- Email: [email protected]
The changing glycemic response to exercise during pregnancy
The changing glycemic response to exercise during pregnancy James F. Clapp III, MD, and Eleanor L. Capeless, MD Cleveland, Ohio, and Burlington, Vermont This study was designed to test the hypothesis that pregnancy reverses the nonpregnant hyperglycemic response to sustained exercise. Serial data were obtained from 75 exercising women. Before pregnancy, exercise produced an intensity-dependent increase in blood glucose that averaged 1.5 mmol/L at high intensities. By the eighth week this response was blunted and blood glucose increased only when exercise intensity exceeded 80% of maximum. At 15 weeks this progressed and was not associated with a change in either the insulin or catecholamine response. By the twenty-third week exercise produced a decrease in blood glucose that was no longer related to exercise intensity. We conclude that the hypothesis is correct and speculate that the early change in the response is related to decreased hepatic glucose release coupled with increased glucose oxidation. In late pregnancy this is probably accentuated by fetoplacental demands. (AM J OBSTET GVNECOL 1991 ;165:1678-83.)
Key words: Pregnancy, exercise, glucose
The change in blood glucose level that occurs during exercise represents a shift in the balance between splanchnic glucose production and peripheral glucose uptake. I When production exceeds uptake, blood glucose rises, and when uptake exceeds production, it falls. In exercising men the balance between production and uptake is variable and appears to be dependent on multiple factors such as food intake, catecholamine response, physical condition, and the type, intensity, and duration of the exercise. I - 5 Although detailed data are not available in women, it is probable that the relationships and mechanisms are similar. During pregnancy the amount of maternal glucose available for fetoplacental uptake varies directly with maternal levels, and in large animal models maternal hypoglycemia is associated with fetal growth restriction. 6 . 7 Thus a maternal hypoglycemic response to regular recreational exercise during pregnancy potentially could restrict fetal glucose availability and result in some degree of fetal growth restriction. This potential concern has been reinforced by several findings. First, data obtained in a limited number of From the Department of Reproductive Biology, Case Western Reserve University School of Medicine, MetroHealth Medical Center, and the Department of Obstetrics and Gynecology, University of Vermont College of Medicine. Supported in part by National Institutes of Health grants Nos. R01 21268 andP50 211089 and funds from MetroHealth Medical Center. Presented in part at the Thirty-eighth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 20-23, 1991. Reprint requests: James F. Clapp III, MD, Department of Reproductive Biology, Case Western Reserve University School of Medicine, MetroHealth Medical Center, 3395 Scranton Road, Cleveland, OH 44109. 6/6/33016
1678
subjects indicate that the consistent increase in blood glucose levels seen in response to recreational running in fit, nonpregnant women was initially blunted and then reversed after similar exercise sessions during pregnancy.7.S Second, recreational athletes who continued a regular running or aerobics regimen in late pregnancy were delivered at term of infants with morphometric evidence of mild asymmetric growth restriction. 9 The current study was undertaken to explore the relationship between pregnancy and the glycemic response to exercise in greater detail. It was designed to test the hypothesis that when fit female recreational athletes continue to exercise during pregnancy the preconceptional hyperglycemic response to field exercise is reversed. The findings support the hypothesis and suggest that the pattern of food intake before, during, and after exercise may need to be altered in the recreational athlete during pregnancy.
Methods After obtaining informed consent in accord with institutional guidelines, 75 healthy, fit, female recreational athletes who either ran (n = 40) or performed aerobics (n = 35) three or more times each week were enrolled for prospective study. Each subject was studied serially before conception and every 6 to 8 weeks throughout pregnancy. All subjects maintained their regular weekly exercise regimens above 60% of their preconceptional performance levels throughout pregnancy. All pregnancies were singleton, accurately dated, and clinically normal. Total pregnancy weight gain (mean 12.7 kg; range 7.3 to 17.3 kg) was adequate and agreed with the reported approximate caloric intake at the various time points of study (mean 40
Volume 165 Number 6, Part I
kcal/kg; range 30 to 55 kcal/kg). The caloric mix was typical for an exercise populace (high complex carbohydrate, low fat, adequate protein). To reproduce the conditions of each individual's exercise sessions outside the laboratory, all subjects were encouraged to eat their usual diet and were studied at a time relative to food intake that coincided with their regular exercise regimen (90 to 270 minutes after food intake). In 70 of the 75 subjects the study protocol began with 10 minutes of rest followed by 20 minutes of representative exercise. A free-flowing forearm venous blood sample was obtained at the end of the rest and exercise period for determination of glucose, insulin, and norepinephrine content. The exercise consisted of either a treadmill run or a standardized aerobics regimen and was conducted at a representative exercise intensity for each individual at each given time point in pregnancy. Daily exercise intensity was monitored with a portable heart rate monitor,1O and the average exercise heart rate for the preceding week was maintained during each laboratory exercise session. Exercise intensity was determined by monitoring oxygen consumption for the latter half of the exercise session and was expressed as a percentage of each individual's preconceptional maximum capacity.9 In the remaining five subjects an indwelling forearm venous sampling catheter was placed before study and free-flowing blood samples were obtained every 15 minutes throughout the protocol. The protocol began with 30 minutes of standing rest followed by a 45-minute, intensity-controlled, uphill treadmill walk and ended with 30 minutes of sitting rest. On each test occasion, the exercise intensity, determined by oxygen consumption, was maintained at a level equivalent to 50% of each individual's preconceptional maximumcapacity. Glucose levels were determined in duplicate on whole blood with a glucose oxidase method. II Serum insulin was measured with a double antibody radioimmunoassay, modified from that described by Starr et al.,12 with interassay and intraassay coefficients of variation of <10% and <5%, respectively. Blood samples for norepinephrine were collected in chilled tubes containing ethylene glycol-bis(f3-amino-ethyl ether)N,N,N',N'-tetraacetic acid and glutathione and stored at - 70° C until analysis by high-performance liquid chromatography with a modification of the method of Kontur and Dawson l 3 (sensitivity 15 pg/ml; coefficient of variation <5%). The data were initially grouped by the intensity of the exercise and analyzed for significant change from rest and from preconceptional values at each study point, by means of analysis of variance and Duncan's multiple range test. Linear regression was used to detect relationships between exercise intensity and the
Glycemic response to exercise in pregnancy
1679
change in glucose. Significant changes over time were sought with analysis of variance and Duncan's multiple range test. Significance was set at the p < 0.05 level.
Results The effect of 20 minutes of running or aerobics on blood glucose levels under conditions similar to those experienced during daily exercise in the field are shown in Table I. The number of subjects exercising within a given intensity range is not shown, as it changed from one time point to another. Before conception 16 exercised below 60% of their maximum and 33 exercised above 69% of their maximum , with an overall average exercise intensity of 68% of maximum. At 8 weeks 25 exercised below 60% and 25 above 69% of their maximum, with the average overall value being 64% of maximum capacity. At 15 and 23 weeks the size of the low-intensity group was stable at 25, but only 18 exercised above 69% of maximum with mean intensities of 63% and 61 % of maximum, respectively. After the thirtieth week, 29 exercised at <60% of maximum while 18 continued to exercise above 69% of maximum, and the overall mean intensity was maintained at 62% of maximum capacity. Before pregnancy postexercise blood glucose levels exceeded preexercise values 96% of the time and the increase was significant at all exercise intensities. The average postexercise elevation varied directly with exercise intensity over a blood glucose range of 3.5 to 7.95 mmol/L at exercise intensities between 42% and 87% of maximum (change in glucose = - 2.02 mmol/L + 0.042 x percent maximum, r = 0.5425). This predicts that blood glucose will rise at exercise intensities in excess of 48% of maximum capacity at a rate of 0.44 mmol!L for each 10% increase in exercise intensity above that level. At <60% of maximum intensity the increase averaged 0.32 ± 0.10 mmol! L (mean ± SEM) rising to 1.54 ± 0.29 mmollL at intensities >80% of maximum. By the eighth week of gestation the response changed, with postexercise glucose levels exceeding preexercise values <40% of the time. A significant increase postexercise occurred only when intensity exceeded 80% of maximum capacity, with the average increase in blood glucose falling by 33% to 1 mmol / L. There was a significant downward shift in both the slope and the intercept of the relationship between exercise intensity and the change in blood glucose (change in glucose = - 2.04 mmol!L + 0.034 x percent maximum, r = 0.4868). This predicts that blood glucose will rise only at exercise intensities >63% of maximum capacity at a rate of 0.34 mmol/L for each 10% increase in intensity above that level. This change suggests that a shift in the relationship between hepatic glucose pro-
1680 Clapp and Capeless
December 199 1 Am J Obstet Gynecol
Table I. Blood glucose levels before and after exercise Postexercise level by exercise intensity (mmollL) Time
Before pregnancy 8 wk 15 wk 23 wk ~30 wk
Preexercise level at rest (mmollL)
4.72 4.60 4.56 4.41 4.66
±
<60%
0.43
5.04 4.15 3.96 3.88 3.96
± 0.44
0.46 0.44 ± 0.55
± ±
±
0.92*
±
0.55*
±
0.56*
60%-69%
1
Maximum
Maximum
5.17 4.59 4.29 3.96 4. 14
± 0.61
± 0.58*
±
0.68*
± ± ±
0.50 0.57* 0.45*
70%-79 %
I
I
Maximum
5.95 4.87 4.32 4.17 4.16
± 0.62
±
0.95*
± ± ±
0.32 0.55 0.47*
± 0.64
>80%
Maximum
6.27 5.60 4.83 4.12 4.41
±
0.98*
± 0.77*
0.47 0.54 ± 0.40
± ±
Data presented as mean ± SD. *Significantly different from rest (P < 0.05).
Table II. Insulin and norepinephrine levels before and after exercise Norepinephrine (pglml)
Insulin (ILL' l ml) Time
Before pregnancy 8 wk 15 wk 23 wk ~30wk
Preexet'cise
15.9 12.4 15.3 15.6 31.6
± ± ± ± ±
8.8 9.6 11.2 9.0 25.1 *
I
Postexercise
6.3 6.1 6.1 6.2 13.7
4.6 3. 1 ± 3.2 ± 3.9 ± 10.3* ± ±
Preexercise
508 530 573 566 494
± ± ± ± ±
166 189 221 177 171
I
Postexercise
2180 1545 1553 1203 1152
± 862 ± 487* ± ± ±
558* 429* 737*
Data presented as mean ± SD. *Significantly different from value before pregnancy, p < 0.05. duction and peripheral use during exercise begins early in pregnancy. In the fifteenth week the change progressed with postexercise glucose levels exceeding preexercise levels <25% of the time. Postexercise values were significantly lower (-0.62 ± 0.12 mmoilL) than those observed preexercise at intensities <60% of maximum capacity and were unchanged at all higher intensities. The regression equation (change in glucose = - 2.04 mmoilL + 0.027 x percent maximum, r = 0.4494) predicts that exercise in the fifteenth week of gestation will increase blood glucose only at intensities >76% of maximum capacity at a rate of only 0.27 mmoilL for each 10% increase in intensity above that level. By the twenty-third week, postexercise glucose levels rose < 20% of the time and a significant decrease in glucose levels of similar magnitude (- 0.60 ± 0.15 mmoilL) occurred after exercise at all exercise intensities <70% of maximum capacity. Regression of exercise intensity versus the change in glucose level was no longer significant (r = 0.1988) with a slope of 0.0 1O. After the thirtieth week, postexercise glucose levels rose < 10% of the time, a significant decrease in glucose levels occurred after exercise at all exercise intensities <80% of maximum capacity, and the relationship between exercise intensity and the change in blood glucose remained nonsignificant. In the subjects who maintained their exercise intensity at <60% of maximum capacity, which approximates the recommendations of the American College
of Obstetricians and Gynecologists,14 exercise was associated with a consistent, significant decrease in blood glucose from the fifteenth week onward. In the twentythird, thirtieth, and thirty-seventh week >20% of the individuals had postexercise values ::;3.3 mmoilL with mean postexercise levels approximating 3.9 mmol/L. Table II details the serial changes observed in the insulin and norepinephrine levels after 20 minutes of representative exercise during pregnancy. Unfortunately, between 10 (14%) and 16 (23%) of the sample sets were partially incomplete at each time point, but their distribution over the various intensity ranges was random. Serum insulin levels decreased significantly after exercise at all time points, and the magnitude was unchanged by increasing exercise intensity. Both preexercise and postexercise levels remained at or near those observed before conception through the twentythird week. After the thirtieth week, both the preexercise and postexercise values rose significantly above the concentrations observed before pregnancy, reflecting the development of relative insulin resistance in all subjects in late pregnancy. Preexercise norepinephrine levels did not change significantly with advancing gestation. However, the mean postexercise level fell significantly in the eighth week and remained at that level for the remainder of the pregnancy. This appeared to be due to the 4% to 7% decrease in mean exercise intensity during pregnancy. When the raw data were corrected for exercise intensity, a significant decrease fro m preconceptional levels was seen only after the
Glycemic response to exercise in pregnancy
Volume 165 Number 6, Part 1
1681
Table III. Respiratory exchange ratio during exercise Exercise intensity during measurement Time
Before pregnancy 8 wk 15 wk 23 wk 2:30 wk
60%-69%
<60%
0,910 0.953 0.944 0.939 0.942
± ± ± ± ±
70%-79%
Maximum
Maximum
0,929 0.946 0.983 0.960 0.957
0,039 0.038* 0.046* 0.038* 0.031 *
± ± ± ± ±
>80%
Maximum
Maximum
0,038 0.042* 0.036* 0.021* 0.028*
0.945 0.974 0.976 0.965 0.952
0.955 0.990 0.968 0.980 0.947
± 0.037 ± ± ± ±
0.018* 0.022* 0.037 0.019
± 0.041
± 0.015* ± ± ±
0.033 0.041 0.040
Data presented as mean ± SD. *Significantly different from value before pregnancy, p < 0.05.
Table IV. Insulin level and respiratory exchange ratio before and during exercise at 50% of maximum capacity Insulin (/LV ImZ) Time
Rest
Before pregnancy
10.2 ± 3.9
8 wk
15 wk 23 wk 30 wk
9.9 ± 4.0 10.4 ± 3.6 10.9 ± 3.8 26.7 ± 10.2
I
Respiratory exchange ratio Exercise
Rest
3.6 ± 1.3 3.4 ± 1.1
2.6 ± 1.6 2.9 ± 1.0 6.3 ± 2.5*
0.834 0.850 0.834 0.864 0.871
± ± ± ± ±
0.044 0.029 0.081 0.059 0.054*
I
Exercise
0.866 0.934 0.912 0.924 0.921
± 0.050
0.025* 0.024* 0.022* ± 0.026* ±
± ±
Data presented as mean ± SD. *Significantly different from value before pregnancy, p < 0.05. thirtieth week at exercise intensities >69% of maximum capacity. Table III details the serial changes observed in the respiratory exchange ratio during exercise. It increased significantly by the eighth week at all exercise intensities, and the elevation was maintained at all intensities through the fifteenth week. Thereafter, changes in the respiratory exchange ratio became nonsignificant at the higher intensities, but a significant increase in the ratio was maintained at exercise intensities <69% of maximum capacity. With Lusk's calculation,lS the pregnancy-associated increase in the ratio approximates a 5% to 7% increase in the fractional oxidation of carbohydrate during exercise in pregnancy. The serial changes in the five subjects who exercised at a constant intensity (50% max) throughout pregnancy were similar to those observed in the larger group. Before pregnancy, a small but significant increase in blood glucose (+ 0.35 ± 0.06 mmollL) was observed by the thirtieth minute of exercise, with a gradual return to preexercise levels during recovery. In the eighth week this did not occur and by the fifteenth week the response was reversed with a significant decrease in blood glucose (- 0.62 ± 0.08 mmollL) by the fifteenth minute of exercise that persisted (- 0.46 to - 0.64 mmollL) for the remainder of the exercise and recovery period (60 minutes). A similar pattern was observed at 23, 30, and 37 weeks, with average decreases in blood glucose level during exercise and
recovery of - 0.63, - 0.56, and - 0.61 mmoliL at the three respective time points of study. Norepinephrine levels at rest, during exercise, and at recovery were similar to those observed before conception until the thirtieth week. Thereafter, levels at rest and during exercise were significantly decreased (rest, 727 to 457 and 529 pg/ml; exercise, 1240 to 848 and 782 pg/ml), but the absolute change from rest to exercise was significantly decreased only at 37 weeks (change in norepinephrine, 513 to 271 pg/ml). Table IV details the serial changes in insulin and the respiratory exchange ratio during exercise in these five subjects. Again there was no significant change in insulin levels until the thirtieth week, when the levels both at rest and during exercise increased. The respiratory exchange ratio at standing rest increased significantly after the thirtieth week, but the values during exercise were increased significantly above preconceptional values throughout pregnancy (+ 0.056 ± 0.006 units). Comment
These serial data, gathered under conditions designed to mimic an individual's daily habits (dietary, temporal, exercise type, and intensity), demonstrate that the nonpregnant hyperglycemic response to recreational running and aerobics is initially blunted and then reversed early in the course of pregnancy. Because weight gain, caloric intake, and caloric mix were relatively constant and adequate, this suggests that some of
1682 Clapp and Capeless
the metabolic adaptations induced by pregnancy alter the balance between hepatic glucose production and peripheral use, which becomes apparent only when peripheral use is increased. A similar hypoglycemic shift, the so-called accelerated starvation of pregnancy,'6 is another example of this phenomenon. While the underlying mechanism for this shift during pregnancy is unclear, the increase in the respiratory exchange ratio during exercise in early pregnancy suggests that this may be due in part to a pregnancy-associated increase in the fractional use of carbohydrate by muscle during exercise. The superimposition of the constant and ever-increasing glucose use by the growing placenta and fetus also must increase the demand for glucose during exercise in late pregnancy.6. 7. '7 This additional demand is reflected in the increase in the respiratory exchange ratio at rest observed after the thirtieth week. The combination of increased glucose use by both exercising muscle and the fetoplacental unit probably is responsible for both the increase in the magnitude and the consistency of the hypoglycemic response during exercise in late pregnancy. In addition, the fact that there is no change in either the insulin or catecholamine response, coupled with the intensity dependence of the glycemic response reversal early in gestation, suggests that during exercise there is also a pregnancy-associated decrease in hepatic glucose production that is not apparent at rest.'7 This physiologic change also should alter glucose homeostasis, favoring a fall in blood glucose levels during exercise. A definitive answer to both possibilities (a pregnancy-associated increase in fractional glucose use by exercising muscle and a decrease in hepatic glucose production during exercise) will require further serial studies with the use of stable isotope infusion. The fact that the pregnancy-associated increase in the respiratory exchange ratio during exercise was consistently abolished at high exercise intensities with advancing gestation is an apparent contradiction and requires explanation. Unfortunately we do not have all the data necessary to identify the underlying mechanism. However, the consistency of the data suggests that it is real and not an interpretive error. The possibility. that carbon dioxide retention occurred during highintensity exercise in late pregnancy (which would factitiously lower the ratio) is unlikely because minute ventilation progressively increased and expired carbon dioxide concentration decreased during pregnancy. Thus we are left with the speculation that muscle glycogen stores at rest are significantly reduced in late pregnancy (possibly because of the increasing requirements of the placenta and fetus) with a resultant depletion of muscle glycogen earlier during high-intensity
December 1991 Am J Obstet Gynecol
exercise. When combined with the presumed suppression of hepatic glucose production and fall in blood glucose levels, this change could easily limit glucose availability to exercising muscle earlier in the latter half of pregnancy. A definitive answer would require serial muscle biopsy in a limited number of subjects during protracted high-intensity exercise. In terms of fetoplacental substrate availability, the data indicate that relatively brief, low- and moderateintensity recreational exercise rapidly lowers maternal blood glucose levels by 0.5 to 0.7 mmol/L on average and that the decrease persists for as much as 30 minutes after exercise. Whether an intermittent change of this degree is sufficient to significantly restrict fetal substrate supply and ultimately growth rate is currently unclear. However, the observation that this degree of running and aerobics in late pregnancy was associated with morphometric evidence of asymmetric growth restriction suggests that, coupled with other physiologic thanges (such as a probable reduction in placental bed blood flow), it may have an appreciable effect that potentially has therapeutic value. 7 . '0 For example, >70% of the exercise-associated decrease in birth weight was due to a 220 gm reduction in body fat, suggesting that reduced fat cell number could have prophylactic value. Likewise, regular, low- to moderate-intensity exercise in late pregnancy has been shown to be as effective as insulin in maintaining euglycemia in a limited number of subjects. '8. '9 Given a larger experience, it is likely that it will also be shown to reduce birth weight, which should reduce morbidity in the large-for-gestational age infant. Finally, it is unfortunate that there are no background data addressing the factors that influence glucose met:abolism during exercise in nonpregnant women. This study suggests that fit women who exercise regularly may have a different response from men ,2.5 because the response to sustained but shortduration exercise is consistently a hyperglycemic one whose magnitude is intensity dependent. In summary, the preconceptional hyperglycemic response to running and aerobics, seen in female recreational athletes, is reversed during pregnancy. The reversal is initially intensity dependent and does not appear to be related to changes in either insulin or norepinephrine response. The data suggest that it may reflect a pregnancy-associated decrease in hepatic glucose production and a pregnancy-associated increase in fractional glucose use by muscle during exercise. In late pregnancy the increased peripheral use of glucose is further increased by the demands of the fetoplacental unit, which magnify the change in blood glucose response and probably obscure its intensity dependence.
Glycemic response to exercise in pregnancy
Volume 165 Number 6, Part 1
The relative hypoglycemia associated with exercise persists for an as yet undefined period of time after exercise and may influence fetal growth,
REFERENCES 1, Coggan AR. Plasma glucose metabolism during exercise in humans. Sports Med 1991;11:102-24. 2. Montain SJ, Hopper MK, Coggan AR, Coyle EF. Exercise metabolism at different time intervals after a meal. J Appl PhysioI1991;70:882-8. 3. Cooper DM, Barsrow TJ, Bergner A, Lee WNP. Blood glucose turnover during high and low-intensity exercise. Am J Physiol 1989;257:E405-12. 4. Ahlborg G, Felig P, Hagenfeldt L, Hendler R, Wahren J. Substrate turnover during prolonged exercise in man. Splanchnic and leg metabolism of glucose, free fatty acids, and amino acids. J Clin Invest 1974;53:1080-90. 5. Ahlborg G, Felig P. Lactate and glucose exchange across the forearm, legs and splanchnic bed during and after prolonged exercise. J Clin Invest 1982;69:45-54. 6. DiGiacomo JE, Hay WW J r. Fetal glucose metabolism and oxygen consumption during sustained hypoglycemia. Metabolism 1990;39: 193-202. 7. Clapp JF. Utero-placental blood flow and fetal growth. In: Sharp F, Milner RDG, Fraser RB, eds. Fetal growth. Ashton-under-Lyne, England: Royal College of Obstetricians and Gynaecologists, Peacock Press, 1989:235-44. 8. Hauth JC, Gilstrap LC, Widner K. Fetal heart rate reactivity before and after maternal jogging during the third trimester. AMJ OBSTET GYNECOL 1982;142:545-7. 9. Clapp JF, Capeless EL. Neonatal morphometrics after endurance exercise during pregnancy. AM J OBSTET GyNECOL 1990; 163: 1805-11.
1683
10. Seaward BL, Sleamaker RH, McAuliffe T, Clapp JF. The precision and accuracy of a portable heart rate monitor. Biomed Instrum Technol 1990;24:37-41. 11. Clapp JF, Wesley M, Sleamaker RH. Thermoregulatory and metabolic responses to jogging prior to and during pregnancy. Med Sci Sports Exerc 1987;19:124-30. 12. Starr JI, Horwitz DL, Mako ME. Insulin, proinsulin and c-peptide. In: Jaffe BM, Berman HR, eds. Methods of hormone radioimmunoassay. New York: Academic Press, 1979:613-28. 13. Kontur P, Dawson R. Manipulation of mobile phase parameters for HPLC separation of endogenous monoamines in rat brain tissue. J Neurosci Methods 1984; 11 :518. 14. American College of Obstetricians and Gynecologists. Home exercise programs-exercise during pregnancy and the postnatal period. Washington: American College of Obstetricians and Gynecologists, 1985. 15. Lusk G. Science of nutrition. 4th ed. Philadelphia: WB Saunders, 1928. 16. Metzger BE, Ravnikar V, Vileisis RA, Freinkel N. "Accelerated starvation" and the skipped breakfast in late normal pregnancy. Lancet 1982; 1:588-92. 17. Kalhan SC, D'Angelo LJ, Savin SM, Adam PAJ. Glucose production in pregnant women at term gestation. Sources of glucose for the human fetus. J Clin Invest 1979;63:38894. 18. Bung P, Artal R, Khodiginan N, Kjos S. Exercise in gestational diabetes: an optional therapeutic approach. Diabetes 1991; [In press]. 19. Jovanovic-Peterson L, Durak E, Peterson CM. Randomized trial of diet versus diet plus cardiovascular conditioning on glucose levels in gestational diabetes. AM J OBSTET GYNECOL 1990;162:754-6.
Bound volumes available to subscribers Bound vol~mes of the AMERICAN JOURNAL OF OBSTETRICS AND GYNECOLOGY are available to subscribers (only) for the 1991 issues from the Publisher, at a cost of $61.00 for domestic, $88.27 for Canada, and $84.00 for international for Vol. 164 (January-June) and Vol. 165 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800) 325-4177, ext. 4351, or (314) 453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.
Key words: Pregnancy, exercise, glucose
The change in blood glucose level that occurs during exercise represents a shift in the balance between splanchnic glucose production and peripheral glucose uptake. I When production exceeds uptake, blood glucose rises, and when uptake exceeds production, it falls. In exercising men the balance between production and uptake is variable and appears to be dependent on multiple factors such as food intake, catecholamine response, physical condition, and the type, intensity, and duration of the exercise. I - 5 Although detailed data are not available in women, it is probable that the relationships and mechanisms are similar. During pregnancy the amount of maternal glucose available for fetoplacental uptake varies directly with maternal levels, and in large animal models maternal hypoglycemia is associated with fetal growth restriction. 6 . 7 Thus a maternal hypoglycemic response to regular recreational exercise during pregnancy potentially could restrict fetal glucose availability and result in some degree of fetal growth restriction. This potential concern has been reinforced by several findings. First, data obtained in a limited number of From the Department of Reproductive Biology, Case Western Reserve University School of Medicine, MetroHealth Medical Center, and the Department of Obstetrics and Gynecology, University of Vermont College of Medicine. Supported in part by National Institutes of Health grants Nos. R01 21268 andP50 211089 and funds from MetroHealth Medical Center. Presented in part at the Thirty-eighth Annual Meeting of the Society for Gynecologic Investigation, San Antonio, Texas, March 20-23, 1991. Reprint requests: James F. Clapp III, MD, Department of Reproductive Biology, Case Western Reserve University School of Medicine, MetroHealth Medical Center, 3395 Scranton Road, Cleveland, OH 44109. 6/6/33016
1678
subjects indicate that the consistent increase in blood glucose levels seen in response to recreational running in fit, nonpregnant women was initially blunted and then reversed after similar exercise sessions during pregnancy.7.S Second, recreational athletes who continued a regular running or aerobics regimen in late pregnancy were delivered at term of infants with morphometric evidence of mild asymmetric growth restriction. 9 The current study was undertaken to explore the relationship between pregnancy and the glycemic response to exercise in greater detail. It was designed to test the hypothesis that when fit female recreational athletes continue to exercise during pregnancy the preconceptional hyperglycemic response to field exercise is reversed. The findings support the hypothesis and suggest that the pattern of food intake before, during, and after exercise may need to be altered in the recreational athlete during pregnancy.
Methods After obtaining informed consent in accord with institutional guidelines, 75 healthy, fit, female recreational athletes who either ran (n = 40) or performed aerobics (n = 35) three or more times each week were enrolled for prospective study. Each subject was studied serially before conception and every 6 to 8 weeks throughout pregnancy. All subjects maintained their regular weekly exercise regimens above 60% of their preconceptional performance levels throughout pregnancy. All pregnancies were singleton, accurately dated, and clinically normal. Total pregnancy weight gain (mean 12.7 kg; range 7.3 to 17.3 kg) was adequate and agreed with the reported approximate caloric intake at the various time points of study (mean 40
Volume 165 Number 6, Part I
kcal/kg; range 30 to 55 kcal/kg). The caloric mix was typical for an exercise populace (high complex carbohydrate, low fat, adequate protein). To reproduce the conditions of each individual's exercise sessions outside the laboratory, all subjects were encouraged to eat their usual diet and were studied at a time relative to food intake that coincided with their regular exercise regimen (90 to 270 minutes after food intake). In 70 of the 75 subjects the study protocol began with 10 minutes of rest followed by 20 minutes of representative exercise. A free-flowing forearm venous blood sample was obtained at the end of the rest and exercise period for determination of glucose, insulin, and norepinephrine content. The exercise consisted of either a treadmill run or a standardized aerobics regimen and was conducted at a representative exercise intensity for each individual at each given time point in pregnancy. Daily exercise intensity was monitored with a portable heart rate monitor,1O and the average exercise heart rate for the preceding week was maintained during each laboratory exercise session. Exercise intensity was determined by monitoring oxygen consumption for the latter half of the exercise session and was expressed as a percentage of each individual's preconceptional maximum capacity.9 In the remaining five subjects an indwelling forearm venous sampling catheter was placed before study and free-flowing blood samples were obtained every 15 minutes throughout the protocol. The protocol began with 30 minutes of standing rest followed by a 45-minute, intensity-controlled, uphill treadmill walk and ended with 30 minutes of sitting rest. On each test occasion, the exercise intensity, determined by oxygen consumption, was maintained at a level equivalent to 50% of each individual's preconceptional maximumcapacity. Glucose levels were determined in duplicate on whole blood with a glucose oxidase method. II Serum insulin was measured with a double antibody radioimmunoassay, modified from that described by Starr et al.,12 with interassay and intraassay coefficients of variation of <10% and <5%, respectively. Blood samples for norepinephrine were collected in chilled tubes containing ethylene glycol-bis(f3-amino-ethyl ether)N,N,N',N'-tetraacetic acid and glutathione and stored at - 70° C until analysis by high-performance liquid chromatography with a modification of the method of Kontur and Dawson l 3 (sensitivity 15 pg/ml; coefficient of variation <5%). The data were initially grouped by the intensity of the exercise and analyzed for significant change from rest and from preconceptional values at each study point, by means of analysis of variance and Duncan's multiple range test. Linear regression was used to detect relationships between exercise intensity and the
Glycemic response to exercise in pregnancy
1679
change in glucose. Significant changes over time were sought with analysis of variance and Duncan's multiple range test. Significance was set at the p < 0.05 level.
Results The effect of 20 minutes of running or aerobics on blood glucose levels under conditions similar to those experienced during daily exercise in the field are shown in Table I. The number of subjects exercising within a given intensity range is not shown, as it changed from one time point to another. Before conception 16 exercised below 60% of their maximum and 33 exercised above 69% of their maximum , with an overall average exercise intensity of 68% of maximum. At 8 weeks 25 exercised below 60% and 25 above 69% of their maximum, with the average overall value being 64% of maximum capacity. At 15 and 23 weeks the size of the low-intensity group was stable at 25, but only 18 exercised above 69% of maximum with mean intensities of 63% and 61 % of maximum, respectively. After the thirtieth week, 29 exercised at <60% of maximum while 18 continued to exercise above 69% of maximum, and the overall mean intensity was maintained at 62% of maximum capacity. Before pregnancy postexercise blood glucose levels exceeded preexercise values 96% of the time and the increase was significant at all exercise intensities. The average postexercise elevation varied directly with exercise intensity over a blood glucose range of 3.5 to 7.95 mmol/L at exercise intensities between 42% and 87% of maximum (change in glucose = - 2.02 mmol/L + 0.042 x percent maximum, r = 0.5425). This predicts that blood glucose will rise at exercise intensities in excess of 48% of maximum capacity at a rate of 0.44 mmol!L for each 10% increase in exercise intensity above that level. At <60% of maximum intensity the increase averaged 0.32 ± 0.10 mmol! L (mean ± SEM) rising to 1.54 ± 0.29 mmollL at intensities >80% of maximum. By the eighth week of gestation the response changed, with postexercise glucose levels exceeding preexercise values <40% of the time. A significant increase postexercise occurred only when intensity exceeded 80% of maximum capacity, with the average increase in blood glucose falling by 33% to 1 mmol / L. There was a significant downward shift in both the slope and the intercept of the relationship between exercise intensity and the change in blood glucose (change in glucose = - 2.04 mmol!L + 0.034 x percent maximum, r = 0.4868). This predicts that blood glucose will rise only at exercise intensities >63% of maximum capacity at a rate of 0.34 mmol/L for each 10% increase in intensity above that level. This change suggests that a shift in the relationship between hepatic glucose pro-
1680 Clapp and Capeless
December 199 1 Am J Obstet Gynecol
Table I. Blood glucose levels before and after exercise Postexercise level by exercise intensity (mmollL) Time
Before pregnancy 8 wk 15 wk 23 wk ~30 wk
Preexercise level at rest (mmollL)
4.72 4.60 4.56 4.41 4.66
±
<60%
0.43
5.04 4.15 3.96 3.88 3.96
± 0.44
0.46 0.44 ± 0.55
± ±
±
0.92*
±
0.55*
±
0.56*
60%-69%
1
Maximum
Maximum
5.17 4.59 4.29 3.96 4. 14
± 0.61
± 0.58*
±
0.68*
± ± ±
0.50 0.57* 0.45*
70%-79 %
I
I
Maximum
5.95 4.87 4.32 4.17 4.16
± 0.62
±
0.95*
± ± ±
0.32 0.55 0.47*
± 0.64
>80%
Maximum
6.27 5.60 4.83 4.12 4.41
±
0.98*
± 0.77*
0.47 0.54 ± 0.40
± ±
Data presented as mean ± SD. *Significantly different from rest (P < 0.05).
Table II. Insulin and norepinephrine levels before and after exercise Norepinephrine (pglml)
Insulin (ILL' l ml) Time
Before pregnancy 8 wk 15 wk 23 wk ~30wk
Preexet'cise
15.9 12.4 15.3 15.6 31.6
± ± ± ± ±
8.8 9.6 11.2 9.0 25.1 *
I
Postexercise
6.3 6.1 6.1 6.2 13.7
4.6 3. 1 ± 3.2 ± 3.9 ± 10.3* ± ±
Preexercise
508 530 573 566 494
± ± ± ± ±
166 189 221 177 171
I
Postexercise
2180 1545 1553 1203 1152
± 862 ± 487* ± ± ±
558* 429* 737*
Data presented as mean ± SD. *Significantly different from value before pregnancy, p < 0.05. duction and peripheral use during exercise begins early in pregnancy. In the fifteenth week the change progressed with postexercise glucose levels exceeding preexercise levels <25% of the time. Postexercise values were significantly lower (-0.62 ± 0.12 mmoilL) than those observed preexercise at intensities <60% of maximum capacity and were unchanged at all higher intensities. The regression equation (change in glucose = - 2.04 mmoilL + 0.027 x percent maximum, r = 0.4494) predicts that exercise in the fifteenth week of gestation will increase blood glucose only at intensities >76% of maximum capacity at a rate of only 0.27 mmoilL for each 10% increase in intensity above that level. By the twenty-third week, postexercise glucose levels rose < 20% of the time and a significant decrease in glucose levels of similar magnitude (- 0.60 ± 0.15 mmoilL) occurred after exercise at all exercise intensities <70% of maximum capacity. Regression of exercise intensity versus the change in glucose level was no longer significant (r = 0.1988) with a slope of 0.0 1O. After the thirtieth week, postexercise glucose levels rose < 10% of the time, a significant decrease in glucose levels occurred after exercise at all exercise intensities <80% of maximum capacity, and the relationship between exercise intensity and the change in blood glucose remained nonsignificant. In the subjects who maintained their exercise intensity at <60% of maximum capacity, which approximates the recommendations of the American College
of Obstetricians and Gynecologists,14 exercise was associated with a consistent, significant decrease in blood glucose from the fifteenth week onward. In the twentythird, thirtieth, and thirty-seventh week >20% of the individuals had postexercise values ::;3.3 mmoilL with mean postexercise levels approximating 3.9 mmol/L. Table II details the serial changes observed in the insulin and norepinephrine levels after 20 minutes of representative exercise during pregnancy. Unfortunately, between 10 (14%) and 16 (23%) of the sample sets were partially incomplete at each time point, but their distribution over the various intensity ranges was random. Serum insulin levels decreased significantly after exercise at all time points, and the magnitude was unchanged by increasing exercise intensity. Both preexercise and postexercise levels remained at or near those observed before conception through the twentythird week. After the thirtieth week, both the preexercise and postexercise values rose significantly above the concentrations observed before pregnancy, reflecting the development of relative insulin resistance in all subjects in late pregnancy. Preexercise norepinephrine levels did not change significantly with advancing gestation. However, the mean postexercise level fell significantly in the eighth week and remained at that level for the remainder of the pregnancy. This appeared to be due to the 4% to 7% decrease in mean exercise intensity during pregnancy. When the raw data were corrected for exercise intensity, a significant decrease fro m preconceptional levels was seen only after the
Glycemic response to exercise in pregnancy
Volume 165 Number 6, Part 1
1681
Table III. Respiratory exchange ratio during exercise Exercise intensity during measurement Time
Before pregnancy 8 wk 15 wk 23 wk 2:30 wk
60%-69%
<60%
0,910 0.953 0.944 0.939 0.942
± ± ± ± ±
70%-79%
Maximum
Maximum
0,929 0.946 0.983 0.960 0.957
0,039 0.038* 0.046* 0.038* 0.031 *
± ± ± ± ±
>80%
Maximum
Maximum
0,038 0.042* 0.036* 0.021* 0.028*
0.945 0.974 0.976 0.965 0.952
0.955 0.990 0.968 0.980 0.947
± 0.037 ± ± ± ±
0.018* 0.022* 0.037 0.019
± 0.041
± 0.015* ± ± ±
0.033 0.041 0.040
Data presented as mean ± SD. *Significantly different from value before pregnancy, p < 0.05.
Table IV. Insulin level and respiratory exchange ratio before and during exercise at 50% of maximum capacity Insulin (/LV ImZ) Time
Rest
Before pregnancy
10.2 ± 3.9
8 wk
15 wk 23 wk 30 wk
9.9 ± 4.0 10.4 ± 3.6 10.9 ± 3.8 26.7 ± 10.2
I
Respiratory exchange ratio Exercise
Rest
3.6 ± 1.3 3.4 ± 1.1
2.6 ± 1.6 2.9 ± 1.0 6.3 ± 2.5*
0.834 0.850 0.834 0.864 0.871
± ± ± ± ±
0.044 0.029 0.081 0.059 0.054*
I
Exercise
0.866 0.934 0.912 0.924 0.921
± 0.050
0.025* 0.024* 0.022* ± 0.026* ±
± ±
Data presented as mean ± SD. *Significantly different from value before pregnancy, p < 0.05. thirtieth week at exercise intensities >69% of maximum capacity. Table III details the serial changes observed in the respiratory exchange ratio during exercise. It increased significantly by the eighth week at all exercise intensities, and the elevation was maintained at all intensities through the fifteenth week. Thereafter, changes in the respiratory exchange ratio became nonsignificant at the higher intensities, but a significant increase in the ratio was maintained at exercise intensities <69% of maximum capacity. With Lusk's calculation,lS the pregnancy-associated increase in the ratio approximates a 5% to 7% increase in the fractional oxidation of carbohydrate during exercise in pregnancy. The serial changes in the five subjects who exercised at a constant intensity (50% max) throughout pregnancy were similar to those observed in the larger group. Before pregnancy, a small but significant increase in blood glucose (+ 0.35 ± 0.06 mmollL) was observed by the thirtieth minute of exercise, with a gradual return to preexercise levels during recovery. In the eighth week this did not occur and by the fifteenth week the response was reversed with a significant decrease in blood glucose (- 0.62 ± 0.08 mmollL) by the fifteenth minute of exercise that persisted (- 0.46 to - 0.64 mmollL) for the remainder of the exercise and recovery period (60 minutes). A similar pattern was observed at 23, 30, and 37 weeks, with average decreases in blood glucose level during exercise and
recovery of - 0.63, - 0.56, and - 0.61 mmoliL at the three respective time points of study. Norepinephrine levels at rest, during exercise, and at recovery were similar to those observed before conception until the thirtieth week. Thereafter, levels at rest and during exercise were significantly decreased (rest, 727 to 457 and 529 pg/ml; exercise, 1240 to 848 and 782 pg/ml), but the absolute change from rest to exercise was significantly decreased only at 37 weeks (change in norepinephrine, 513 to 271 pg/ml). Table IV details the serial changes in insulin and the respiratory exchange ratio during exercise in these five subjects. Again there was no significant change in insulin levels until the thirtieth week, when the levels both at rest and during exercise increased. The respiratory exchange ratio at standing rest increased significantly after the thirtieth week, but the values during exercise were increased significantly above preconceptional values throughout pregnancy (+ 0.056 ± 0.006 units). Comment
These serial data, gathered under conditions designed to mimic an individual's daily habits (dietary, temporal, exercise type, and intensity), demonstrate that the nonpregnant hyperglycemic response to recreational running and aerobics is initially blunted and then reversed early in the course of pregnancy. Because weight gain, caloric intake, and caloric mix were relatively constant and adequate, this suggests that some of
1682 Clapp and Capeless
the metabolic adaptations induced by pregnancy alter the balance between hepatic glucose production and peripheral use, which becomes apparent only when peripheral use is increased. A similar hypoglycemic shift, the so-called accelerated starvation of pregnancy,'6 is another example of this phenomenon. While the underlying mechanism for this shift during pregnancy is unclear, the increase in the respiratory exchange ratio during exercise in early pregnancy suggests that this may be due in part to a pregnancy-associated increase in the fractional use of carbohydrate by muscle during exercise. The superimposition of the constant and ever-increasing glucose use by the growing placenta and fetus also must increase the demand for glucose during exercise in late pregnancy.6. 7. '7 This additional demand is reflected in the increase in the respiratory exchange ratio at rest observed after the thirtieth week. The combination of increased glucose use by both exercising muscle and the fetoplacental unit probably is responsible for both the increase in the magnitude and the consistency of the hypoglycemic response during exercise in late pregnancy. In addition, the fact that there is no change in either the insulin or catecholamine response, coupled with the intensity dependence of the glycemic response reversal early in gestation, suggests that during exercise there is also a pregnancy-associated decrease in hepatic glucose production that is not apparent at rest.'7 This physiologic change also should alter glucose homeostasis, favoring a fall in blood glucose levels during exercise. A definitive answer to both possibilities (a pregnancy-associated increase in fractional glucose use by exercising muscle and a decrease in hepatic glucose production during exercise) will require further serial studies with the use of stable isotope infusion. The fact that the pregnancy-associated increase in the respiratory exchange ratio during exercise was consistently abolished at high exercise intensities with advancing gestation is an apparent contradiction and requires explanation. Unfortunately we do not have all the data necessary to identify the underlying mechanism. However, the consistency of the data suggests that it is real and not an interpretive error. The possibility. that carbon dioxide retention occurred during highintensity exercise in late pregnancy (which would factitiously lower the ratio) is unlikely because minute ventilation progressively increased and expired carbon dioxide concentration decreased during pregnancy. Thus we are left with the speculation that muscle glycogen stores at rest are significantly reduced in late pregnancy (possibly because of the increasing requirements of the placenta and fetus) with a resultant depletion of muscle glycogen earlier during high-intensity
December 1991 Am J Obstet Gynecol
exercise. When combined with the presumed suppression of hepatic glucose production and fall in blood glucose levels, this change could easily limit glucose availability to exercising muscle earlier in the latter half of pregnancy. A definitive answer would require serial muscle biopsy in a limited number of subjects during protracted high-intensity exercise. In terms of fetoplacental substrate availability, the data indicate that relatively brief, low- and moderateintensity recreational exercise rapidly lowers maternal blood glucose levels by 0.5 to 0.7 mmol/L on average and that the decrease persists for as much as 30 minutes after exercise. Whether an intermittent change of this degree is sufficient to significantly restrict fetal substrate supply and ultimately growth rate is currently unclear. However, the observation that this degree of running and aerobics in late pregnancy was associated with morphometric evidence of asymmetric growth restriction suggests that, coupled with other physiologic thanges (such as a probable reduction in placental bed blood flow), it may have an appreciable effect that potentially has therapeutic value. 7 . '0 For example, >70% of the exercise-associated decrease in birth weight was due to a 220 gm reduction in body fat, suggesting that reduced fat cell number could have prophylactic value. Likewise, regular, low- to moderate-intensity exercise in late pregnancy has been shown to be as effective as insulin in maintaining euglycemia in a limited number of subjects. '8. '9 Given a larger experience, it is likely that it will also be shown to reduce birth weight, which should reduce morbidity in the large-for-gestational age infant. Finally, it is unfortunate that there are no background data addressing the factors that influence glucose met:abolism during exercise in nonpregnant women. This study suggests that fit women who exercise regularly may have a different response from men ,2.5 because the response to sustained but shortduration exercise is consistently a hyperglycemic one whose magnitude is intensity dependent. In summary, the preconceptional hyperglycemic response to running and aerobics, seen in female recreational athletes, is reversed during pregnancy. The reversal is initially intensity dependent and does not appear to be related to changes in either insulin or norepinephrine response. The data suggest that it may reflect a pregnancy-associated decrease in hepatic glucose production and a pregnancy-associated increase in fractional glucose use by muscle during exercise. In late pregnancy the increased peripheral use of glucose is further increased by the demands of the fetoplacental unit, which magnify the change in blood glucose response and probably obscure its intensity dependence.
Glycemic response to exercise in pregnancy
Volume 165 Number 6, Part 1
The relative hypoglycemia associated with exercise persists for an as yet undefined period of time after exercise and may influence fetal growth,
REFERENCES 1, Coggan AR. Plasma glucose metabolism during exercise in humans. Sports Med 1991;11:102-24. 2. Montain SJ, Hopper MK, Coggan AR, Coyle EF. Exercise metabolism at different time intervals after a meal. J Appl PhysioI1991;70:882-8. 3. Cooper DM, Barsrow TJ, Bergner A, Lee WNP. Blood glucose turnover during high and low-intensity exercise. Am J Physiol 1989;257:E405-12. 4. Ahlborg G, Felig P, Hagenfeldt L, Hendler R, Wahren J. Substrate turnover during prolonged exercise in man. Splanchnic and leg metabolism of glucose, free fatty acids, and amino acids. J Clin Invest 1974;53:1080-90. 5. Ahlborg G, Felig P. Lactate and glucose exchange across the forearm, legs and splanchnic bed during and after prolonged exercise. J Clin Invest 1982;69:45-54. 6. DiGiacomo JE, Hay WW J r. Fetal glucose metabolism and oxygen consumption during sustained hypoglycemia. Metabolism 1990;39: 193-202. 7. Clapp JF. Utero-placental blood flow and fetal growth. In: Sharp F, Milner RDG, Fraser RB, eds. Fetal growth. Ashton-under-Lyne, England: Royal College of Obstetricians and Gynaecologists, Peacock Press, 1989:235-44. 8. Hauth JC, Gilstrap LC, Widner K. Fetal heart rate reactivity before and after maternal jogging during the third trimester. AMJ OBSTET GYNECOL 1982;142:545-7. 9. Clapp JF, Capeless EL. Neonatal morphometrics after endurance exercise during pregnancy. AM J OBSTET GyNECOL 1990; 163: 1805-11.
1683
10. Seaward BL, Sleamaker RH, McAuliffe T, Clapp JF. The precision and accuracy of a portable heart rate monitor. Biomed Instrum Technol 1990;24:37-41. 11. Clapp JF, Wesley M, Sleamaker RH. Thermoregulatory and metabolic responses to jogging prior to and during pregnancy. Med Sci Sports Exerc 1987;19:124-30. 12. Starr JI, Horwitz DL, Mako ME. Insulin, proinsulin and c-peptide. In: Jaffe BM, Berman HR, eds. Methods of hormone radioimmunoassay. New York: Academic Press, 1979:613-28. 13. Kontur P, Dawson R. Manipulation of mobile phase parameters for HPLC separation of endogenous monoamines in rat brain tissue. J Neurosci Methods 1984; 11 :518. 14. American College of Obstetricians and Gynecologists. Home exercise programs-exercise during pregnancy and the postnatal period. Washington: American College of Obstetricians and Gynecologists, 1985. 15. Lusk G. Science of nutrition. 4th ed. Philadelphia: WB Saunders, 1928. 16. Metzger BE, Ravnikar V, Vileisis RA, Freinkel N. "Accelerated starvation" and the skipped breakfast in late normal pregnancy. Lancet 1982; 1:588-92. 17. Kalhan SC, D'Angelo LJ, Savin SM, Adam PAJ. Glucose production in pregnant women at term gestation. Sources of glucose for the human fetus. J Clin Invest 1979;63:38894. 18. Bung P, Artal R, Khodiginan N, Kjos S. Exercise in gestational diabetes: an optional therapeutic approach. Diabetes 1991; [In press]. 19. Jovanovic-Peterson L, Durak E, Peterson CM. Randomized trial of diet versus diet plus cardiovascular conditioning on glucose levels in gestational diabetes. AM J OBSTET GYNECOL 1990;162:754-6.
Bound volumes available to subscribers Bound vol~mes of the AMERICAN JOURNAL OF OBSTETRICS AND GYNECOLOGY are available to subscribers (only) for the 1991 issues from the Publisher, at a cost of $61.00 for domestic, $88.27 for Canada, and $84.00 for international for Vol. 164 (January-June) and Vol. 165 (July-December). Shipping charges are included. Each bound volume contains a subject and author index and all advertising is removed. Copies are shipped within 60 days after publication of the last issue in the volume. The binding is durable buckram with the JOURNAL name, volume number, and year stamped in gold on the spine. Payment must accompany all orders. Contact Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis, MO 63146-3318, USA; phone (800) 325-4177, ext. 4351, or (314) 453-4351. Subscriptions must be in force to qualify. Bound volumes are not available in place of a regular JOURNAL subscription.