Drugs in Pregnancy and Lactation: Tenth Edition


Respiratory Drug (Antitussive)




The available human data on the reproductive effects of dextromethorphan do not demonstrate a major teratogenic risk. Except for one study involving chick embryos, there are no published animal reproductive studies. Unpublished pregnant rat and rabbit data have been cited, though, that indicated there was no embryo or fetal harm with the doses used. Extrapolation of the chick embryo data to humans is not possible because of the lethal dose used and the absence of maternal and placental metabolizing systems. Moreover, prior teratology studies have not shown agreement between the chick and mammalian models (1). Fetuses of women with the phenotype for slow dextromethorphan metabolism should have higher concentrations of the drug than fetuses of mothers with normal metabolism, but this may not be clinically significant in the absence of a demonstrated dose–effect. Use of liquid preparations of dextromethorphan that contain ethanol, however, should be avoided during pregnancy because ethanol is a known teratogen.

Many authors consider dextromethorphan to be safe for consumption during pregnancy. This opinion appears to have been based on the low incidence of congenital defects reported in surveillance studies and its wide appeal as a cough suppressant, rather than on evidence derived from human pregnancy research with dextromethorphan or from studies in any animal model. The latest data, however, provide more assurance that dextromethorphan is not a major teratogen.


Dextromethorphan, a derivative of the narcotic analgesic levorphanol (see also Levorphanol) that is widely used as a cough suppressant in over-the-counter (OTC) preparations, produces little or no CNS depression. Although it is an antitussive without expectorant, analgesic, or addictive characteristics, abuse of the liquid product, possibly because of the ethanol vehicle used in some proprietary mixtures, is a potential complication (see case below). The agent is available either alone (as capsules, lozenges, or oral solutions) or in combination with a large variety of other compounds used for upper respiratory tract symptoms. Combination products containing ethanol should be avoided during pregnancy (see also Ethanol).

No information is available on the placental transfer of dextromethorphan. The molecular weight (about 271) is low enough that transfer to the fetus should be expected.

Only one published animal reproduction study involving dextromethorphan has been located (see reference 2 for unpublished data). A 1998 report examined the effects of dextromethorphan on chick embryos (3). The authors hypothesized that N-methyl-D-aspartate (NMDA) receptor antagonists, such as ethanol and dextromethorphan, induced neural crest (craniofacial and cardiac septal defects) and neural tube defects (NTDs). Dextromethorphan, an NMDA receptor antagonist that acts as a channel blocker at the receptor, was injected into chick embryos in ovo for 3 consecutive days at doses of 0.5, 5, 50, and 500 nmol/embryo/day. The embryos were examined 24 hours after the third dose (none of the embryos were allowed to hatch). Dextromethorphan caused a dose-related increase in embryo mortality with rates of 14.1% with the 50-nmol dose (p <0.05 vs. controls) and 56.7% with the 500-nmol dose (p<0.001 vs. controls). Seven (1.2%) of the 595 control embryos injected only with the vehicle (normal saline) had anomalies (spinal defect [N = 1], craniofacial defects [N = 4], and multiple defects [N = 2]), whereas 30 (8.0%) of all dextromethorphan-treated embryos had anomalies (spinal defect [N = 1], craniofacial defects [N = 12], multiple defects [N = 16], and other defects [N = 1]). Only the number of defects (about 15%) in the 500-nmol group, however, was significantly increased (p <0.001) over the number in the control group. The authors cited published evidence that the receptors blocked by dextromethorphan in the chick embryos are analogous to receptors in other animals, including humans, during early development and that the drug would also block these receptors, resulting in similar malformations (3).

Interpretation of the results of the above study have been criticized (1,2,4) and defended (5). The primary concerns raised were the inappropriateness of the chick embryo model for determining human teratogenicity and the design of the study (1,4). Of particular concern, no embryos were allowed to hatch (the doses used were lethal) so it could not be determined if dextromethorphan was actually teratogenic. In addition, one author cited unpublished animal reproduction data from a drug manufacturer showing that in pregnant rats and rabbits, daily doses up to 20 and 100 times the human therapeutic dose on a body weight basis, respectively, caused no embryo or fetal harm in comparison with controls (1).

Metabolism of dextromethorphan has been shown to be primarily a result of O-demethylation to dextrorphan (6). The ability to metabolize many drugs, including O-demethylation of dextromethorphan, is genetically determined in adults. In this study of 155 adult volunteers, 144 (93%) metabolized dextromethorphan rapidly, and 11 (7%) were poor metabolizers, but the poor (slow) drug metabolizer phenotype has been reported in 5%–10% of whites (6). In poor metabolizers, unmetabolized drug was the main excretion product, implying that these individuals had much higher and more prolonged plasma concentrations of dextromethorphan than did those who were extensive metabolizers. Moreover, using microsomal preparations obtained from aborted 10- to 30-week-old human fetuses and live newborn infants, the average activity of O-demethylation was less than 1% of the adult value and did not begin to rise until after birth (6). Thus, accumulation of unmetabolized dextromethorphan in the fetal compartment is a potential result of maternal ingestion of the drug during pregnancy (6).

The Collaborative Perinatal Project monitored 50,282 mother–child pairs, 300 of whom took dextromethorphan during the 1st trimester (7, p. 378). Twenty-four of the infants exposed in utero had a congenital malformation, a standardized relative risk (SRR) of 1.18. When only malformations showing uniform rates by hospital were considered, 17 (SRR 1.21) infants had a congenital defect (7, p. 379). Of these 17, 9 had major defects (SRR 1.10) and 8 had minor defects (SRR 1.30) (7, p. 382). For use anytime during pregnancy, 580 exposures were recorded, 15 of which had a malformation (SRR 1.39) (7, pp. 438, 442). Ten of these defects were inguinal hernias (7, p. 496). The SSRs do not support a relationship between the drug and congenital malformations.

A case report published in 1981 described a woman who consumed 480–840 mL/day of a cough syrup throughout pregnancy (8). The potential maximum daily doses based on 840 mL of syrup were 1.68 g of dextromethorphan, 16.8 g of guaifenesin, 5.0 g of pseudoephedrine, and 79.8 mL of ethanol. The infant had facial features of the fetal alcohol syndrome (bilateral epicanthal folds; short palpebral fissures; short, upturned nose; hypoplastic philtrum and upper lip with thinned vermilion; and a flattened midface [8]). (See also Ethanol.) Other defects noted were an umbilical hernia and labia that appeared hypoplastic. The infant displayed irritability, tremors, and hypertonicity. It is not known if dextromethorphan or the drugs other than ethanol were associated with the adverse effects observed in the infant.

The use of dextromethorphan in four of five cases of a rare and distinct malformation complex was reported in 1984 (9). The complex of defects included absence of external genitalia, urinary, genital, and anal orifices, and persistence of the cloaca (9).

Chromosome analysis in four of the cases was normal and genetics did not appear to be a cause of the defects. Although a causative relationship could not be determined, the authors noted that three of the pregnancies, possibly all five, were exposed to doxylamine during the first 50 days of pregnancy (9). Dextromethorphan, however, was not thought to be related to the outcomes because only four of the mothers had symptoms of respiratory infection or took dextromethorphan during the critical period.

A surveillance study published in 1985 examined the prevalence of certain major birth defects among liveborn infants of 6509 mothers (10). Dextromethorphan was assumed to have been used by 59 of the mothers, only one of whom gave birth to an infant with a major anomaly. This study found no strong association between any of the commonly used drugs and the congenital malformations surveyed (10).

Data from the Spanish Collaborative Study of Congenital Malformations (ECEMC) evaluating prenatal exposure to cough medicines containing dextromethorphan were published in 2001 (11). Using standardized methods, all newborn infants born in more than 77 hospitals throughout Spain were examined during the first 3 days of life for major and/or minor congenital defects. The case–control study was conducted between 1976 and 1998 and included 1,575,388 liveborn infants, 27,864 of whom had congenital defects detected during the first 3 days. Each case infant (those with defects) and its control (the next nonmalformed, same sex-infant born) were obtained from the same hospital. Among the case and control mothers, 0.26% (N = 70) and 0.18% (N = 48), respectively, had taken cough medicines containing dextromethorphan during the 1st trimester. The data were primarily analyzed for NTD and cardiac defects to test the hypothesis raised in the chick embryo study cited above (reference 3), but about 600 different categories of defects were also examined. Most of these categories, however, had no cases or controls. The adjusted (for maternal age, fever, drugs other than dextromethorphan, flu/cold, first-degree relatives with the same defects) odds ratio and 95% confidence intervals for selected anomalies were NTD 0.67 (0.09–4.94), CNS defects 1.36 (0.39–4.72), hydrocephaly 3.39 (0.38–30.35), congenital heart defects 0.92 (0.13–6.61), oral clefts 4.72 (0.55–40.24), and cleft palate 3.26 (0.35–30.34). When data on the amount consumed were available, the estimated total dextromethorphan dose and duration for cases and controls was about 101 mg/2.69 days and 117 mg/3.96 days, respectively. The investigators concluded that the use of dextromethorphan during the 1st trimester was not associated with an increase in congenital defects (11).

A 1984 review on the effect of OTC drugs on human pregnancy concluded that dextromethorphan was safe to use during this period (12). Other reference sources (1316) have also concluded that this antitussive does not pose a risk to the human fetus, and a 1998 source states that dextromethorphan is one of the drugs of choice during pregnancy for cough (17). Two of these references recommended a combination of guaifenesin plus dextromethorphan as the preferred antitussive in pregnant asthmatic patients (15,16).

A study published in 2001 described the outcomes of 184 pregnancies exposed to dextromethorphan, 128 of which were exposed in the 1st trimester, compared with 184 matched controls (18). The subjects were women who had called a teratogen information service concerning their use of dextromethorphan during pregnancy. There were six major birth defects (type not specified) in the subject group. One was a chromosomal abnormality and two were born to women who had used the antitussive after the 1st trimester. Among controls, there were five major birth defects (type not specified), one of which was a chromosomal abnormality. There were also no statistical differences between all subjects and controls in other outcomes: live births (172 vs. 174), spontaneous abortions (10 vs. 8), therapeutic abortions (1 vs. 2), stillbirths (1 vs. 0), minor malformations (10 vs. 8), and birth weight (3381 g vs. 3446 g) (18).


No reports describing the use of dextromethorphan during human lactation or measuring the amount, if any, excreted into milk have been located. The relatively low molecular weight of dextromethorphan (about 271) suggests that passage into milk probably occurs. Many preparations containing dextromethorphan also contain ethanol. These products should be avoided during nursing (see Ethanol). Preparations without ethanol, however, are probably safe to use during breastfeeding.


1.Brent RL. Studies of the fetal effects of dextromethorphan in ovo. Teratology 1999;60:57–8. (Originally published as an untitled letter to the editor: Pediatr Res 1998;44:415–6.)

2.Polifka JE, Shepard TH. Studies of the fetal effects of dextromethorphan in ovo. Teratology 1999;60:56–7. (Originally published as an untitled letter to the editor: Pediatr Res 1998;44:415.)

3.Andaloro VJ, Monaghan DT, Rosenquist TH. Dextromethorphan and other N-methyl-D-aspartate receptor antagonists are teratogenic in the avian embryo model. Pediatr Res 1998;43:1–7.

4.Brent RL, Shepard TH, Polifka JE. Response to Dr. Rosenquist’s comments pertaining to the paper by Andaloro et al. (‘98) “Dextromethorphan and other N-methyl-D-aspartate receptor antagonists are teratogenic in the avian embryo model” and letters to the editor by Polifka JE and Shepard TH (‘98) and Brent RL (‘98). Teratology 1999;60:61–2.

5.Rosenquist TH. Studies of the fetal effects of dextromethorphan in ovo. Teratology 1999;60:56–60. (Originally published as an untitled author’s response: Pediatr Res 1998;44:416–7.)

6.Jacqz-Aigrain E, Cresteil T. Cytochrome P450-dependent metabolism of dextromethorphan: fetal and adult studies. Dev Pharmacol Ther 1992;18:161–8.

7.Heinonen OP, Slone D, Shapiro S. Birth Defects and Drugs in Pregnancy. Littleton, MA: Publishing Sciences Group, 1977.

8.Chasnoff IJ, Diggs G, Schnoll SH. Fetal alcohol effects and maternal cough syrup abuse. Am J Dis Child 1981;135:968.

9.Robinson HB Jr, Tross K. Agenesis of the cloacal membrane. A probable teratogenic anomaly. Perspect Pediatr Pathol 1984;8:79–96.

10.Aselton P, Jick H, Milunsky A, Hunter JR, Stergachis A. First-trimester drug use and congenital disorders. Obstet Gynecol 1985;65:451–5.

11.Martinez-Frias ML, Rodriguez-Pinilla E. Epidemiologic analysis of prenatal exposure to cough medicines containing dextromethorphan: no evidence of human teratogenicity. Teratology 2001;63:38–41.

12.Rayburn WF. OTC drugs and pregnancy. Perinatol Neonatol 1984;8:21–7.

13.Berglund F, Flodh H, Lundborg P, Prame B, Sannerstedt R. Drug use during pregnancy and breast-feeding. A classification system for drug information. Acta Obstet Gynecol Scand Suppl 1984;126:1–55.

14.Onnis A, Grella P. The Biochemical Effects of Drugs in Pregnancy. Vol. 2. West Sussex, England: Ellis Horwood Limited, 1984:62–3.

15.Clark SL. Asthma in Pregnancy. National Asthma Education Program Working Group on Asthma and Pregnancy, National Institutes of Health, National Heart, Lung, and Blood Institute. Obstet Gynecol 1993;82:1036–40.

16.Report of the Working Group on Asthma and Pregnancy. Executive summary: management of asthma during pregnancy. J Allergy Clin Immunol 1994;93:139–62.

17.Koren G, Pastuszak A, Ito S. Drugs in pregnancy. N Engl J Med 1998;338:1128–37.

18.Einarson A, Lyszkiewicz D, Koren G. The safety of dextromethorphan in pregnancy. Results of a controlled study. Chest 2001;119:466–9.