Pregnancy has little effect, if any, on gastrointestinal secretion or absorption, but it has a major effect on motility throughout the gastrointestinal tract. Recent studies suggest that motility changes are related to increased levels of circulating female sex hormones rather than, as once believed, to the enlarging uterus. Given the number of pregnancies per year, the costs in managing these problems can be substantial. Thus, our purpose was to review the effects of pregnancy on gastrointestinal motility to expand the knowledge of internists, gastroenterologists, and obstetricians about these common problems.

Methods

Top Methods Author & Article Info References

To compile a thorough review of pregnancy-related gastrointestinal motility disorders, their pathophysiology, associated clinical manifestations, and treatment, we used a variety of sources. We first did computerized literature searches of Index Medicus and MEDLINE using pregnancy and gastrointestinal system as the key words. Further references (including abstracts) were obtained by hand searches. References were selected for analysis if the article focused on alteration of motility during pregnancy, particularly if they emphasized hormonal effects. Animal studies, both in vivo and in vitro, that tried to illustrate hormonal effects on the gastrointestinal tract were also included. The latest editions of standard drug textbooks (PDR, American Hospital Formulary Service) and references pertaining to drug therapy during pregnancy were used to provide most of the information on the appropriate pharmacologic management of these disorders during pregnancy. Specific pharmaceutical corporations were contacted to obtain information about drug safety during pregnancy. We did not receive financial support from these companies.

Esophagus

Gastroesophageal Reflux Disease

Heartburn is estimated to occur in 30% to 50% of all pregnancies, with the incidence approaching 80% in some patient groups [2]. Although some authors [3] find that most women experience reflux symptoms after 5 months of gestation, Castro [4] suggests that these symptoms are reported only when they become most troublesome and that they begin much earlier in pregnancy. After detailed interviews of 60 patients, he noted that 52% first experienced heartburn during the first trimester of pregnancy; 24%, in the second trimester; and 8.8%, in the last trimester. Although heartburn tends to recur in subsequent pregnancies, no significant differences were noted in the incidence of heartburn among multiparous and primiparous women [5].

Pathophysiology

The pathogenesis of gastroesophageal reflux during pregnancy involves both mechanical and intrinsic factors that adversely affect lower esophageal sphincter tone. Early in pregnancy, lower esophageal sphincter pressure falls, returning to normal in the postpartum period. Van Thiel and colleagues [6] studied four previously asymptomatic pregnant patients sequentially at 12, 24, and 36 weeks of gestation and 1 to 4 weeks postpartum. Figure 1 shows that during all stages of pregnancy, resting lower esophageal sphincter pressures were less than the lower limits of normal for their motility laboratory, reaching a nadir at 36 weeks and returning to normal in the postpartum period. All four patients complained of frequent heartburn by 36 weeks. Fisher and colleagues [7] studied eight women during early pregnancy (mean gestation, 16 weeks) and 6 weeks after elective abortion. A mean lower esophageal sphincter pressure of 22.1 ± 2.4 mm Hg (± SE) was found before abortion, which was unchanged from the postabortion mean sphincter pressure of 22.6 ± 2.3 mm Hg. Although these patients had normal resting lower esophageal sphincter pressures, their sphincter pressures did not increase appropriately when challenged with injections of pentagastrin, edrophonium chloride, metacholine, and a protein meal. During each of these maneuvers, lower esophageal sphincter pressure was significantly lower during early pregnancy than after abortion. Thus, during pregnancy, not only is lower esophageal sphincter pressure decreased but adaptive responses of the sphincter may be reversibly inhibited.

View larger version (12K): [in this window] [in a new window]   Figure 1. Lower esophageal sphincter pressures. Data were recorded in four volunteer women during pregnancy and in the postpartum period. Area between the dotted lines shows the range of lower esophageal sphincter pressures in normal nonpregnant women. Horizontal bars represent the mean ± SE for each period. Lower esophageal sphincter pressure declined progressively during pregnancy but returned to normal in the postpartum period. (Adapted from Van Thiel DH, Gravaler JS, Joshi SN, et al. Heartburn of pregnancy. Gastroenterology. 1977; 72:666-8, with permission.).

Animal and human experiments have helped elucidate the important effects of female hormones on the lower esophageal sphincter. Fisher and colleagues [8] used in-vitro studies to examine strips of circular smooth muscle from the lower esophageal sphincter of opossums. Dose-response curves were constructed for gastrin and acetylcholine alone and with 17-ß-estradiol, progesterone, or the combination of sex hormones added to the system. The lower esophageal sphincter dose-response curve to acetylcholine and gastrin was blunted by estradiol as well as progesterone; progesterone was more effective in inhibiting the maximum response. The combination of progesterone and estradiol, however, was more potent than either hormone alone. In-vivo models using the opossum show a substantial reduction of lower esophageal sphincter pressure with the administration of both estradiol and progesterone [9]. Whether the decrease in lower esophageal sphincter function is due to estrogen, progesterone, or both is unclear. Filipone and colleagues [10] attempted to address this issue by studying five transsexual patients during a control period without hormonal stimulation, followed by periods of administration of estrogen, progesterone, or both. Resting lower esophageal sphincter pressure was significantly decreased during combination hormonal therapy (5.0 ± 0.1 mm Hg, mean ± SE) when compared with the control period (11.2 ± 2.1 mm Hg). No decrease in sphincter pressure was observed during either estrogen or progesterone administration alone. In addition, the adaptive lower esophageal sphincter pressure response to a protein meal was diminished by progesterone and by combination hormone therapy but not by estrogen alone. Similar observations were made by Van Thiel and colleagues [11] after examining normal menstruating women taking sequential oral contraceptives. The women's baseline mean lower esophageal sphincter pressure was 20.8 ± 1.7 mm Hg (mean ± SE) and did not change significantly during estrogen administration. Sphincter pressure, however, decreased to 9.4 ± 1.2 mm Hg during the administration of estrogen and progesterone. These studies suggest that progesterone is the mediator of lower esophageal sphincter smooth muscle relaxation; however, estrogen may be needed as a "primer" for this action to occur.

It was assumed for many years that the gravid uterus was responsible for an increase in abdominal pressure that overcame lower esophageal sphincter pressure and produced gastroesophageal reflux. However, Van Thiel and Wald [12] presented evidence that refuted this theory in a novel study using adult men with tense ascites as a model of pseudopregnancy. Before diuresis, the mean lower esophageal sphincter pressure was 30.9 ± 1.7 mm Hg (mean ± SE), which decreased significantly to 24.0 ± 1.6 mm Hg after loss of ascites. Thus, a compensatory increase in lower esophageal sphincter pressure usually parallels any increase in intra-abdominal pressure. This protective mechanism should prevent the enlarging uterus from promoting acid reflux.

Presentation and Diagnosis

The clinical features of gastroesophageal reflux in pregnancy do not differ from those found in the general population. Heartburn and regurgitation are the predominant symptoms, worsening as pregnancy advances. Not unexpectedly, Castro [4] found that most patients reported the adverse effects of food ingestion, causing some to restrict their meals to once a day because of intense postprandial heartburn. Likewise, the recumbent position aggravated heartburn in 82% of the patients, requiring some to sleep upright in a chair.

The diagnostic evaluation of pregnant patients with gastroesophageal reflux disease generally requires only a complete history. Barium studies should be avoided because of the radiation risk to the fetus. Endoscopy, although not commonly needed, is safe especially with current monitoring equipment [4]. Prolonged esophageal pH monitoring may be useful in atypical presentations. Fortunately, complications of reflux disease during pregnancy, such as esophagitis, are quite uncommon [4], and reflux symptoms are usually limited to pregnancy with no prolonged adverse effects on the mother or fetus.

Treatment

Because of potential teratogenicity of systemic drugs taken during gestation, lifestyle modifications are particularly important in treating pregnant patients with troubling heartburn. Patients are instructed to elevate the head of the bed and to avoid stooping, bending, or assuming other positions that tend to worsen reflux symptoms. Eating small frequent meals and refraining from ingesting foods or liquids other than water within 3 hours of bedtime are advised.

Nonsystemic drug therapy with antacids is the first step in treating the pregnant patient who does not respond to lifestyle modifications. Animal studies show no teratogenic effects from constant ingestion of antacids during pregnancy [13]. Adverse effects of antacids include interference with iron absorption and, specifically, metabolic alkalosis and fluid overload in both the fetus and mother with ingestion of sodium bicarbonate [14]. Sucralfate (Carafate), like antacids, appears to be safe because none of the drug is absorbed. In a randomized study involving 66 patients, Ranchet and colleagues [15] compared sucralfate (1 g three times daily) with lifestyle modifications in the treatment of heartburn associated with pregnancy. The sucralfate group had a significant improvement in both heartburn and regurgitation after 15 and 30 days. In addition, a greater proportion of sucralfate-treated patients than controls had complete remission of burning (90% versus 43%) and acidic regurgitation (83% versus 27%).

Systemic acid-inhibiting medications are used during pregnancy, although controlled studies are not available for ethical reasons. Animal studies show that H₂ blockers cross the placental blood barrier and are excreted in breast milk [16-18]. Teratologic studies find no adverse effect on litter parameters or early development of young rabbits [19, 20]. A study in rats, however, showed that cimetidine, but not ranitidine, when taken during pregnancy and the immediate postpartum period, can produce undesired feminization of the male fetus [21]. The largest human experience using H₂ blockers is with cimetidine. Data that SmithKline Beecham Pharmaceuticals collected in approximately 50 patients taking cimetidine during pregnancy are generally favorable. (Palmer R. Personal communication.) All women delivered normal babies after treatment with cimetidine ranging from 400 mg to 1 g per day for as long as the entire pregnancy. Armentano and colleagues [22] reported the case of a 39-year-old woman taking ranitidine, 150 mg/d, for reflux esophagitis during the last 6 months of her second pregnancy. Excellent control of reflux symptoms occurred, and a healthy baby was delivered. Cipriani and coworkers [23] described three more cases of reflux esophagitis in pregnancy treated with ranitidine, commenting on the well-being of the newborn. Even fewer human data are available concerning use of famotidine (Pepcid) and nizatidine (Axid) during pregnancy. Information on omeprazole (Prilosec) is limited to animal studies. Omeprazole produces dose-related increases in embryo—lethality, fetal resorptions, and pregnancy disruptions when administered in high doses to rabbits (up to 172 times the human dose). In rats, dose-related embryo-fetal toxicity and postnatal developmental toxicity were observed in offspring from parents treated with 35 to 345 times the human dose (Schackleford R. Personal communication). Omeprazole also crosses the placenta in near-term pregnant sheep [24].

Finally, promotility agents have no role in treating heartburn associated with pregnancy. Data supporting efficacy are lacking, and safer treatment options are available. Table 1 addresses the use and safety of specific drugs.

Gastroesophageal Reflux and the Risk for Aspiration during Anesthesia

Pregnant women requiring anesthesia during delivery are at high risk for aspirating gastric contents. In fact, aspiration during labor is the most frequent cause of obstetric morbidity and death due to anesthesia [25]. Predisposing factors are the same as those promoting gastroesophageal reflux during pregnancy, with the addition of recumbency and administration of anesthetic agents. Because gastric acidity is the key factor determining the severity of aspiration pneumonitis [26], prevention should be directed toward increasing the pH level of gastric contents. Early reports recommended prophylactic ingestion of antacids to maintain gastric pH levels greater than 2.5 during labor [27, 28]. More recently, H₂ blockers have been advocated for use during obstetric anesthesia. Hodgkinson and colleagues [29] evaluated the use of cimetidine or antacid therapy for elective cesarean section under general anesthesia. When both regimens were used, the mean gastric pH level at induction of labor was modestly elevated (pH 5.5). The mean volume of gastric contents in the cimetidine group was one third that measured in the antacid-treated group (P < 0.01). The effects of omeprazole on intragastric acidity were recently assessed during obstetric anesthesia [30]. Omeprazole (80 mg) was given orally the evening before surgery. The average pH level of stomach contents immediately after endotracheal intubation and immediately before extubation was 5.0. The use of promotility agents in this setting is limited. Brock-Utne and colleagues [31] examined the effect of intravenous metoclopramide (Reglan), at a 10-mg dose, on lower esophageal sphincter pressure in late pregnancy. Basal sphincter pressure increased 15% to 25% above baseline. Others [25] have found that metoclopramide may be beneficial in this setting by promoting gastric emptying. Antacids and H₂ blockers, however, remain the cornerstone of therapy for the prevention of aspiration pneumonitis.

Stomach

Nausea and Vomiting

Nausea is the most common gastrointestinal symptom of pregnancy, occurring in 50% to 90% of all pregnancies. Vomiting is an associated complaint in 25% to 55% of women [32-35]. Nausea accompanied by vomiting ("morning sickness") is more common in younger women, obese women, women from Western cultures, women with fewer than 12 years' education [33], women who experience nausea and vomiting while taking oral contraceptives [32], and women with a corpus luteum in the right ovary [36]. The nausea and vomiting of pregnancy begins shortly after the first missed menstrual period, peaks in the third month, and usually disappears by the fourth month of pregnancy. Symptoms generally occur in the early morning and improve later in the day. In one large study, 50% of women experienced symptoms in the morning, 7% reported nausea in the evening, 7% had nausea in the morning and evening, and 36% complained of symptoms all day [32]. The prognosis for both the mother and infant is excellent, with no increased risk for fetal deaths, low birth weights, or congenital malformations [37].

Pathophysiology

The cause of nausea and vomiting during pregnancy is unknown. Elevated serum levels of human chorionic gonadotropin, progesterone, and estradiol during pregnancy have been implicated, although psychological factors may be relevant, especially in women with severe and prolonged nausea and vomiting.

Data are conflicting about the relation between human chorionic gonadotropin and the nausea and vomiting of pregnancy. Serum levels are maximal during the first trimester of pregnancy when symptoms are most common [38, 39] and are elevated in women with hydatidiform mole who have a high incidence of nausea and vomiting [39]. More recent studies, however, have not found a specific relation between levels of chorionic gonadotropin and the incidence and severity of nausea in either pregnant women or women with molar pregnancies [40, 41].

Alterations in gastric motility associated with elevated levels of progesterone are implicated in the nausea of pregnancy. Animal studies show inhibitory effects of progesterone on gastrointestinal smooth muscle. Bruce and Behsudi [42] noted a significant reduction in contractile activity of excised esophageal, antrum, and colonic tissue from male rats pretreated with progesterone compared to nontreated male animals. In addition, Kumar and colleagues [43] showed that progesterone inhibits spontaneous contractile activity of human gastric muscle and colon in vitro, whereas estradiol and corticosteroids do not affect these contractile responses.

Several investigators have reported gender and menopausal differences in the gastric emptying rates of liquids and solids. Hutson and colleagues [44] evaluated mixed liquid and solid meal gastric emptying in men and three groups of women. Premenopausal and postmenopausal women with and without estrogen and progesterone replacement had slower gastric emptying of the liquid component than did men. Premenopausal and postmenopausal women being treated with sex hormone replacement had a decreased rate of gastric emptying of solids compared to men. In contrast, postmenopausal women without hormone replacement had rates of solid emptying similar to those of men. Datz and colleagues [45] reported similar results, but other investigators have not observed a difference in liquid or solid emptying between the sexes [46, 47]. These conflicting data may result from not controlling for the menopausal status of the women, caloric content of the meals, or scintigraphic imaging techniques.

Few human studies are available that evaluate gastric motor function during pregnancy because of the possibility of adverse fetal effects. In a preliminary study, Schade and colleagues [48] evaluated gastric liquid emptying using scintigraphy in 10 pregnant women during the second trimester before and 6 weeks after a therapeutic abortion. No impairment of liquid emptying was found; however, solid emptying was not evaluated and nausea was not present during the studies. Davison and colleagues [49] determined gastric emptying by serially aspirating stomach contents mixed with a known volume and concentration of phenol red after ingesting a liquid test meal in three groups: healthy, nonpregnant women, women in their third trimester of pregnancy, and women in established labor. A significant delay in liquid emptying was found during labor but not in the third trimester of pregnancy. In the labor group, gastric emptying of liquids returned to normal after delivery. Pain and emotional disturbances during labor were suggested as more likely causes of delayed gastric emptying than were altered hormone levels.

Recent studies using the electrogastrogram have found an increased prevalence of gastric dysrhythmias in nauseated pregnant women. This technique records gastric myoelectric activity by means of cutaneous electrodes positioned over the abdomen. Koch and colleagues [50] evaluated gastric myoelectric activity in 32 pregnant women: 26 women with nausea and 6 with either minimal or no nausea. Gastric dysrhythmias were demonstrated in all nauseated women: 17 had tachygastrias (4 to 9 cycle-per-minute [cpm] waves), 5 had bradygastrias (1 to 2 cpm waves), and 4 had a flat-line pattern (Figure 2). Of the six women with minimal to no nausea, all had normal 3-cpm patterns. Six women were studied again after delivery when not nauseated and had reverted to normal 3-cpm patterns. This technique shows great promise in evaluating the nausea and vomiting of pregnancy while correlating hormone levels with alterations of gastric myoelectrical activity. The test is simple, safe, and noninvasive, and may be easily repeated for serial studies.

View larger version (21K): [in this window] [in a new window]   Figure 2. Nausea scores and gastric rhythms in pregnant women. Mean nausea scores ± SE were significantly greater (P < 0.05) in women with flat-line dysrhythmias, 1- to 2-cpm waves, and 4-to 9-cpm tachygastrias compared with pregnant women with normal 3-cpm electrogastrogram rhythms. Differences in nausea scores among women in the various gastric dysrhythmias did not differ significantly. (From Koch KL, Stern RM, Vasey JJ, et al. Gastric dysrhythmias and nausea of pregnancy. Dig Dis Sci. 1990;35:961-8, with permission.).

Psychological factors may also play a role in the nausea and vomiting of pregnancy. In 1943, William Meninger [51] wrote that 50% of obstetricians regarded nausea and vomiting of pregnancy as psychologically based. Because symptoms disappeared by the fourth month of pregnancy, he postulated, nausea and vomiting represented an unconscious attempt to reject an unwanted child; symptoms ceased with the onset of fetal movement when further denial of the pregnancy was impossible. A recent psychological study suggests that women who have the nausea and vomiting of pregnancy are more likely to have more unplanned, unwanted pregnancies and have negative relationships with their mothers [34].

Treatment

The severity of nausea and vomiting dictates therapy. Mild symptoms are managed supportively with reassurance and avoidance of precipitating factors. Recommended dietary therapy consists of ingesting multiple small meals high in carbohydrates and low in fats.

Pharmacotherapy with antiemetics (see Table 1 such as the antihistamine derivative, meclizine [Bonine, Antivert], and the phenothiazine derivative, promethazine (Phenergan), may be necessary if symptoms are intractable. Adverse effects to the human fetus have not been reported; however, neither is recommended for routine use during pregnancy [52-54]. Bendectin (doxylamine succinate plus pyridoxine HCl), previously the only antiemetic approved for use during pregnancy, was withdrawn from the market because of suspected but unproved teratogenicity [55]. Data on the harmful fetal effects of other antiemetics such as prochlorperazine (Compazine), diphenhydramine (Benadryl), and trimethobenzamide (Tigan) preclude their use during pregnancy. Other phenothiazines such as chlorpromazine (Thorazine) should be avoided because of the potential for embryo toxicity, jaundice, and prolonged extrapyramidal effects in neonates [14]. Metoclopramide (Reglan) has been used as an antiemetic during pregnancy in Europe for almost 30 years [56]. Metoclopramide does not cause teratogenic effects in animals when given in doses 12 to 250 times the maximum recommended human dose [52], although it crosses the placenta and produces substantial fetal blood levels [57]. As with other antiemetics, controlled human data are not available regarding its use during pregnancy. However, anecdotal data from various nations have not found fetal abnormalities in several dozen pregnancies [58, 59].

Recently, Sahakian and colleagues [60] reported on a double-blind, placebo-controlled trial evaluating pyridoxine (vitamin B₆) in pregnant women with nausea or vomiting. Fifty-nine women completed the study: thirty-one received 25-mg tablets of vitamin B₆ orally every 8 hours and 28 received placebo. Fifteen B₆-treated patients had vomiting before therapy compared with 10 of 28 in the placebo group. After 3 days of treatment, the number of vitamin-treated women reporting vomiting was significantly decreased (8 of 31 patients, 26%) compared with the control group (15 of 28 patients, 54%, P < 0.005). Nausea was unchanged in women with mild or moderate complaints; however, severe nausea was significantly decreased (P < 0.05) in women treated with vitamin B₆. Vitamin B₆ has no known teratogenic risks or adverse side effects; therefore it may be a worthwhile therapeutic alternative in patients with severe nausea or vomiting.

Hyperemesis Gravidarum

Hyperemesis gravidarum is characterized by intractable vomiting occurring early in pregnancy. Unlike the uncomplicated nausea and vomiting of pregnancy, hyperemesis gravidarum is associated with fluid and electrolyte imbalances. The average incidence is 3.5 per 1000 deliveries [61]. Hyperemesis gravidarum is usually self-limited, disappearing by the third month of pregnancy. Although it rarely persists throughout pregnancy, frequent relapses may occur [61, 62].

The cause of hyperemesis gravidarum is poorly understood, but endocrine and psychological factors are thought to play a role. Like the nausea and vomiting of pregnancy, abnormal levels of human chorionic gonadotropin [41, 61], progesterone [41], estradiol [41], and even thyroid hormones [63] may be important. Social and psychological factors have also been associated with hyperemesis gravidarum. Fairweather [62] reported that maternal personality characteristics are important and suggested that women with immature or dependent personalities or both, are predisposed to this syndrome.

Metabolic consequences of hyperemesis gravidarum occur as a result of fluid and electrolyte depletion. Treatment is directed at replacement of fluids and electrolytes and at correction of nutritional deficiencies, if present. Small frequent meals high in carbohydrates are recommended. Although no antiemetic is approved for use during pregnancy, promethazine or metoclopramide may be necessary to increase the patient's threshold for emesis. Total parenteral nutrition may be required in patients with persistent weight loss, acidosis, and malnutrition [64]. In addition to medical therapy, adjunctive psychotherapy may be helpful in many cases [61].

Small Bowel and Colon

Abdominal Bloating and Constipation

Pregnant women frequently complain of abdominal bloating and constipation; however, data are conflicting regarding their prevalences. Levy and colleagues [65] interviewed 1000 healthy, Jewish women regarding their bowel habits before and during pregnancy. Fifty-five percent of women had no change in bowel frequency, 34% had an increase, and 11% had a decrease in frequency during pregnancy. Constipation requiring laxative treatment was rare (1.5%). Other surveys, however, found that nearly one third of pregnant women report constipation, usually during the first and third trimesters [66]. For example, Anderson [67] found that 38% of British women studied complained of constipation during their pregnancy, particularly in the last trimester. These varying demographic studies probably represent differences in dietary and cultural habits.

Pathophysiology

Abdominal bloating and constipation during pregnancy are probably caused by hormonally related changes in both small-bowel and colonic motility. Animal and human studies evaluating gastrointestinal transit during pregnancy and the menstrual cycle support this concept.

Small-Bowel Studies

Scott and DeFlora [68] compared small-bowel transit in rats during pregnancy (days 10, 15, and 20), the estrous cycle, and after castration by determining the distribution of a chromium-51 bolus introduced into the duodenum through an implanted catheter. Transit time, as assessed by geometric mean, was significantly faster (P < 0.05) in the castrated rats (4.96 ± 0.025) compared to the other groups (pregnancy, 3.95 ± 0.39; estrus, 3.35 ± 1.12). No differences in intestinal transit were observed between the pregnant and cycling rats. In another study, Scott and colleagues [69] studied small intestinal myoelectric activity in pregnant rats (days 12 to 18 of the 22-day gestation period) and in postpartum, nonpregnant, and castrated females. In all animals the interdigestive myoelectric complex was present and migrated from the proximal electrode to the distal electrode. However, the pregnant rats showed intervals between activity fronts that were quite prolonged compared to the other groups.

Small-bowel transit can be safely determined during pregnancy using the lactulose hydrogen breath test. After administration of a lactulose meal, breath hydrogen concentrations are determined at 10-minute intervals. The concentrations increase when the lactulose reaches the colon and undergoes bacterial fermentation. This technique minimizes the effect of gastric emptying by using a small volume of liquid, thereby primarily assessing small-bowel transit time. Using this technique, Wald and colleagues [70] measured gastrointestinal transit and sex hormone levels in 15 women during the third trimester of pregnancy and 4 to 6 weeks after delivery. Small-bowel transit time was significantly greater during pregnancy when progesterone and estradiol levels were substantially increased compared to the postpartum period (134 ± 14 minutes versus 93 ± 7 minutes, respectively) (P < 0.01) (Figure 3). Similarly, Braverman and colleagues [71] evaluated small-bowel transit in 10 women during the third trimester of pregnancy and on the second and fourth days postpartum. These findings were compared to eight age-matched women in the follicular phase of the menstrual cycle. Small-bowel transit time was prolonged during pregnancy and returned to values similar to controls after delivery. Changes in transit time paralleled the increase and subsequent postpartum decrease of serum progesterone levels. Finally, Lawson and colleagues [72] noted that intestinal transit times were prolonged in both the second and third trimesters of pregnancy (125 ± 48 minutes and 137 ± 58 minutes, respectively) compared to the first trimester of pregnancy or the postpartum period (99 ± 39 minutes and 75 ± 33 minutes, respectively). The transit times and levels of progesterone measured in the first trimester of pregnancy were not significantly different from those during the postpartum period.

View larger version (17K): [in this window] [in a new window]   Figure 3. Lactulose hydrogen breath test. Results from two women studied during the third trimester of pregnancy and after delivery. Breath hydrogen on the ordinate is shown plotted against time on the abscissa for two representative women, each of whom was studied twice. The open circles and broken lines represent studies done in late pregnancy, whereas the closed circles and solid lines represent studies performed in the postpartum period. Time is recorded from the moment of lactulose ingestion. Small bowel transit time was significantly greater during pregnancy compared to the postpartum period. (From Wald A, Van Thiel DH, Hoeschstetter L, et al. Effect of pregnancy on gastrointestinal transit. Dig Dis Sci. 1982; 27: 1015-8, with permission.).

These findings suggest that small-bowel transit is prolonged in late pregnancy, probably because of increased levels of progesterone. However, alternative factors may also contribute to the prolongation. During pregnancy, plasma levels of motilin, a stimulatory gastrointestinal hormone, are reduced, returning to normal levels 1 week postpartum. It has been suggested that progesterone may inhibit motilin release, thereby contributing to the delayed motility of pregnancy [73]. Further, it cannot be excluded that the gravid uterus may cause some mechanical impedance to small-bowel transit, especially late in pregnancy.

Colon Studies

Gill and colleagues [74] examined the effects of progesterone on the electrical and mechanical activities of the circular and longitudinal smooth muscle layers of the canine colon. Progesterone added to the muscle bath solution produced a dose-dependent reversible inhibition of the contractile forces in both types of muscle. The frequency of large-amplitude contractions and their electrical correlates recorded from longitudinally oriented strips also were decreased in a dose-dependent manner (Figure 4). Bruce and Behsudi [42] found that mid-transverse colon strips obtained from male Sprague-Dawley rats pretreated with progesterone showed a 53% inhibition of muscle activity compared to controls. Similarly, Scott and colleagues [75] compared the contractility of colon muscle from nonpregnant female rats and rats in late pregnancy. Their results confirmed that pregnancy decreased colonic muscle activity, with progesterone playing the major inhibitory role. Finally, Ryan and Bhojwani [76] examined the influence of female sex hormones on colonic transit in vivo using radiolabeled markers. Rats in a high estrogen-progesterone cycle stage had significantly slower colon transit times, as assessed by geometric mean, compared with animals in a low hormonal stage (1.97 ± 0.05, 4.25 ± 0.57, respectively; P < 0.001). Ovariectomized rats pretreated with estrogen and progesterone had significantly slower transit times than untreated ovariectomized animals (1.94 ± 0.19, 4.19 ± 0.17, respectively; P < 0.01). Pregnant rats had mean colonic transit times essentially identical to those of animals pretreated with hormone.

View larger version (17K): [in this window] [in a new window]   Figure 4. The effect of progesterone on canine colonic smooth muscle. Top panel. Increasing progesterone concentrations inhibited the contractile force (\#9679;) but not the frequency of circular smooth muscle as compared with the force () and frequency () of controls. Bottom panel. Increasing progesterone concentrations inhibited both the contractile force and frequency of longitudinal smooth muscle as compared with controls (Adapted from Gill RC, Bowes KL, Kingma YJ. Effect of progesterone on canine colonic smooth muscle. Gastroenterology. 1985; 88:1941-7, with permission.).

Human studies evaluating the influence of sex hormones on colonic motility and transit are few and contradictory. In muscle bath studies, Kumar [43] found that progesterone inhibited both the amplitude and frequency of spontaneous colon muscle activity, whereas estrogens and hydrocortisone had no effect. A later study by Hinds and associates [77] examined colon transit using radiopaque markers in women during the follicular and luteal phases of the menstrual cycle as well as in women taking oral contraceptives. These results were compared with transit times in male controls. No significant differences in colon transit were found between either phase of the menstrual cycle or between women and men. However, the authors suggested that their methods and small sample size may have precluded finding minimal differences among groups. Studies in pregnant women are not available because of potential radiation risks to the fetus.

Treatment

Use of laxatives during pregnancy may have serious side-effects to both mother and fetus [14] (see Table 1). Certain anthraquinone laxatives [Alon, Danthron] are associated with an increased risk for congenital malformations [78, 79]. Castor oil can initiate premature uterine contractions. Saline laxatives such as Milk of Magnesia may produce sodium retention in the mother [80]. Excessive use of mineral oil can decrease maternal absorption of fat-soluble vitamins resulting in neonatal hypoprothrombinemia and hemorrhage [81]. Some laxatives also produce diarrhea in the neonate [80]. Stimulant laxatives appear safe during pregnancy [14]; however, the phenolphthalein laxatives are excreted in breast milk and may cause colic in breast-fed infants [82]. Lactulose, sorbitol, and glycerin are not teratogenic in animals and thus are safe for use during pregnancy. The safest agents, however, are bulk-forming preparations containing fiber (for example, Metamucil, Citrucel, and Perdiem) because no systemic absorption occurs. Increasing dietary fiber content or using bulk agents is thus the preferred treatment of constipation. Information is minimal regarding the secretion of laxatives in breast milk; thus care must be taken using these medications while breast feeding.

Diarrhea

Diarrhea occurring during pregnancy has the same cause as in the nonpregnant patient (for example, drugs, osmotic diarrhea, malabsorption, infections) and a routine evaluation is therefore indicated. Nonspecific diarrhea can be managed with various medications (see Table 1). Nonsystemic medications should be tried first. Bulk agents may be useful because of their ability to absorb water. Kaolin with pectin [Kaopectate] is not absorbed from the gastrointestinal tract and has no known systemic effects [83]. Systemic therapies for diarrhea during pregnancy are limited by safety concerns. Loperamide (Imodium) given to pregnant rats and rabbits at doses up to 30 times the human dose produced no harm to the fetus; however, higher doses impaired maternal and neonatal survival. Teratogenic effects were not observed [84]. Human data are limited to 21 spontaneous reports of patients ingesting loperamide during pregnancy in the United States and abroad. Unfortunately, only 9 of 21 women had follow-up. Four newborns were unaffected, two pregnancies resulted in spontaneous abortions, two babies were born with congenital malformations, and one had bilateral erbs palsies (Cohen K. Personal communication). Nevertheless, we believe it is probably safe if given only when clearly needed and in small, limited doses. It is unknown whether loperamide is distributed in breast milk [83]. Diphenoxylate with atropine (Lomotil) produces teratogenic effects in animals, including impaired fertility and growth retardation in female rats treated with 20 mg/kg body weight per day [50]. In humans, malformations occurred in several infants exposed to Lomotil in the first trimester of pregnancy [79]. Both agents are excreted in breast milk, and atropine may inhibit lactation [14]. Bismuth subsalicylate should not be used during pregnancy because salicylate absorption may occur. Salicylates are teratogenic in animals, and neonatal hemorrhage has been reported in humans [85]. The effects of bismuth absorption on the fetus are unknown [14].

Conclusions

Motility disturbances of the gastrointestinal tract are common in uncomplicated pregnancies. These disturbances are largely attributed to alterations in sex hormones, resulting in various clinical manifestations. Gastrointestinal symptoms resulting from or aggravated by pregnancy include heartburn, nausea and vomiting, abdominal bloating, and constipation. Patient education, environmental modification, and reassurance that these symptoms are a normal occurrence in pregnancy may be all that is needed. Drug therapy is reserved for patients with intractable symptoms. An understanding of the potential teratogenicity and safety of drug therapies used during pregnancy is vital for the clinician managing these patients.

Part of this material was presented at a symposium of the American Gastroenterological Association in May 1991.