17.2)t - 4.9 ( - 7.9 to - 1.8)*
124 123
75.4 (72.0 to 78.9) 73.5 (69.9 to 77.1) - 1.9 ( - 6.9 to 3.1)
77.2 (74.0 to 80.4) 73.3 (69.8 to 76.8) - 3.9 ( - 8.7 to 0.9)
1.8 ( - 1.6 to 5.2) - 0.2 ( - 4.4 to 3.9)
124 123
1.06 (1.00 to 1.12) 1.02 (0.97 to 1.07) - 0.04 ( - 0.11 to 0.03)
1.03 (0.98 to 1.08) 1.00 (0.94 to 1.06) - 0.03 ( - 0 . 1 1 to 0.05)
- 0.03 ( - 0.06 to 0.00) - 0.02 ( - 0.06 to 0.02)
124 123
28.6 (26.8 to 30.4) 28.1 (26.5 to 29.7) - 0.5 ( - 2.9 to 1.9)
32.1 (29.8 to 34.4) 29.2 (27.4 to 31.0) - 2.9 ( - 5.8 to 0.0)||
3.5 (2.0 to 5.0)t 1.1 ( - 0.2 to 2.5)
115 117
1.28 (1.27 to 1.29) 1.28 (1.27 to 1.29) 0.00 ( - 0.01 to 0.01)
1.28 (1.27 to 1.29) 1.28 (1.27 to 1.29) - 0.01 ( - 0.01 to 0.01)
0.00 ( - 0.01 to 0.00) 0.01 (0.00 to 0.01)
123 122
1.15 (1.13 to 1.17) 1.17(1.15 to 1.19) 0.02 ( - 0.01 to 0.05)
1.22 (1.20 to 1.24) 1.20 (1.17 to 1.23) - 0.02 ( - 0.06 to 0.02)
0.07 (0.04 to 0.08)t 0.03 (0.01 to 0.05)11
125 124
529 (486 to 572) 535 (495 to 575) 6 ( - 53 to 65)
486 (450 to 522) 518 (479 to 557) 32 ( - 21 to 85)
- 43 ( - 74 to - 12)** - 17 ( - 50 to 17)
* Values are means with 95% confidence intervals (CI). t Period 2 differs from period 1, P < 0.001. $ Period 2 differs from period 1, P < 0.005. 8 Treatment groups differ, P < 0.001. || Treatment groups differ, P = 0.05. 11 Period 2 differs from period 1, P < 0.05. ** Period 12 differs from period 1, P < 0.01.
sity of the spine. In addition to changes in bone, fatand lean-tissue mass change during the year. These changes are large enough to affect the interpretation of clinical studies in which bone density and body composition are measured. The contents of this publication do not necessarily reflect the views or policies of the U.S. Department of Agriculture nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. government. Acknowledgments: The authors thank Loran Salamone, Jennifer Johnson, Katherine Deady, Judy King, Sherry Allen, and the staffs of the Metabolic Research Unit, Nutrition Evaluation Laboratory, and Division of Scientific Computing for their contributions. Grant Support: By the U.S.D.A. Human Nutrition Research Center on Aging at Tufts University (contract no. 53-3K06-5-10) and the Procter and Gamble Company. Requests for Reprints: Bess Dawson-Hughes, MD, Calcium and Bone Metabolism Laboratory, U.S.D.A. Human Nutrition Research Center on Aging at Tufts University, 711 Washington Street, Boston, MA 02111. Current Author Addresses: Drs. Dawson-Hughes, Dallal, and Krall and Ms. Harris, Ms. Sokoll, and Ms. Falconer: U.S.D.A. Human Nutrition Research Center on Aging at Tufts University, 711 Washington Street, Boston, MA 02111.
References 1. Smith RW, Rizek J, Frame B. Determinants of serum antirachitic activity. Special reference to involutional osteoporosis. Am J Clin Nutr. 1964;14:98-108.
Stamp TC, Round JM. Seasonal changes in human plasma levels of 25-hydroxyvitamin D. Nature. 1974;247:563-5. Omdahl JL, Garry PJ, Hunsaker LA, Hunt WC, Goodwin JS. Nutritional status in a healthy elderly population: vitamin D. Am J Clin Nutr. 1982;36:1225-33. Lips P, Hackeng WH, Jongen MJ, van Ginkel FC, Netelenbos JC. Seasonal variation in serum concentrations of parathyroid hormone in elderly people. J Clin Endocrinol Metab. 1983;57:204-6. Meller Y, Kestenbaum RS, Galinsky D, Shany S. Seasonal variation in serum levels of vitamin D metabolites and parathormone in geriatric patients with fractures in Southern Israel. Isr J Med Sci. 1986;22:8-11. Krall EA, Sahyoun N, Tannenbaum S, Dallal GE, Dawson-Hughes B. Effect of vitamin D intake on seasonal variations in parathyroid hormone secretion in postmenopausal women. N Engl J Med. 1989; 321:1777-83. Lips P, van Ginkel FC, Jongen MJ, Rubertus F, van der Vijgh WJ, Netelenbos JC. Determinants of vitamin D status in patients with hip fracture and in elderly control subjects. Am J Clin Nutr. 1987;46: 1005-10. Malm OJ. Calcium requirement and adaptation in adult men. Scand J Clin Lab Invest. 1958;10(Supp 36): 1-290. Tothill P, Kennedy NS, Nicoll J, Smith MA, Reid DM, Nuki G. The seasonal variation of total body calcium. Clin Phys Physiol Meas. 1986;7:361-7. Aitken JM, Anderson JB, Horton PW. Seasonal variation in bone mineral content after the menopause. Nature. 1973;241:59-60. Hyldstrup L, McNair P, Jensen GF, Transbol I. Seasonal variations in indices of bone formation precede appropriate bone mineral changes in normal men. Bone. 1986;7:167-70. Krolner B. Seasonal variation of lumbar spine bone mineral content in normal women. Calcif Tissue Int. 1983;35:145-7. Bergstralh EJ, Sinaki M, Offord KP, Wahner HW, Melton LJ 3d. Effect of season on physical activity score, back extensor muscle strength, and lumbar bone mineral density. J Bone Min Res. 1990; 5:371-7.
1 October 1991 • Annals
of Internal
Medicine
• V o l u m e 115 • N u m b e r 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
511
14. Overgaard K, Nilas L, Johansen JS, Christiansen C. Lack of seasonal variation in bone mass and biochemical estimates of bone turnover. Bone. 1988;9:285-8. 15. Dark J, Zucker I, Wade GN. Photoperiodic regulation of body mass, food intake, and reproduction in meadow voles. Am J Physiol. 1983;245:R334-8. 16. Wade GN, Gray JM. Gonadal effects on food intake and adiposity: a metabolic hypothesis. Physiol Behav. 1979;22:583-93. 17. Tashiro Y. Studies on the relation between body composition and function. No. 3: Relation between body composition and basal metabolism with special reference to its seasonal fluctuation. Taishitsu Igaku Kenkyusho Houkoku. 1961;12:589-97. 18. Acheson KJ, Campbell IT, Edholm OG, Miller DS, Stock MJ. A longitudinal study of body weight and body fat changes in Antarctica. Am J Clin Nutr. 1980;33:972-7. 19. Sato H, Watanuki S, Iwanaga K, Baba N. Seasonal variation in body composition in women. Ann Physiol Anthropol. 1984;3:152-6. 20. Heymsfield SB, Wang J, Heshka S, Kehayias JJ, Pierson RN. Dualphoton absorptiometry: comparison of bone mineral and soft tissue mass measurements in vivo with established methods. Am J Clin Nutr. 1989;49:1283-9. 21. Dawson-Hughes B, Dallal GE, Krall EA, Sadowski L, Sahyoun N, Tannenbaum S. A controlled trial of the effect of calcium supplementation on bone density in postmenopausal women. N Engl J Med. 1990;323:878-83. 22. Smith KT, Heaney RP, Flora L, Hinders SM. Calcium absorption from a new calcium delivery system (CCM). Calcif Tissue Int. 1987;41:351-2. 23. Kriska AM, Sandler RB, Cauley JA, LaPorte RE, Horn DL, Pambianco G. The assessment of historical physical activity and its relation to adult bone parameters. Am J Epidemiol. 1988;127:1053-63. 24. Johnson J, Dawson-Hughes B. Precision and stability of dual-energy X-ray absorptiometry measurements. Calcif Tissue Int. 1991 ;49: 174-8. 25. Reinhardt TA, Horst RL, Orf JW, Hollis BW. A microassay for 1,25-dihydroxyvitamin D not requiring high performance liquid chromatography: application to clinical studies. J Clin Endocrinol Metab. 1984;58:91-8. 26. Preece MA, O'Riordan JL, Lawson DE*, Kodicek E. A competitive protein-binding assay for 25-hydroxycholecalciferol and 25-hydroxyergocalciferol in serum. Clin Chem Acta. 1974;54:235-42. 27. Kaplan A, Szabo LL. Clinical Chemistry: Interpretation and Techniques. 2d ed. Philadelphia: Lea & Febiger; 1983. 28. Snedecor GW, Cochran WG. Statistical Methods. 7th ed. Ames, Iowa: The Iowa State University Press; 1980. 29. National Research Council. Subcommittee on the Tenth Edition of the RDAs. Recommended dietary allowances: Subcommittee on the Tenth Edition of the RDAs, Food and Nutrition Board, Commission on Life Sciences, National Research Council. 10th ed. Washington, DC: National Academy Press; 1989.
512
30. Frost HM. The origin and nature of transients in human bone remodeling dynamics. In: Frame B, Parfitt AM, Duncan H; eds. Clinical Aspects of Metabolic Bone Disease. International Conference Series no. 270. Amsterdam: Excerpta Medica; 1973:124-37. 31. Lukert BP, Carey M, McCarty B, Tiemann S, Goodnight L, Helm M, et al. Influence of nutritional factors on calcium-regulating hormones and bone loss. Calcif Tissue Int. 1987;40:119-25. 32. Delvin EE, Imbach A, Copti M. Vitamin D nutritional status and related biochemical indices in an autonomous elderly population. Am J Clin Nutr. 1988;48:373-8. 33. Vir SC, Love AH. Nutritional status of institutionalized and noninstitutionalized aged in Belfast, Northern Ireland. Am J Clin Nutr. 1979;32:1934-47. 34. Lamberg-Allardt C. Vitamin D intake, sunlight exposure and 25hydroxyvitamin D levels in the elderly during one year. Ann Nutr Metab. 1984;28:144-50. 35. Abraham S, Carroll MD, Dresser CM, Johnston CL. Dietary Intake Findings, United States 1976-1980. Hyattsville, Maryland: Department of Health and Human Services; 1983. (DHHS publication no. (PHS) 83-1681.) 36. Cohn SH, Vaswani A, Zanzi I, Aloia JF, Roginsky MS, Ellis KJ. Changes in body chemical composition with age measured by totalbody neutron activation. Metabolism. 1976;25:85-95. 37. Krolner B, Toft B, Pors Nielsen S, Tondevold E. Physical exercise as prophylaxis against involutional vertebral bone loss: a controlled trial. Clin Sci. 1983;64:541-6. 38. Pocock NA, Eisman JA, Yeates MG, Sambrook PN, Eberl S. Physical fitness is a major determinant of femoral neck and lumbar spine bone mineral density. J Clin Invest. 1986;78:618-21. 39. Dalsky GP, Stocke KS, Ehsani AA, Slatopolsky E, Lee WC, Birge SJ Jr. Weight-bearing exercise training and lumbar bone mineral content in postmenopausal women. Ann Intern Med. 1988;108:824-8. 40. Touitou Y, Sulon J, Bogdan A, Reinberg A, Sodoyez JC, DemeyPonsart E. Adrenocortical hormones, ageing and mental condition: seasonal and circadian rhythms of plasma 18-hydroxy-l 1-deoxycorticosterone, total and free Cortisol and urinary corticosteroids. J Endocrinol. 1983;96:53-64. 41. Del Ponte A, Guagnano MT, Sensi S. Time-related behavior of endocrine secretion: circannual variations of FT 3 , Cortisol, HGH and serum basal insulin in healthy subjects. Chronobiol Int. 1984; 1:297-300. 42. Kennaway DJ, Royles P. Circadian rhythms of 6-sulphatoxy melatonin, Cortisol and electrolyte excretion of the summer and winter solstices in normal men and women. Acta Endocrinol (Copenh). 1986;13:450-6. 43. Nicolau GY, Lakatua D, Sackett-Lundeen L, Haus E. Circadian and circannual rhythms of hormonal variables in elderly men and women. Chronobiol Int. 1984;1:301-19.
1 October 1991 • Annals of Internal Medicine • Volume 115 • Number 7
Downloaded From: http://annals.org/pdfaccess.ashx?url=/data/journals/aim/19735/ by a University of California San Diego User on 04/30/2017
