- Email: [email protected]
Changes in bone mineral and bone formation rates during pregnancy and lactation in rats
Bone, 7, 283-287 (1986) Printed in the USA. All rights reserved.
Copyright
8756-3282186 $3.00 + .OO 0 1986 Pergamon Journals Ltd.
Changes in Bone Mineral and Bone Formation Rates during Pregnancy and Lactation in Rats SC.
MILLER,
J.G. SHUPE,
Division of Rad/obio/ogy,
E.H. REDD, M.A. MILLER, and T.H. OMURA
Department
Address for correspondence 84112, USA
of Pharmacology,
School of Medfone,
and reprints:: Dr. Scott C. Miller, Division of Radiobiology,
Abstract
of Utah, Salt Lake Oty. Utah, USA Bldg. 351, University of Utah, Salt Lake City, UT
there is a loss of bone mass during lactation, apparently irrespective of the amount of calcium in the diet (Spray, 1950; Komarkova et al., 1967; Brommage and DeLuca, 1985). The lactational bone loss is particularly evident in spongiosa regions of the skeleton (Ellinger et al., 1952), although cortical bone is also affected (Miller et al., 1982). Recent studies from our laboratory indicate that the rates of maternal endochondral bone elongation (Redd et al., 1984) and dentin apposition (Miller et al., 1985) are elevated during pregnancy, particularly at midpregnancy. These trends were reversed after parturition, during lactation. The purpose of the present study was to describe changes in bone mineral content, cortical and trabecular bone morphometry, and bone formation rates at the periosteal and endosteal surfaces dunng pregnancy and lactation. These results indicate that significant changes occur in skeletal modeling and metabolism during pregnancy and lactation in rats
Changes in bone chemistry, cortical bone morphometry, and periosteal and trabecular bone formation rates were determined during pregnancy and lactation in rats. Data were obtained using chemical, static morphometric, and fluorochrome-based histomorphometric methods in pregnant or lactating animals and compared with age-matched, unmated controls. There were significant increases in bone weight, ash weight, calcium content, and femoral cross-sectional area by late pregnancy and decreases in these same parameters during lactation. There were also decreases in bone mass and increases in surface:volume ratios of trabecular bone during lactation. There were increases in femoral periosteal and endosteal perimeters associated with the reproductive cycle. Bone formation and appositional rates at the periosteal and endosteal surfaces were elevated during pregnancy, particularly evident at midpregnancy. Periosteal bone formation rates declined during lactation, but trabecular bone formation rates increased. These results indicate that during pregnancy there are increases in bone formation rates contributing to the increases in skeletal mass. During lactation in rats, reductions in skeletal mass are accompanied by increases in bone turnover, particularly evident in trabecular bone. Key Words: Bone-Bone tation-Calcium.
Univemty
Materials and Methods Female Sprague-Dawley rats were obtatned at 60.-70 days of age and housed in a light and temperature-controlled environment The animals were fed a semlpurified diet containing 0 6% calcium and 0.5% phosphorus and allowed free access to distilled water When aged between 90 and 105 days, the rats were mated, and the first day of pregnancy was established by the presence of sperm in vaginal smears In each case, when pregnancy was es tablished, an age-matched, unmated, control rat was entered in the study. To equalize the calcium and phosphate demand on the mothers during lactation, the litter sizes were adjusted at 3 days postpartum to the weights of the mothers (9 pups. 290-320 g, 10 pups. 320 g) Five days before death, each rat (mated and control) was given an intraperitoneal injection of 30 mgikg body weight of Calcein (fluorescein-methylene-iminodiacetic acid. Sigma Chemical Co., St. LOUIS, MO) Three days later, each rat was given an intraperitoneal injection of 25 mg/kg body weight of tetracycline-HCI (Achromycin, Lederle Division, American CyanamIde Co Pearl River, NY) Rats were killed during pregnancy at day 7 (deslgnated as days after the pregnancy was established, P + 7), day 14 (P + 14), and day 21 (P + 21) and during lactation at day 7 (designated as days of lactation, L - 7), day 14 (L - 14). day 21 (L - 21), and day 28 (L - 28) All procedures were performed in the morning to minimize circadian effects. At least five pairs of rats were included in each experimental group. At autopsy the intact right femurs were cleaned and Immediately weighed. The bones were then dehydrated and defatted In acetone and anhydrous ethyl ether, dried, and reweighed The bones were ashed in a muffled furnace and the ash was weighed
Formation-Pregnancy-Lac-
Introduction There are significant alterations in calcium metabolism dunng pregnancy and lactation in mammals (Pitkln, 1975), but specific changes that might occur in osseous tissues are not well understood. Previous studies suggested that the maternal skeleton might accumulate calcium during pregnancy. Benzie et al. (1955) reported an increase in skeletal mass in ewes during pregnancy, and mineral balance studies suggest that the maternal skeleton accumulates calcium during pregnancy in humans (Heaney and Skillman, 1971) and rats (Goss and Schmidt, 1930). Lactation, on the other hand, is known to produce considerable demands on the calcium homeostatic system, particularly in those species with multiple offspring. In rats 283
284
SC. Miller et al
: Bone changes in pregnancy and lactation
The ash was then dissolved in hydrochloric acid, and total bone calcium content was determined by atomic absorption spectros-
by the second week of pregnancy pared with age-matched, unmated
copy. The left femurs and lumbar vertebrae were immersion fixed in acetone, defatted and dehydrated in anhydrous ethyl ether, and embedded undecalcified in polystyrene resin. Cross-sections of the femoral middiaphyseal shaft and longitudinal sectlons of the lumbar vertebrae were cut on a bone saw. The sections were mounted on plastic slides, ground to about 30 km in thickness, and stained with toluidine blue and basic fuchsin prior to vlewing by fluorescent microscopy. Three sections from the femoral middiaphyseal shaft from each animal were quantified using a digitizer to obtain the following parameters: periosteal perimeter, endosteal perimeter, total bone area, periosteal perimeter with a double label, and mineral appositional rate at the periosteal surface. The appositional rate was determined by dividing the distance between the double labels by the number of days between the Calcein and tetracycline injections. The bone formation rate was calculated by multiplying the fraction of double-labeled surface by the mineral appositional rate. Endosteal bone formation rates were determined from trabecular bone in the lumbar vertebral body The area sampled was 0.7 mm wide and extended from the middle of the proximal epiphysls to the distal epiphysis. The following parameters were measured in the trabecular bone using a digitizer. percentage of bone, surface:volume ratio, fraction of surface with a double label, and the mean distance between the double labels The appositional rate was corrected for obliqueness of the bone profile (Frost, 1983) The bone formation rate was calculated as described above. Within each experimental group the difference between pregnant or lactating rats and their controls was tested for significance
mained elevated during the third week (P + 21) of pregnancy. After parturition and during lactation, this trend was reversed. From the second week of lactation (L - 14) through the fourth week (L - 28), there were significant reductions in dry weight, ash weight, and total bone calcium. There were also significant reductions in the ash weight:dry weight ratio at the later stages of lactation. During pregnancy there were increases in the crosssectional area of cortical bone, medullary cavity area, endosteal perimeter, and periosteal perimeter (Table II). The cortical bone area declined during lactation, and the medullary cavity area and endosteal perimeter were increased over controls. The periosteal perimeter was significantly elevated over controls during the first half of the lactational period. The fraction of double-labeled surface and bone formation rates at the periosteal surface were increased throughout pregnancy. At midpregnancy (P + 14) and late pregnancy (P + 21), the periosteal apposltional rates were increased. During lactation this trend was reversed, with all periosteal bone formation parameters lower than controls by the third week of lactation (L - 21). After 4 weeks of lactation (L - 28) most of the pups had been weaned by the mothers, and there appeared to be some restoration of bone formation parameters to control levels
using the Student’s t-test. The data are expressed as the mean ? the standard deviation (SD)
Results There was a significant increase in wet weight, dry weight, ash weight, and total calcium content of the intact femurs
I). The dry weights,
ash weights,
(P + 14) when comcontrol animals (Table and calcium
P + 7 (7)b Control Pregnant P + 14 (5) Control Pregnant P + 21 (8) Control Pregnant L - 7 (5) Control Lactating L - 14 (6) Control Lactating L - 21 (6) Control Lactating L - 28 (6) Control Lactating
re-
(Table Ill). In trabecular bone of the lumbar vertebral body there was a significant increase at midpregnancy (P + 14) in the appositional rate and bone formation rate (Table IV). During lactation there was a progressive loss of bone mass and an increase in the surface:volume ratios of the trabecular bone. The fraction of double-labeled bone sur-
Table I. Changes in femoral weights and calcium content of rats during pregnancy or lactation and in age-matched, Group
content
unmated controls
Wet Weight (mg)”
Dry Weight (mg)
Ash Weight (mg)
% Ash of Dry Weight
847 ? 38 859 2 82
587 2 29 583 ? 48
388 ? 21 379 * 29
66.1 2 0.6 65.5 t 0.8
137 t 134 ”
825 2 39 941 f 98=
575 2 19 631 2 46c
375 ? 17 409 ” 306
65.2 2 1.4 64.9 2 1.2
137 2 7 153 * 1oc
912 T 100 985 f 118
594 2 76 668 * 68d
383 ? 14 424 ? 24e
62.7 ? 3.9 63.8 2 4.1
141 ? 6 154 rf- 12’
965 2 38 957 f 64
643 t 26 644 t 34
420 ? 15 416 ? 16
65.2 ? 1.0 64.6 ” 2.1
154 i 9 149 2 5
922 2 70 877 ? 26
674 ? 40 602 2 12e
425 2 27 365 2 18e
63.1 2 0.6 61.9 ? 1.8
149 2 9 133 ? 8e
964 _’ 85 859 2 59”
646 2 18 517 ? 29e
412 ? 10 307 * 20”
63.8 2 1.6 59.4 2 1 .oe
147 2 4 110 -t 8e
972 * 111 808 ? 87’
668 2 61 521 2 48e
435 ? 34 320 2 30e
65.3 2 1.2 61.5 2 1.g8
153 2 13 115 2 12e
a Data are expressed as means * SD b Number of animal pairs in each group c Significantly different from controls, P < 0.025. d Significantly different from controls, P < 0.05. e Significantly different from controls, P < 0.005. ’ Significantly different from controls, P < 0.01.
Total Calcium (mg)
11 14
285
SC. Miller et al.: Bone changes in pregnancy and lactation
Table II. Changes in femoral cortical bone measured from middiaphyseal cross-sections in pregnant and lactating rats and agematched, unmated controls. Medullary Cavity Area (mm2)
Endosteal Perimeter (mm)
Periosteal Perimeter (mm)
5.35 t 0.33 5.39 * 0.31
3.66 2 0.17 3.79 * 0.43
6.99 ? 0.27 7.21 ~fr0.49
11.03 * 0.13 10.95 2 0.33
5.53 2 0.09 5.75 -+ 0.27
3.19 * 0.44 3.28 f 0.36
6.58 2 0.44 6.77 ? 0.40
10.66 !z 0.23 11 08 2 O.lgc
5.55 2 0.20 6.00 * O.&r
3.16 2 0.30 3.74 2 0.368
6.57 ‘-c 0.34 7.17 2 0.368
10.80 ? 0.26 11.23 + 0.41’
5.89 2 0.40 6.30 2 0.38
3.34 f 0.38 3.65 ~fr0.34
6.75 -t- 0.38 7.04 2 0.27
10.96 ‘- 0.47 11.54 2 0.32d
6.09 f 0.32 6.24 2 0.18
3.14 * 0.44 4.06 ? 0.70’
6.67 * 0.66 7.58 f 0.60e
10.87 ? 0.36 11.72 ? 0.3ge
5.99 ? 0.26 5.08 2 0.3g8
2.91 2 0.17 4.17 2 0.63e
6.38 ? 0.18 7 65 * 0.54e
10.93 5 0.27 11.12 ? 0.40
6.39 2 0.75 5.05 ? 0.46e
3.25 * 0.39 3.92 & 0.26e
6.72 ? 0 53 7.63 2 0.72e
11.33 2 0.61 11.17 2 0.38
Bone Area (mm2)a
Group P + 7 (7)b Control Preonant P + 14 (5) Control Preanant P + 2’1 (8) Control Pregnant L - 7 (5) Control Lactatina L - 14(6) Control Lactating L - 21 (6) Control ’ Lactating L - 28 (6) Control Lactating
a Data are expressed as means ? SD. b Number of animal pairs in each group. c Significantly different from controls, P < d Significantly different from controls, P < BSignificantly different from controls, P < f Significantly different from controls, P <
0.01. 0.05. 0.005 0.025.
face was markedly increased by the third week of lactation (L + 21),resulting in significant increases in the bone formation rate. After 4 weeks of lactation (L - 28), when most
pups had been weaned from the mothers, the fraction of double-labeled surface, the appositional rate, and the bone formation rate were increased over controls.
Table Ill. Changes in femoral periosteal bone format’mn parameters measured from middiaphyseal cross-sections in pregnant and lactating rats and age-matched, unmated controls
Group P + 7 (7)b Control Pregnant P + 14 (5) Control Pregnant P + 21 (8) Control Pregnant L - 7 (5) Control Lactating L - 14 (6) Control Lactating L - 21 (6) Control Lactating L - 28 (6) Control Lactating
Fraction of Double-Labeled Surfacea
Appositional Rate (pm/year)
Bone Formation Rate (mm2/mm/year)
0.321 2 0.079 0.400 2 0.097c
2.0 + 0.4 2.2 + 0 4
0 24 t 0 07 0.33 -e 0 15c
0.279 2 0.087 0.525 r 0.258c
17201 2 8 i 0.4d
0.18 ” 0.04 0.51 -t 0 34e
0.358 ? 0.139 0 484 2 0.096=
1 6 ” 0.5 2 0 ? 0.4c
0.22 + 0 12 0.35 2 0 07e
0.244 2 0.120 0.309 2 0.148
1 5 + 0.1 1 4 i 0.1
0 13 2 0.06 0.16 ? 0 08
0.264 2 0.110 0.207 2 0.134
14?01 1 2 lr- 0.2c
0.14 -+ 0.06 0.09 c 0 05
0.295 ? 0.248 0.039 2 0.046c
1 3 t 0.1 1 1 ” 0.2c
015 i 014 0.01 t 0 01e
0.192 2 0.152 0.125 2 0.098
0.9 * 0.2 0.9 ? 0.2
011 t 013 0.04 2 0 04
a Data are expressed as means ? SD. b Number of animal pairs in each group. c Significantly different from controls, P < 0.05. d Significantly different from controls, P C 0.005. e Significantly different from controls, P < 0.025.
286
S.C Miller et al.: Bone changes
Table IV. Changes
in bone and bone formation parameters lactation and in age-matched, unmated controls.
Group P + 7 (7)b Control Pregnant P + 14 (6) Control Pregnant P + 21 (7) Contrdl ’ Pregnant L - 7 (5) Control Lactating L - 14 (6) Control Lactating L - 21 (6) Control Lactating L - 28 (6) Control Lactating
of trabecular
bone in the lumbar vertebrae
In pregnancy of rats dunng
and lactation pregnancy
% Bonea
Surface. Volume Ratio (mm2/mm3)
Fraction of Double-labeled Surface
39.8 2 10.8 41.5 k 5.8
22 4 -r- 3.4 224 2 27
0 055 2 0 026 0 045 -’ 0 028
16202 16kO3
0.03 2 0.02 0 03 t 0 02
41.7 t 4.9 42.9 k 3.1
24 2 2 3.3 238273
0048 k 0018 0.067 k 0.029
16201 19kO4C
003 2 001 0 05 t 0 02d
39.5 2 5 5 424 + 10.8
209 231
? 2.1 271
0 080 k 0.045 0071 k 0.040
16204 1 5 2 0.2
0 04 -+ 0 03 0 04 i- 0 02
40.2 2 9.0 362263
21.2 t 56 22.2 t 4.8
0 073 t 0.053 0 029 t 0.026
1.8 t 0.4 20206
0 05 -t 0 04 0 02 k 0 01
423 k 52 30 4 ” 1 7e
21 8 t 3.9 26 8 2 2.5c
0 094 k 0.035 0 083 + 0.023
19104 16504
0.06 k 0 03 0 05 t 0 02
41.6 2 72 21 5 5 6.4e
19 1 2 2.3 31 7 -c 71e
0.079 + 0 040 0.229 i- 0 086”
1.6 k 06 19203
0 05 2 0 04 016 k 008”
358 i- 11 0 21 4 ~fr 82c
23.9 k 5.5 25 8 k 5.6
0 058 -c 0.037 0 250 t 0 029e
17 -+ 04 2 1 i 0.2d
0 04 2 0 03 019 -‘003e
AppositIonal Rate (wmiday)
or
Bone FormatIon Rate (mmz/mmlvear)
a Data are expressed as means k SD b Number of animal pairs in each group c Significantly different from controls, P < 0.025. d Significantly different from controls, P < 0.05. e Significantly different from controls, P < 0 005
Discussion Perhaps the most striking observation made in this study was the increase in bone formation parameters during pregnancy. Periosteal bone formation rates were elevated throughout pregnancy, with peak values at midpregnancy. The endosteal bone formation rate was also increased at midpregnancy. The observation that formation Indices were increased over controls at midpregnancy is consistent with our previous observations of increased maternal endochondral bone elongation (Redd et al., 1984) and dentin apposItIonal rates (Miller et al., 1985), also at midpregnancy. The increases tn bone formation rates and endochondral growth rates would account for the increases in bone weights and calcium content during pregnancy, as observed in this study. These results support earlier work showing that the maternal skeleton stores calcium during pregnancy, perhaps in excess of fetal need (Goss and Schmidt, 1930; Spray, 1950). There were small but significant increases in femoral periosteal perimeters from midpregnancy to midlactation. These increases were likely due to accelerated bone formation rates at this surface. The reason the periosteal perimeter did not remain elevated over the controls In late lactation was probably due to decreased periosteal bone formation rates during lactation. This would allow normal periosteal formation in the control rats, which contmue to model their bones, to catch up with the bones from lactating mothers. In this regard it is interesting that Smith (1967) has noted increased periosteal diameter and cortical bone thickness in women with four or more children when compared with nulliparous women. This suggests accelerated periosteal bone formation rates with the reproductive cycle in humans, consistent with observations made in the present study.
Whereas periosteal bone formation rates were decreased during lactation, there were increased trabecular bone formation rates at the later stages of lactation. The fraction of double-labeled surface was substantially increased In late lactation, although bone surface perimeter was decreased due to lactational bone loss. This supports a previous study that demonstrated an increase In tetracycline-labeled trabecular bone surface during lactation In vitamin D-replete rats (Miller et al., 1982). The mineral appositional rate was also elevated over controls during the fourth week of lactation, a period when most pups had been weaned. The elevated bone formation rates during late lactation with reductions in bone mass and elevated surface:volume ratios indicate an Increase in bone turnover during lactation. Accelerated bone turnover during lactation has also been suggested from isotopic studies in calcium-deficient rats (Wong et al., 1981) Thus, it appears that changes In skeletal modeling and perhaps remodeling during pregnancy and lactation are locatlon-speclflc, that IS, the periosteal surface appears to respond somewhat differently from the trabecular endosteal surface The marked decreases in bone weight, calcium content. cortical bone area, and percentage of trabecular bone and the increase in trabecular bone surface:volume ratios during lactation were certainly expected from previous studtes on lactational bone loss in the rat (Komarkova et al., 1967; Brommage and DeLuca, 1985). The Increases in endosteal bone formation rates at the end of lactation might represent a mechanism to restore mineral to the maternal skeleton preparatory to the next reproductive cycle. The extent of reconstitution of the bone mass lost following lactation is not well establlshed. It would be of considerable interest to define the endocrine changes that might be responsible for accelerated bone formation rates observed during pregnancy and
S C Miller et al
Bone changes
in pregnancy
and lactation
bone loss during lactatron. There are a number of hormones that might be involved in maternal skeletal metabolism during pregnancy, and some have been discussed in previous reports (Redd et al., 1984; Miller et al., 1985). Perhaps the best correlation that we have found between the midpregnancy peaks in bone formation rates, as observed in this study, and previously reported mrdpregnancy peaks in dentin appositron (Miller et al., 1985) and endochondral growth (Redd et al., 1984) is with placental lactogens and somatomedins. Placental lactogen in the rat reaches its peak dunng the second week of pregnancy (Tonkowicz and Voogt, 1983) and other lactogens might appear at midpregnancy (Robertson and Friesen, 1981), during the transition from pituitary to placental control of pregnancy. The effects of some lactogens might be mediated via the somatomedin system (Hurley et al., 1977; Daughaday and Kapadia, 1978), and somatomedins are known to stimulate bone formation in vitro (Canalis et al., 1977). Thus, rt is not unreasonable to suspect involvement of the somatomedin system in some of the skeletal changes that occur during pregnancy. The endocrine events responsible for lactational bone loss are not clear. The loss of bone can occur independent of the vitamin D endocrine system (Halloran and DeLuca, 1980; Miller et al., 1982) and perhaps independent of parathyrord hormone, calcitonin, estrogens, and glucocorticords (Brommage and DeLuca, 1985). The regulation of this process might possibly involve hormones, such as prolactin, or yet to be identified hormonal effecters (Brommage and DeLuca, 1985).
Acknowledgement. This work was supported by Grant DE-06007 from the U.S. Public Health Service and Contract DEAC0276EY00119 from the U S. Department of Energy
References Benzle D
Boyne A D Dalgarno AC Duckworth J.. HIII R and Walker D M Studies of the skeleton of the sheep I The effect of different levels of dietary caiclum during pregnancy and lactation on Individual
bones J. Agnc. So 46 425444, 1955. Brommage R and DeLuca H F Regulatron of bone mrneral loss during lactation Am J. Physioi 246 E182-187, 1985 Canalrs E M Hlntr R L Dretrich J W Malna D M and Rarsz L G Effect of
207
somatomedin
and growth hormone on bone collagen synthesis in vitro
Metaboiism 26 1079- 1087, 1977 Daughaday W H and Kapadra M Maintenance trvity rn hypophysectomlzed
pregnant
of serum somatomedln
rats Endocrinology
ac
102 1317
1320. 1978 Ellrnger G M Duckworth J Dalgarno A C and Ouenoullle M H Skeletal changes durrng pregnancy and lactatron in the rat Effect of drfferenl levels of dietary calcrum 81 J Nulr 6 235-253 1952 Frost H M Bone histomorphometry Analysis of the labeling “escape error In Bone H/stomorphometry Techniques and interpretation R Reeker, ed CRC Press, Boca Raton. FL. 1983, pp 133- 142 Goss H and Schmrdt C L A Calcum and phosphorus metabolrsm in rats durrng pregnancy and lactation and the rnfluence of the reaction of the diet thereon J. Biol Chem 66 417-432. 1930 Halloran B P and DeLuca H F Skeletal changes during pregnancy and lactatron The role of vrtamin D Endocnnoiogy 107 1923- 1928. 1980 Heaney R P and Skrllman T G Calcrum metabolrsm in normal human pregnancy J Clan Endocnnol 33 661-670. 1971 Hurley T S D’Ercole A J Handwerger S Underwood L E Furlanetto R W and Fellows R E Ovrne placental lactogen induces somatomedrn A possible role rn fetal growth Endocnnoiogy 101 1635-1638. 1977 Komarkova A, Zahor Z and Crabanova V The effect of lactation on the composrtron of long bone in rats J Lab Ci~n Med 69 102-109. 1967 Mrller S C Omura T H and Smrth L J Changes in dentrn appositional rates durrng pregnancy and lactation In rats J Dent Res 64 10621064, 1985 Mrller S C Halloran B P DeLuca H F and Jee W S S Role of vrtamrn D rn maternal skeletal changes durrng pregnancy and lactatron a hrslomorphometnc study Caini Tissue Inf 34 245-252. 1982 Prtkrn R M Calcium metabolrsm rn pregnancy A revrew Am. J Obsref Gynecoi 121 724-737, 1975 Redd E H Miller SC and Jee W S S Changes In endochondral bone elongatron rates durrng pregnancy and lactation in rats Calcif Issue Inf. 36 697-701, 1984 Robertson M C and Frresen H G Two forms of rat placental lactogen revealed by radroimmunoassay Endocnnology 106 238882390. 1981 Smrth R W Jr Dietary and hormonal factors rn bone loss fed Proc 26 1737m 1746, 1967 Spray C M A study of some aspects of reproduction by means of chemical analysis Er J Nuti 4:354-360, 1950 Tonkowrcz P A and Voogt J L Examlnatron of rat placental lactogen and prolactrn at 6-hr Intervals dunng mrdpregnancy Proc Sot Exp B/o/ Med 173 583-587. 1983 Wong K M , Singer L Ophaug R H and Klern L Effect of lactation and calcrum deficiency, and of fluoride Intake, on bone turnover rn rats Isotopic measurements of bone resorption and formatton J Nutr 111 184881854, 1981
Recewed May 28. 1985 Rewsed February 1 I. 1986 Accepted February 13, 1986
Copyright
8756-3282186 $3.00 + .OO 0 1986 Pergamon Journals Ltd.
Changes in Bone Mineral and Bone Formation Rates during Pregnancy and Lactation in Rats SC.
MILLER,
J.G. SHUPE,
Division of Rad/obio/ogy,
E.H. REDD, M.A. MILLER, and T.H. OMURA
Department
Address for correspondence 84112, USA
of Pharmacology,
School of Medfone,
and reprints:: Dr. Scott C. Miller, Division of Radiobiology,
Abstract
of Utah, Salt Lake Oty. Utah, USA Bldg. 351, University of Utah, Salt Lake City, UT
there is a loss of bone mass during lactation, apparently irrespective of the amount of calcium in the diet (Spray, 1950; Komarkova et al., 1967; Brommage and DeLuca, 1985). The lactational bone loss is particularly evident in spongiosa regions of the skeleton (Ellinger et al., 1952), although cortical bone is also affected (Miller et al., 1982). Recent studies from our laboratory indicate that the rates of maternal endochondral bone elongation (Redd et al., 1984) and dentin apposition (Miller et al., 1985) are elevated during pregnancy, particularly at midpregnancy. These trends were reversed after parturition, during lactation. The purpose of the present study was to describe changes in bone mineral content, cortical and trabecular bone morphometry, and bone formation rates at the periosteal and endosteal surfaces dunng pregnancy and lactation. These results indicate that significant changes occur in skeletal modeling and metabolism during pregnancy and lactation in rats
Changes in bone chemistry, cortical bone morphometry, and periosteal and trabecular bone formation rates were determined during pregnancy and lactation in rats. Data were obtained using chemical, static morphometric, and fluorochrome-based histomorphometric methods in pregnant or lactating animals and compared with age-matched, unmated controls. There were significant increases in bone weight, ash weight, calcium content, and femoral cross-sectional area by late pregnancy and decreases in these same parameters during lactation. There were also decreases in bone mass and increases in surface:volume ratios of trabecular bone during lactation. There were increases in femoral periosteal and endosteal perimeters associated with the reproductive cycle. Bone formation and appositional rates at the periosteal and endosteal surfaces were elevated during pregnancy, particularly evident at midpregnancy. Periosteal bone formation rates declined during lactation, but trabecular bone formation rates increased. These results indicate that during pregnancy there are increases in bone formation rates contributing to the increases in skeletal mass. During lactation in rats, reductions in skeletal mass are accompanied by increases in bone turnover, particularly evident in trabecular bone. Key Words: Bone-Bone tation-Calcium.
Univemty
Materials and Methods Female Sprague-Dawley rats were obtatned at 60.-70 days of age and housed in a light and temperature-controlled environment The animals were fed a semlpurified diet containing 0 6% calcium and 0.5% phosphorus and allowed free access to distilled water When aged between 90 and 105 days, the rats were mated, and the first day of pregnancy was established by the presence of sperm in vaginal smears In each case, when pregnancy was es tablished, an age-matched, unmated, control rat was entered in the study. To equalize the calcium and phosphate demand on the mothers during lactation, the litter sizes were adjusted at 3 days postpartum to the weights of the mothers (9 pups. 290-320 g, 10 pups. 320 g) Five days before death, each rat (mated and control) was given an intraperitoneal injection of 30 mgikg body weight of Calcein (fluorescein-methylene-iminodiacetic acid. Sigma Chemical Co., St. LOUIS, MO) Three days later, each rat was given an intraperitoneal injection of 25 mg/kg body weight of tetracycline-HCI (Achromycin, Lederle Division, American CyanamIde Co Pearl River, NY) Rats were killed during pregnancy at day 7 (deslgnated as days after the pregnancy was established, P + 7), day 14 (P + 14), and day 21 (P + 21) and during lactation at day 7 (designated as days of lactation, L - 7), day 14 (L - 14). day 21 (L - 21), and day 28 (L - 28) All procedures were performed in the morning to minimize circadian effects. At least five pairs of rats were included in each experimental group. At autopsy the intact right femurs were cleaned and Immediately weighed. The bones were then dehydrated and defatted In acetone and anhydrous ethyl ether, dried, and reweighed The bones were ashed in a muffled furnace and the ash was weighed
Formation-Pregnancy-Lac-
Introduction There are significant alterations in calcium metabolism dunng pregnancy and lactation in mammals (Pitkln, 1975), but specific changes that might occur in osseous tissues are not well understood. Previous studies suggested that the maternal skeleton might accumulate calcium during pregnancy. Benzie et al. (1955) reported an increase in skeletal mass in ewes during pregnancy, and mineral balance studies suggest that the maternal skeleton accumulates calcium during pregnancy in humans (Heaney and Skillman, 1971) and rats (Goss and Schmidt, 1930). Lactation, on the other hand, is known to produce considerable demands on the calcium homeostatic system, particularly in those species with multiple offspring. In rats 283
284
SC. Miller et al
: Bone changes in pregnancy and lactation
The ash was then dissolved in hydrochloric acid, and total bone calcium content was determined by atomic absorption spectros-
by the second week of pregnancy pared with age-matched, unmated
copy. The left femurs and lumbar vertebrae were immersion fixed in acetone, defatted and dehydrated in anhydrous ethyl ether, and embedded undecalcified in polystyrene resin. Cross-sections of the femoral middiaphyseal shaft and longitudinal sectlons of the lumbar vertebrae were cut on a bone saw. The sections were mounted on plastic slides, ground to about 30 km in thickness, and stained with toluidine blue and basic fuchsin prior to vlewing by fluorescent microscopy. Three sections from the femoral middiaphyseal shaft from each animal were quantified using a digitizer to obtain the following parameters: periosteal perimeter, endosteal perimeter, total bone area, periosteal perimeter with a double label, and mineral appositional rate at the periosteal surface. The appositional rate was determined by dividing the distance between the double labels by the number of days between the Calcein and tetracycline injections. The bone formation rate was calculated by multiplying the fraction of double-labeled surface by the mineral appositional rate. Endosteal bone formation rates were determined from trabecular bone in the lumbar vertebral body The area sampled was 0.7 mm wide and extended from the middle of the proximal epiphysls to the distal epiphysis. The following parameters were measured in the trabecular bone using a digitizer. percentage of bone, surface:volume ratio, fraction of surface with a double label, and the mean distance between the double labels The appositional rate was corrected for obliqueness of the bone profile (Frost, 1983) The bone formation rate was calculated as described above. Within each experimental group the difference between pregnant or lactating rats and their controls was tested for significance
mained elevated during the third week (P + 21) of pregnancy. After parturition and during lactation, this trend was reversed. From the second week of lactation (L - 14) through the fourth week (L - 28), there were significant reductions in dry weight, ash weight, and total bone calcium. There were also significant reductions in the ash weight:dry weight ratio at the later stages of lactation. During pregnancy there were increases in the crosssectional area of cortical bone, medullary cavity area, endosteal perimeter, and periosteal perimeter (Table II). The cortical bone area declined during lactation, and the medullary cavity area and endosteal perimeter were increased over controls. The periosteal perimeter was significantly elevated over controls during the first half of the lactational period. The fraction of double-labeled surface and bone formation rates at the periosteal surface were increased throughout pregnancy. At midpregnancy (P + 14) and late pregnancy (P + 21), the periosteal apposltional rates were increased. During lactation this trend was reversed, with all periosteal bone formation parameters lower than controls by the third week of lactation (L - 21). After 4 weeks of lactation (L - 28) most of the pups had been weaned by the mothers, and there appeared to be some restoration of bone formation parameters to control levels
using the Student’s t-test. The data are expressed as the mean ? the standard deviation (SD)
Results There was a significant increase in wet weight, dry weight, ash weight, and total calcium content of the intact femurs
I). The dry weights,
ash weights,
(P + 14) when comcontrol animals (Table and calcium
P + 7 (7)b Control Pregnant P + 14 (5) Control Pregnant P + 21 (8) Control Pregnant L - 7 (5) Control Lactating L - 14 (6) Control Lactating L - 21 (6) Control Lactating L - 28 (6) Control Lactating
re-
(Table Ill). In trabecular bone of the lumbar vertebral body there was a significant increase at midpregnancy (P + 14) in the appositional rate and bone formation rate (Table IV). During lactation there was a progressive loss of bone mass and an increase in the surface:volume ratios of the trabecular bone. The fraction of double-labeled bone sur-
Table I. Changes in femoral weights and calcium content of rats during pregnancy or lactation and in age-matched, Group
content
unmated controls
Wet Weight (mg)”
Dry Weight (mg)
Ash Weight (mg)
% Ash of Dry Weight
847 ? 38 859 2 82
587 2 29 583 ? 48
388 ? 21 379 * 29
66.1 2 0.6 65.5 t 0.8
137 t 134 ”
825 2 39 941 f 98=
575 2 19 631 2 46c
375 ? 17 409 ” 306
65.2 2 1.4 64.9 2 1.2
137 2 7 153 * 1oc
912 T 100 985 f 118
594 2 76 668 * 68d
383 ? 14 424 ? 24e
62.7 ? 3.9 63.8 2 4.1
141 ? 6 154 rf- 12’
965 2 38 957 f 64
643 t 26 644 t 34
420 ? 15 416 ? 16
65.2 ? 1.0 64.6 ” 2.1
154 i 9 149 2 5
922 2 70 877 ? 26
674 ? 40 602 2 12e
425 2 27 365 2 18e
63.1 2 0.6 61.9 ? 1.8
149 2 9 133 ? 8e
964 _’ 85 859 2 59”
646 2 18 517 ? 29e
412 ? 10 307 * 20”
63.8 2 1.6 59.4 2 1 .oe
147 2 4 110 -t 8e
972 * 111 808 ? 87’
668 2 61 521 2 48e
435 ? 34 320 2 30e
65.3 2 1.2 61.5 2 1.g8
153 2 13 115 2 12e
a Data are expressed as means * SD b Number of animal pairs in each group c Significantly different from controls, P < 0.025. d Significantly different from controls, P < 0.05. e Significantly different from controls, P < 0.005. ’ Significantly different from controls, P < 0.01.
Total Calcium (mg)
11 14
285
SC. Miller et al.: Bone changes in pregnancy and lactation
Table II. Changes in femoral cortical bone measured from middiaphyseal cross-sections in pregnant and lactating rats and agematched, unmated controls. Medullary Cavity Area (mm2)
Endosteal Perimeter (mm)
Periosteal Perimeter (mm)
5.35 t 0.33 5.39 * 0.31
3.66 2 0.17 3.79 * 0.43
6.99 ? 0.27 7.21 ~fr0.49
11.03 * 0.13 10.95 2 0.33
5.53 2 0.09 5.75 -+ 0.27
3.19 * 0.44 3.28 f 0.36
6.58 2 0.44 6.77 ? 0.40
10.66 !z 0.23 11 08 2 O.lgc
5.55 2 0.20 6.00 * O.&r
3.16 2 0.30 3.74 2 0.368
6.57 ‘-c 0.34 7.17 2 0.368
10.80 ? 0.26 11.23 + 0.41’
5.89 2 0.40 6.30 2 0.38
3.34 f 0.38 3.65 ~fr0.34
6.75 -t- 0.38 7.04 2 0.27
10.96 ‘- 0.47 11.54 2 0.32d
6.09 f 0.32 6.24 2 0.18
3.14 * 0.44 4.06 ? 0.70’
6.67 * 0.66 7.58 f 0.60e
10.87 ? 0.36 11.72 ? 0.3ge
5.99 ? 0.26 5.08 2 0.3g8
2.91 2 0.17 4.17 2 0.63e
6.38 ? 0.18 7 65 * 0.54e
10.93 5 0.27 11.12 ? 0.40
6.39 2 0.75 5.05 ? 0.46e
3.25 * 0.39 3.92 & 0.26e
6.72 ? 0 53 7.63 2 0.72e
11.33 2 0.61 11.17 2 0.38
Bone Area (mm2)a
Group P + 7 (7)b Control Preonant P + 14 (5) Control Preanant P + 2’1 (8) Control Pregnant L - 7 (5) Control Lactatina L - 14(6) Control Lactating L - 21 (6) Control ’ Lactating L - 28 (6) Control Lactating
a Data are expressed as means ? SD. b Number of animal pairs in each group. c Significantly different from controls, P < d Significantly different from controls, P < BSignificantly different from controls, P < f Significantly different from controls, P <
0.01. 0.05. 0.005 0.025.
face was markedly increased by the third week of lactation (L + 21),resulting in significant increases in the bone formation rate. After 4 weeks of lactation (L - 28), when most
pups had been weaned from the mothers, the fraction of double-labeled surface, the appositional rate, and the bone formation rate were increased over controls.
Table Ill. Changes in femoral periosteal bone format’mn parameters measured from middiaphyseal cross-sections in pregnant and lactating rats and age-matched, unmated controls
Group P + 7 (7)b Control Pregnant P + 14 (5) Control Pregnant P + 21 (8) Control Pregnant L - 7 (5) Control Lactating L - 14 (6) Control Lactating L - 21 (6) Control Lactating L - 28 (6) Control Lactating
Fraction of Double-Labeled Surfacea
Appositional Rate (pm/year)
Bone Formation Rate (mm2/mm/year)
0.321 2 0.079 0.400 2 0.097c
2.0 + 0.4 2.2 + 0 4
0 24 t 0 07 0.33 -e 0 15c
0.279 2 0.087 0.525 r 0.258c
17201 2 8 i 0.4d
0.18 ” 0.04 0.51 -t 0 34e
0.358 ? 0.139 0 484 2 0.096=
1 6 ” 0.5 2 0 ? 0.4c
0.22 + 0 12 0.35 2 0 07e
0.244 2 0.120 0.309 2 0.148
1 5 + 0.1 1 4 i 0.1
0 13 2 0.06 0.16 ? 0 08
0.264 2 0.110 0.207 2 0.134
14?01 1 2 lr- 0.2c
0.14 -+ 0.06 0.09 c 0 05
0.295 ? 0.248 0.039 2 0.046c
1 3 t 0.1 1 1 ” 0.2c
015 i 014 0.01 t 0 01e
0.192 2 0.152 0.125 2 0.098
0.9 * 0.2 0.9 ? 0.2
011 t 013 0.04 2 0 04
a Data are expressed as means ? SD. b Number of animal pairs in each group. c Significantly different from controls, P < 0.05. d Significantly different from controls, P C 0.005. e Significantly different from controls, P < 0.025.
286
S.C Miller et al.: Bone changes
Table IV. Changes
in bone and bone formation parameters lactation and in age-matched, unmated controls.
Group P + 7 (7)b Control Pregnant P + 14 (6) Control Pregnant P + 21 (7) Contrdl ’ Pregnant L - 7 (5) Control Lactating L - 14 (6) Control Lactating L - 21 (6) Control Lactating L - 28 (6) Control Lactating
of trabecular
bone in the lumbar vertebrae
In pregnancy of rats dunng
and lactation pregnancy
% Bonea
Surface. Volume Ratio (mm2/mm3)
Fraction of Double-labeled Surface
39.8 2 10.8 41.5 k 5.8
22 4 -r- 3.4 224 2 27
0 055 2 0 026 0 045 -’ 0 028
16202 16kO3
0.03 2 0.02 0 03 t 0 02
41.7 t 4.9 42.9 k 3.1
24 2 2 3.3 238273
0048 k 0018 0.067 k 0.029
16201 19kO4C
003 2 001 0 05 t 0 02d
39.5 2 5 5 424 + 10.8
209 231
? 2.1 271
0 080 k 0.045 0071 k 0.040
16204 1 5 2 0.2
0 04 -+ 0 03 0 04 i- 0 02
40.2 2 9.0 362263
21.2 t 56 22.2 t 4.8
0 073 t 0.053 0 029 t 0.026
1.8 t 0.4 20206
0 05 -t 0 04 0 02 k 0 01
423 k 52 30 4 ” 1 7e
21 8 t 3.9 26 8 2 2.5c
0 094 k 0.035 0 083 + 0.023
19104 16504
0.06 k 0 03 0 05 t 0 02
41.6 2 72 21 5 5 6.4e
19 1 2 2.3 31 7 -c 71e
0.079 + 0 040 0.229 i- 0 086”
1.6 k 06 19203
0 05 2 0 04 016 k 008”
358 i- 11 0 21 4 ~fr 82c
23.9 k 5.5 25 8 k 5.6
0 058 -c 0.037 0 250 t 0 029e
17 -+ 04 2 1 i 0.2d
0 04 2 0 03 019 -‘003e
AppositIonal Rate (wmiday)
or
Bone FormatIon Rate (mmz/mmlvear)
a Data are expressed as means k SD b Number of animal pairs in each group c Significantly different from controls, P < 0.025. d Significantly different from controls, P < 0.05. e Significantly different from controls, P < 0 005
Discussion Perhaps the most striking observation made in this study was the increase in bone formation parameters during pregnancy. Periosteal bone formation rates were elevated throughout pregnancy, with peak values at midpregnancy. The endosteal bone formation rate was also increased at midpregnancy. The observation that formation Indices were increased over controls at midpregnancy is consistent with our previous observations of increased maternal endochondral bone elongation (Redd et al., 1984) and dentin apposItIonal rates (Miller et al., 1985), also at midpregnancy. The increases tn bone formation rates and endochondral growth rates would account for the increases in bone weights and calcium content during pregnancy, as observed in this study. These results support earlier work showing that the maternal skeleton stores calcium during pregnancy, perhaps in excess of fetal need (Goss and Schmidt, 1930; Spray, 1950). There were small but significant increases in femoral periosteal perimeters from midpregnancy to midlactation. These increases were likely due to accelerated bone formation rates at this surface. The reason the periosteal perimeter did not remain elevated over the controls In late lactation was probably due to decreased periosteal bone formation rates during lactation. This would allow normal periosteal formation in the control rats, which contmue to model their bones, to catch up with the bones from lactating mothers. In this regard it is interesting that Smith (1967) has noted increased periosteal diameter and cortical bone thickness in women with four or more children when compared with nulliparous women. This suggests accelerated periosteal bone formation rates with the reproductive cycle in humans, consistent with observations made in the present study.
Whereas periosteal bone formation rates were decreased during lactation, there were increased trabecular bone formation rates at the later stages of lactation. The fraction of double-labeled surface was substantially increased In late lactation, although bone surface perimeter was decreased due to lactational bone loss. This supports a previous study that demonstrated an increase In tetracycline-labeled trabecular bone surface during lactation In vitamin D-replete rats (Miller et al., 1982). The mineral appositional rate was also elevated over controls during the fourth week of lactation, a period when most pups had been weaned. The elevated bone formation rates during late lactation with reductions in bone mass and elevated surface:volume ratios indicate an Increase in bone turnover during lactation. Accelerated bone turnover during lactation has also been suggested from isotopic studies in calcium-deficient rats (Wong et al., 1981) Thus, it appears that changes In skeletal modeling and perhaps remodeling during pregnancy and lactation are locatlon-speclflc, that IS, the periosteal surface appears to respond somewhat differently from the trabecular endosteal surface The marked decreases in bone weight, calcium content. cortical bone area, and percentage of trabecular bone and the increase in trabecular bone surface:volume ratios during lactation were certainly expected from previous studtes on lactational bone loss in the rat (Komarkova et al., 1967; Brommage and DeLuca, 1985). The Increases in endosteal bone formation rates at the end of lactation might represent a mechanism to restore mineral to the maternal skeleton preparatory to the next reproductive cycle. The extent of reconstitution of the bone mass lost following lactation is not well establlshed. It would be of considerable interest to define the endocrine changes that might be responsible for accelerated bone formation rates observed during pregnancy and
S C Miller et al
Bone changes
in pregnancy
and lactation
bone loss during lactatron. There are a number of hormones that might be involved in maternal skeletal metabolism during pregnancy, and some have been discussed in previous reports (Redd et al., 1984; Miller et al., 1985). Perhaps the best correlation that we have found between the midpregnancy peaks in bone formation rates, as observed in this study, and previously reported mrdpregnancy peaks in dentin appositron (Miller et al., 1985) and endochondral growth (Redd et al., 1984) is with placental lactogens and somatomedins. Placental lactogen in the rat reaches its peak dunng the second week of pregnancy (Tonkowicz and Voogt, 1983) and other lactogens might appear at midpregnancy (Robertson and Friesen, 1981), during the transition from pituitary to placental control of pregnancy. The effects of some lactogens might be mediated via the somatomedin system (Hurley et al., 1977; Daughaday and Kapadia, 1978), and somatomedins are known to stimulate bone formation in vitro (Canalis et al., 1977). Thus, rt is not unreasonable to suspect involvement of the somatomedin system in some of the skeletal changes that occur during pregnancy. The endocrine events responsible for lactational bone loss are not clear. The loss of bone can occur independent of the vitamin D endocrine system (Halloran and DeLuca, 1980; Miller et al., 1982) and perhaps independent of parathyrord hormone, calcitonin, estrogens, and glucocorticords (Brommage and DeLuca, 1985). The regulation of this process might possibly involve hormones, such as prolactin, or yet to be identified hormonal effecters (Brommage and DeLuca, 1985).
Acknowledgement. This work was supported by Grant DE-06007 from the U.S. Public Health Service and Contract DEAC0276EY00119 from the U S. Department of Energy
References Benzle D
Boyne A D Dalgarno AC Duckworth J.. HIII R and Walker D M Studies of the skeleton of the sheep I The effect of different levels of dietary caiclum during pregnancy and lactation on Individual
bones J. Agnc. So 46 425444, 1955. Brommage R and DeLuca H F Regulatron of bone mrneral loss during lactation Am J. Physioi 246 E182-187, 1985 Canalrs E M Hlntr R L Dretrich J W Malna D M and Rarsz L G Effect of
207
somatomedin
and growth hormone on bone collagen synthesis in vitro
Metaboiism 26 1079- 1087, 1977 Daughaday W H and Kapadra M Maintenance trvity rn hypophysectomlzed
pregnant
of serum somatomedln
rats Endocrinology
ac
102 1317
1320. 1978 Ellrnger G M Duckworth J Dalgarno A C and Ouenoullle M H Skeletal changes durrng pregnancy and lactatron in the rat Effect of drfferenl levels of dietary calcrum 81 J Nulr 6 235-253 1952 Frost H M Bone histomorphometry Analysis of the labeling “escape error In Bone H/stomorphometry Techniques and interpretation R Reeker, ed CRC Press, Boca Raton. FL. 1983, pp 133- 142 Goss H and Schmrdt C L A Calcum and phosphorus metabolrsm in rats durrng pregnancy and lactation and the rnfluence of the reaction of the diet thereon J. Biol Chem 66 417-432. 1930 Halloran B P and DeLuca H F Skeletal changes during pregnancy and lactatron The role of vrtamin D Endocnnoiogy 107 1923- 1928. 1980 Heaney R P and Skrllman T G Calcrum metabolrsm in normal human pregnancy J Clan Endocnnol 33 661-670. 1971 Hurley T S D’Ercole A J Handwerger S Underwood L E Furlanetto R W and Fellows R E Ovrne placental lactogen induces somatomedrn A possible role rn fetal growth Endocnnoiogy 101 1635-1638. 1977 Komarkova A, Zahor Z and Crabanova V The effect of lactation on the composrtron of long bone in rats J Lab Ci~n Med 69 102-109. 1967 Mrller S C Omura T H and Smrth L J Changes in dentrn appositional rates durrng pregnancy and lactation In rats J Dent Res 64 10621064, 1985 Mrller S C Halloran B P DeLuca H F and Jee W S S Role of vrtamrn D rn maternal skeletal changes durrng pregnancy and lactatron a hrslomorphometnc study Caini Tissue Inf 34 245-252. 1982 Prtkrn R M Calcium metabolrsm rn pregnancy A revrew Am. J Obsref Gynecoi 121 724-737, 1975 Redd E H Miller SC and Jee W S S Changes In endochondral bone elongatron rates durrng pregnancy and lactation in rats Calcif Issue Inf. 36 697-701, 1984 Robertson M C and Frresen H G Two forms of rat placental lactogen revealed by radroimmunoassay Endocnnology 106 238882390. 1981 Smrth R W Jr Dietary and hormonal factors rn bone loss fed Proc 26 1737m 1746, 1967 Spray C M A study of some aspects of reproduction by means of chemical analysis Er J Nuti 4:354-360, 1950 Tonkowrcz P A and Voogt J L Examlnatron of rat placental lactogen and prolactrn at 6-hr Intervals dunng mrdpregnancy Proc Sot Exp B/o/ Med 173 583-587. 1983 Wong K M , Singer L Ophaug R H and Klern L Effect of lactation and calcrum deficiency, and of fluoride Intake, on bone turnover rn rats Isotopic measurements of bone resorption and formatton J Nutr 111 184881854, 1981
Recewed May 28. 1985 Rewsed February 1 I. 1986 Accepted February 13, 1986