|
|
| |||||||||||||||||||||||||||||||
Background The potential risks and benefits of regular exercise during lactation have not been adequately evaluated. We investigated whether regular aerobic exercise had any effects on the volume or composition of breast milk.
Methods Six to eight weeks post partum, 33 sedentary women whose infants were being exclusively breast-fed were randomly assigned to an exercise group (18 women) or a control group (15 women). The exercise program consisted of supervised aerobic exercise (at a level of 60 to 70 percent of the heart-rate reserve) for 45 minutes per day, 5 days per week, for 12 weeks. Energy expenditure, dietary intake, body composition, and the volume and composition of breast milk were assessed at 6 to 8, 12 to 14, and 18 to 20 weeks post partum. Maximal oxygen uptake and the plasma prolactin response to nursing were assessed at 6 to 8 and 18 to 20 weeks.
Results The women in the exercise group expended about 400 kcal per day during the exercise sessions but compensated for this energy expenditure with a higher energy intake than that recorded by the control women (mean [±SD] intake, 2497 ±436 vs. 2168 ±328 kcal per day at 18 to 20 weeks; P<0.05). Maximal oxygen uptake increased by 25 percent in the exercising women but by only 5 percent in the control women (P<0.001). There were no significant differences between the two groups in maternal body weight or fat loss, the volume or composition of the breast milk, the infant's weight gain, or maternal prolactin levels during the 12-week study.
Conclusions In this study, aerobic exercise performed four or five times per week beginning six to eight weeks post partum had no adverse effect on lactation and significantly improved the cardiovascular fitness of the mothers.
There is very little information available on the effect of exercise on lactation in humans. In an earlier cross-sectional study, we found no differences in milk volume or composition between eight sedentary women and eight highly trained women who exercised11. However, the subjects were self-selected, and the women in the exercise group may have been unusual in their capacity to exercise intensively while lactating. We therefore undertook a randomized study to assess the effect of initiating an exercise program during lactation. In this paper we report the outcomes related to energy expenditure and lactation.
Methods
Study Design
Breast-feeding women were randomly assigned to either the exercise group or the control group. Those in the exercise group undertook a program of aerobic exercise lasting 45 minutes per day, 5 days per week, for 12 weeks beginning 6 to 8 weeks post partum. Those in the control group did not engage in aerobic exercise more than once per week during the same period. The women who were enrolled met the following criteria: they had no chronic illness, used no regular medication, and did not smoke; they had not exercised more than twice per week during the previous 3 months; they planned to breast-feed their infants exclusively for at least 20 weeks; and they had delivered healthy infants at term (37 to 43 weeks' gestation). Measurements were made at base line (6 to 8 weeks post partum), at the midpoint of the study (12 to 14 weeks), and at the end of the study (18 to 20 weeks). The women's energy expenditure, resting metabolic rate, body composition, dietary intake, and breast-milk volume and composition were measured at all three points. Maximal oxygen uptake and plasma prolactin levels were measured at the beginning and end of the study. The protocol was approved by the Human Subjects Review Committee of the University of California, Davis.
The participants were recruited through letters to new parents; 88 eligible women responded. Of these, 50 did not participate: 32 because of time constraints, and 18 because they did not want to risk being randomly assigned to the control group. Five of the 38 participants did not complete the study: 1 was excluded because of unreliable data, 1 had a previously undetected thyroid condition, 2 withdrew, and 1 had very low milk production at base line that required intervention, making her ineligible to continue in the study. All but the last woman had been assigned to the control group. The five women who withdrew were similar to those who remained in terms of their weight, height, age, education, and parity and the sex of their infants, but their infants had significantly lower birth weights than the infants of the women who remained in the study (mean [±SD] birth weight, 3.31 ±0.20 vs. 3.77 ±0.46 kg; P<0.05).
Exercise Program
Exercise sessions were individually tailored and included rapid walking, jogging, or bicycling. An assistant monitored the heart rate of each woman during each session. Sessions began with 10 minutes of stretching and low-intensity exercise, followed by aerobic exercise prescribed to achieve 60 to 70 percent of the heart-rate reserve,12 as measured by the graded exercise (oxygen-consumption) test. The program began with 20-minute sessions, with 5-minute increments every three days until the woman could complete 45 minutes of continuous exercise at the target heart rate. The exercise-frequency goal was five times per week, but the women occasionally missed some sessions because of illness, injury, or other reasons. The average frequency of exercise was 4.5 times per week. If a woman missed three or more days of exercise in any week (as was the case for three women), an extra week was added to the program.
Energy Expenditure and Maximal Oxygen Consumption
The resting metabolic rate was determined by indirect calorimetry on two mornings by methods described previously11,13. The average of the two values was used unless the coefficient of variation was greater than 7 percent, in which case a third day of measurements was scheduled. The mean (±SD) coefficient of variation between the measurements on the two days was 3.2 ±2.5 percent.
Heart-rate monitoring was used to estimate energy expenditure, as described elsewhere13. Calibration was performed individually for each woman during each study period (at base line and at the midpoint and end of the study) by simultaneously measuring the heart rate and the oxygen consumption of each woman while she was sedentary (resting, sitting, or standing) and active (walking on a treadmill at various speeds or riding a stationary bicycle). During each study period, each woman wore a heart-rate monitor (Uniq Heart Watch, model 8799, Computer Instruments, Westbury, N.Y.) for three days (for women in the exercise group, this included two exercise days). Oxygen consumption during sleep was assumed to be equivalent to that during the measurement of the resting metabolic rate13,14. The total daily energy expenditure was calculated by totaling oxygen consumption during periods of sleep and activities other than sleep and multiplying that value by 4.83 kcal per liter of oxygen (20.2 kJ per liter). Group means for energy expenditure based on heart-rate monitoring are very similar to those obtained with the method using doubly labeled water13.
For most women, maximal oxygen consumption was measured on a treadmill at 4.8 to 7.2 km (3.0 to 4.5 miles) per hour, with progressive increases in grade until the woman could no longer continue, as described elsewhere11. Three women whose main form of exercise was bicycling used a bicycle ergometer for the determination of maximal oxygen consumption, with the work rate increased by 25 W every two minutes. If the maximal heart rate was considerably lower than the predicted value (220 minus the woman's age) or if there was no plateau in oxygen consumption with increasing workload, the test was repeated on another day.
Dietary Intake
Dietary intake was recorded by each woman with a tape recorder during the three-day periods of heart-rate monitoring. The portions of food consumed were weighed to the nearest 2 g. Nutrient intake was calculated with the Food Processor II computer program (ESHA Research, Salem, Oreg.), food-composition tables, and data from food manufacturers.
Anthropometric Measurements
The women were weighed on a beam-balance scale to the nearest 10 g. Height was measured with a steel tape measure and headboard to the nearest 0.5 cm. Body composition was determined by hydrostatic weighing15. The infants' weights were measured to the nearest gram on an electronic balance.
Volume and Composition of Breast Milk
Test weighing was used to measure the milk volume during the same three-day periods in which the dietary and heart-rate records were compiled. Mothers weighed their infants before and after each feeding with an electronic balance (Sartorius 3826, Brinkmann Instruments, Westbury, N.Y.) accurate to 1 g. The total was adjusted for insensible water loss, as described elsewhere16. This method compares favorably with indirect measures of milk volume based on isotope-dilution techniques17.
Milk samples were collected by the alternate complete expression of one breast at each feeding for 24 hours, with an electric breast pump, as described elsewhere16. Samples of milk proportional to the volume pumped at each feeding were pooled for analysis. Total and nonprotein nitrogen levels were determined by micro-Kjeldahl analysis,18 and the protein concentration was calculated as 6.25 x (total nitrogen - nonprotein nitrogen). Lipid was measured gravimetrically after a modified Folch extraction19. Lactose was analyzed by the colorimetric procedure described by Dahlqvist20. The mean (±SD) coefficient of variation for duplicate samples was 2.7 ±3.5 percent for protein, 2.2 ±1.4 percent for lipids, and 1.4 ±1.2 percent for lactose. Gross energy density was calculated by multiplying the values for protein, lipid, and lactose by 5.65, 9.25, and 3.95 kcal per gram (23.7, 38.7, and 16.5 kJ per gram), respectively, and summing the products.
Plasma Prolactin Concentrations
Blood samples were collected in the morning, at least two hours after the previous breast-feeding and one hour after the last meal. An indwelling catheter with a heparin lock was inserted into the woman's arm or hand vein. After 20 minutes, a basal sample of blood was obtained. The infant was then allowed to breast-feed, and blood samples were obtained 10, 20, 30, 45, 60, 90, and 120 minutes after nursing began. After centrifugation, the plasma was frozen at -20 °C. Prolactin was analyzed in duplicate (coefficient of variation, 4.8 ±3.6 percent) with a radioimmunoassay kit (Radioassay System Laboratory, Carson City, Calif.).
Statistical Analysis
Data were analyzed with SAS-PC software21. Student's t-tests and chi-square tests were used to compare characteristics in the two groups. Multivariate repeated-measures analysis of variance was used to evaluate differences between the groups over time. Results are reported as means ±SD. All P values are two-tailed.
Results
There were no significant differences in the base-line characteristics of the women in the exercise group (n = 18) and the control group (n = 15) (Table 1). Because the proportion of female infants was somewhat higher in the exercise group than in the control group, the infants' weight at base line averaged 327 g less in the exercise group. All infants were exclusively breast-fed throughout the study.
|
|
The intake of breast milk by the infants was slightly lower in the exercise group than in the control group at all three study points when expressed in terms of grams per day (because the infants of the exercising mothers were smaller initially and thus consumed less milk), but not when expressed in terms of grams per kilogram of body weight per day. The breast-milk composition (lipid, protein, and lactose concentrations, and energy density) did not differ between the two groups except for the concentration of protein, which was significantly higher in the exercise group than in the control group at the final measurement. However, the change in the protein concentration did not differ between the groups (P = 0.80); both declined significantly over time (P<0.001 by repeated-measures analysis of variance). As for the volume of breast milk, the output of energy in milk was somewhat lower in the exercise group than in the control group at all measurements when expressed in terms of kilocalories per day, but not when adjusted for the body weight of the infants (kilocalories per kilogram per day). There were no significant differences between the two groups in the changes in breast-milk intake by the infants, energy output in milk, or the infants' body weight during the study (Figure 1).
|
Basal and peak plasma prolactin levels declined during the study in both groups (Figure 2). The rate of decline did not differ significantly between the two groups. The frequency and duration of nursing also declined significantly in both groups; neither the mean values nor the rate of decline differed significantly between the two groups (data not shown). Menstrual cycles resumed before 20 weeks after delivery in three control women and one woman in the exercise group (P = 0.22).
|
This study demonstrates that aerobic exercise performed four to six times per week beginning six to eight weeks post partum has no adverse effect on lactation and significantly improves the cardiovascular fitness of the mother. The results are consistent with those of our earlier cross-sectional study11 and indicate that exercise is safe, even for previously sedentary women who are breast-feeding their babies. In both studies, women in the exercise group compensated for their increased energy expenditure with a higher energy intake. In this study, as a result, the rate of weight loss and the decline in the percentage of body fat after delivery did not differ between the exercise and control groups. Although a longer-term exercise program would be expected to reduce body fat, breast-feeding women should be aware that exercise alone will not increase their rate of weight loss unless their dietary intake is controlled. They should be advised, however, that the consequences of rapid weight loss during lactation are unknown. The Institute of Medicine recommends that beyond the first month after delivery, lactating women should not lose more than 2 kg per month23.
Although mean energy expenditure was higher in the exercise group than in the control group -- by almost 300 kcal per day at the midpoint of the study -- this difference narrowed by the end of the 12-week program because the exercising women cut back on other activities. This finding suggests that among breast-feeding women, high levels of energy expenditure may be difficult to sustain because of fatigue or time constraints.
Basal prolactin levels did not differ between the two groups, a finding that is consistent with the results of our earlier study11. This finding suggests that the short-term increase in prolactin levels after exercise in nonlactating women that has been reported by others4,5,6 does not influence basal levels. The prolactin response to nursing was highly variable, a fact that probably reflects differences in the infants' sucking and the mothers' basal prolactin levels. In malnourished women, basal prolactin levels are higher during conditions of greater energetic stress,24 but this effect is probably mediated through the infants' frequency and duration of nursing. In the present study, the frequency and duration of nursing did not differ between the exercise group and the control group.
The mothers in the exercise group did not mention any difficulties in nursing after exercise, despite the recent finding of Wallace et al.25 that after exercise breast milk may have higher concentrations of lactic acid, which can affect the taste of the milk and its acceptability to the infant. However, the women in that study had undergone a maximal-exercise test; lactic acid levels in milk are less likely to be elevated after moderate exercise26.
In conclusion, we have demonstrated that breast-feeding women can safely undertake a program of frequent, moderate aerobic exercise without adversely affecting the volume or composition of their breast milk.
Supported by a grant (HD 24112) from the National Institutes of Health.
We are indebted to the Department of Physical Education of the University of California, Davis, for allowing us to use the facilities of the Human Performance Laboratory; to Wendy Kristy and Geanne Lyons for their tireless assistance with data collection; to our enthusiastic team of student assistants; to Kenneth H. Brown for his comments on the manuscript; and, finally, to the study participants, who gave so generously of their time and energy.
Source Information
From the Department of Nutrition, University of California, Davis (K.G.D., L.A.N., M.A.M., B.L.), and the Department of Food, Nutrition and Food Service Management, University of North Carolina, Greensboro (C.A.L.).
Address reprint requests to Dr. Dewey at the Department of Nutrition, University of California, Davis, CA 95616-8669.
References
| |||||||||||||||||||||||||||||||
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
