Fetal Risk Summary

Nonoxynol-9 and octoxynol-9 are vaginal spermicides used to prevent conception. These agents, applied intravaginally, act by inactivating sperm after direct contact. Although human data are lacking, in animals nonoxynol-9 rapidly crosses the vaginal wall into the systemic circulation (1). Octoxynol should also be expected to act in a similar manner. The use of vaginal spermicides just before conception or inadvertently during the early stages of pregnancy has led to investigations of their effects on the fetus. The effects studied include congenital malformations, spontaneous abortions (SABs), low birth weight, stillbirth, sex ratio at birth, frequency of multiple births, and premature delivery.

A causal relationship between vaginal spermicides and congenital abnormalities was first tentatively proposed in a 1981 study comparing 763 spermicide users and 3,902 nonuser controls (2). The total number of infants with malformations was low: 17 (2.2%) in the exposed group and 39 (1.0%) in the nonexposed group. Malformations thought to be associated with spermicide use were limb reduction deformities (3 cases), neoplasms (2 cases), chromosomal abnormalities (Down's syndrome) (3 cases), and hypospadias (2 cases). An earlier investigation, published in 1977, had concluded there was no causal relationship between spermicides and congenital defects, although there was an increased incidence of limb reduction defects in infants of users (11 of 93) as compared with nonusers (8 of 186) (3). Three reports appeared in 1982 that suggested a possible relationship between spermicide use and congenital malformations (4,5 and 6). In a case-control study conducted by one of the co-authors of the 1981 investigation, a positive association with Down's syndrome was proposed when, in a group of 16 affected infants, 4 were from users of spermicides (4). In another case-control study, increased risk ratios after spermicide use, although not statistically significant, were reported for limb reduction defects (relative risk 2.00; six infants) and hypospadias (relative risk 4.00; eight infants) (5). Finally, an English study observed, among other defects, 2 cases each of hypospadias, limb reduction deformity, and Down's syndrome among infants of 1,103 spermicide users (6). The authors stated that their data were not conclusive, but the occurrence rates of these particular defects were higher than those observed in a comparative nonuser group.

Several criticisms have been directed at the original 1981 study (7,8,9 and 10). First, an infant was presumed exposed if the mother had a prescription filled at a designated pharmacy within 600 days of delivery. No attempt was made to ascertain actual use of the product or whether the mothers, either users or nonusers, had purchased a spermicide without a prescription (7,8 and 9). In a subsequent correspondence, all of the study's exposed cases of limb reduction deformity (3 cases), Down's syndrome (3 cases), and neoplasm (2 cases) were re-examined in terms of the exact timing of spermicide use (10). The data suggested that spermicides were not used near the time of conception in these cases. However, this does not eliminate the possibility that spermicides may act directly on the ovum before conception (11). Second, the four types of malformations lack a common cause and time of occurrence (9). Even a single type of defect, such as limb reduction deformity, has a varied origin (9). Third, the total number of infants with malformations was low (2.2% vs. 1.0%). Because these values are comparable to the 2%–5% reported incidence of major malformations in hospital-based studies, the apparent association may have been caused by a lower than expected rate of defects in the nonexposed group rather than an increase in the exposed infants (7,9). Fourth, no confounding variables other than maternal age were adjusted (9).

A number of investigators have been unable to reproduce the results published in 1981 (8,12,13,14,15,16,17,18 and 19). In a study examining 188 infants with chromosomal abnormalities or limb reduction defects, no relationship between periconceptional use of spermicides and these defects was observed (8). No association between spermicide use at conception and any congenital malformation was observed in a study comparing 1,427 cases with 3,001 controls (12). In a prospective study of 34,660 women controlled for age, time in pregnancy, concentration of spermicide used, and other confounding variables, the malformation rate of spermicide users was no greater than in users of other contraceptive methods (13,14 and 15). A cohort study, the Collaborative Perinatal Project involving 50,282 mother-child pairs, found no greater risk for limb reduction deformities, neoplasm, Down's syndrome, or hypospadias in children exposed in utero to spermicides (16). One group of investigators interviewed 12,440 women during delivery and found no relationship between the last contraceptive method used and congenital malformations (17). Spermicides were the last contraceptive method used by 3,891 (31%) of the women. A 1987 case-control study of infants with Down's syndrome (N=265), hypospadias (N=396), limb reduction defects (N=146), neoplasms (N=116), or neural tube defects (N=215) compared with 3,442 control infants with a wide variety of other defects was unable to establish any causal relationship to maternal spermicide use (18). The authors investigated spermicide usage at three different time intervals—preconceptional (1 month before to 1 month after the last menstrual period), first trimester (first 4 lunar months), and any use during lifetime—without producing a positive association. A similar study, involving 13,729 women who had produced 154 fetuses with trisomy, 98 with trisomy 21 (i.e., Down's syndrome), also failed to find any association with spermicides (19). In addition, a letter correspondence from one researcher argued that an association between vaginal spermicides, or any environmental risk factor for that matter, and trisomies was implausible based on an understanding of the origin of these defects (20).

An association between vaginal spermicides and SABs was found in five studies (21,22,23,24 and 25). A strong association was found among subjects who had obtained a spermicide within 12 weeks of conception (21). Another study demonstrated approximately twice the rate of SABs in women who continued to use spermicides after conception compared with users before or close to the time of conception (22). However, a 1985 critique concluded that both sets of investigators had seriously biased their results by failing to adjust for potentially confounding variables (9). In a study involving women aborting spontaneously before the 28th week of gestation and controls delivering after the 28th week, women who used spermicides at the time of conception demonstrated a 5-fold increase in chromosomal anomalies on karyotype examinations in 929 cases (23). Although no association between spermicide use and chromosomally normal abortion was found in a study involving 6,339 women, spermicide use of more than 1 year was more common in cases aborting trisomic conception than in controls (24). In an earlier report, the same authors observed an odds ratio of 4.8 for the association between abortuses with anomalies and unexposed controls (25). Two of these latter studies (23,25) did not adjust for confounding variables.

Three studies have found no association with SABs (6,14,26). No significant risk for SAB was observed in a large cohort study involving 17,032 subjects (6) or in another study examining periconceptional spermicide use (14). In a well-designed, large prospective study involving 32,123 subjects, spermicide use before conception was associated with a significant reduction in SAB during the 2nd trimester (26).

Three studies found no association between spermicide use and birth weight (6,14,22), but one study did find such an association (5). In this latter investigation, spermicide use after the last menstrual period was significantly associated with a lower mean birth weight among female infants of both smoking and nonsmoking mothers. For male births, an association with lower birth weight was found only when the mothers smoked. The authors were unable to determine whether these relationships were causal. Spermicide use before the last menstrual period had no effect on birth weight.

Under miscellaneous effects, a case-control study of 73 nontraumatic stillbirths found no relationship with the use of vaginal spermicides (27). No association between sex ratio at birth or frequency of multiple births and spermicides was found in one study (6). However, in a 1976 national survey, female births were approximately 25% higher among women using spermicides near the time of conception compared with nonusers, a statistically significant difference (22). Finally, a 1985 study found no evidence of a relationship between spermicide use and preterm delivery (14).

A 1990 Reference reported the meta-analysis of previous studies to determine whether maternal spermicide use is detrimental to the developing fetus (28). Negative associations were found between the periconceptual or postconceptual maternal use of spermicides and teratogenicity, SAB, stillbirth, reduced fetal weight, prematurity, or an increased incidence of female births.

In summary, the available evidence indicates that the use of vaginal spermicides, either before or during early pregnancy, does not pose a risk of congenital malformations to the fetus. Three authors of the original 1981 paper reporting a relationship between spermicides and congenital defects have commented that available data now argue against a causal association (29). In addition, the FDA has issued a statement that spermicides do not cause birth defects (30). There is also controversy on whether the 1981 study should have been published (31,32). The data for SABs, low birth weight, stillbirths, sex ratios at birth, frequency of multiple births, and premature delivery also indicate it is unlikely that these factors are influenced by spermicide use.

Breast Feeding Summary

Although human data are lacking, nonoxynol-9 is rapidly excreted into the milk of lactating rats (1). Similar excretion in humans should be expected for both nonoxynol-9 and octoxynol-9. If excretion does occur, the effect on the nursing infant is unknown.

References

1. Chvapil M, Eskelson CD, Stiffel V, Owen JA, Droegemueller W. Studies on nonoxynol-9. II. Intravaginal absorption, distribution, metabolism and excretion in rats and rabbits. Contraception 1980;22:325–39.
2. Jick H, Walker AM, Rothman KJ, Hunter JR, Holmes LB, Watkins RN, D'Ewart DC, Danford A, Madsen S. Vaginal spermicides and congenital disorders. JAMA 1981;245:1329–32.
3. Smith ESO, Dafoe CS, Miller JR, Banister P. An epidemiological study of congenital reduction deformities of the limbs. Br J Prev Soc Med 1977;31:39–41.
4. Rothman KJ. Spermicide use and Down's syndrome. Am J Public Health 1982;72:399–401.
5. Polednak AP, Janerich DT, Glebatis DM. Birth weight and birth defects in relation to maternal spermicide use. Teratology 1982;26:27–38.
6. Huggins G, Vessey M, Flavel R, Yeates D, McPherson K. Vaginal spermicides and outcome of pregnancy: findings in a large cohort study. Contraception 1982;25:219–30.
7. Oakley GP Jr. Spermicides and birth defects. JAMA 1982;247:2405.
8. Cordero JF, Layde PM. Vaginal spermicides, chromosomal abnormalities and limb reduction defects. Fam Plann Perspect 1983;15:16–8.
9. Bracken MB. Spermicidal contraceptives and poor reproductive outcomes: the epidemiologic evidence against an association. Am J Obstet Gynecol 1985;151:552–6.
10. Watkins RN. Vaginal spermicides and congenital disorders: the validity of a study. JAMA 1986;256:3095.
11. Jick H, Walker A, Rothman KJ. Vaginal spermicides and congenital disorders: the validity of a study—in reply. JAMA 1986;256:3095–6.
12. Bracken MB, Vita K. Frequency of non-hormonal contraception around conception and association with congenital malformations in offspring. Am J Epidemiol 1983;117:281–91.
13. Mills JL, Harley EE, Reed GF, Berendes HW. Are spermicides teratogenic? JAMA 1982;248:2148–51.
14. Mills JL, Reed GF, Nugent RP, Harley EE, Berendes HW. Are there adverse effects of periconceptional spermicide use? Fertil Steril 1985;43:442–6.
15. Harlap S, Shiono PH, Ramcharan S. Congenital abnormalities in the offspring of women who used oral and other contraceptives around the time of conception. Int J Fertil 1985;30:39–47.
16. Shapiro S, Slone D, Heinonen OP, Kaufman DW, Rosenberg L, Mitchell AA, Helmrich SP. Birth defects and vaginal spermicides. JAMA 1982;247:2381–4.
17. Linn S, Schoenbaum SC, Monson RR, Rosner B, Stubblefield PG, Ryan KJ. Lack of association between contraceptive usage and congenital malformations in offspring. Am J Obstet Gynecol 1983;147:923–8.
18. Louik C, Mitchell AA, Werler MM, Hanson JW, Shapiro S. Maternal exposure to spermicides in relation to certain birth defects. N Engl J Med 1987;317:474–8.
19. Warburton D, Neugut RH, Lustenberger A, Nicholas AG, Kline J. Lack of association between spermicide use and trisomy. N Engl J Med 1987;317:478–82.
20. Bracken MB. Vaginal spermicides and congenital disorders: study reassessed, not retracted. JAMA 1987;257:2919.
21. Jick H, Shiota K, Shepard TH, Hunter JR, Stergachis A, Madsen S, Porter JB. Vaginal spermicides and miscarriage seen primarily in the emergency room. Teratogenesis Carcinog Mutagen 1982;2:205–10.
22. Scholl TO, Sobel E, Tanfer K, Soefer EF, Saidman B. Effects of vaginal spermicides on pregnancy outcome. Fam Plann Perspect 1983;15:244,249–50.
23. Warburton D, Stein Z, Kline J, Strobino B. Environmental influences on rates of chromosome anomalies in spontaneous abortions (abstract). Am J Hum Genet 1980;32:92A.
24. Strobino B, Kline J, Lai A, Stein Z, Susser M, Warburton D. Vaginal spermicides and spontaneous abortion of known karyotype. Am J Epidemiol 1986;123:431–43.
25. Strobino B, Kline J, Stein Z, Susser M, Warburton D. Exposure to contraceptive creams, jellies and douches and their effect on the zygote (abstract). Am J Epidemiol 1980:112:434.
26. Harlap S, Shiono PH, Ramcharan S. Spontaneous foetal losses in women using different contraceptives around the time of conception. Int J Epidemiol 1980;9:49–56.
27. Porter JB, Hunter-Mitchell J, Jick H, Walker AM. Drugs and stillbirth. Am J Public Health 1986;76:1428–31.
28. Einarson TR, Koren G, Mattice D, Schechter-Tsafriri O. Maternal spermicide use and adverse reproductive outcome: a meta-analysis. Am J Obstet Gynecol 1990;162:655–60.
29. Jick H, Walker AM, Rothman KJ. The relation between vaginal spermicides and congenital disorders—in reply. JAMA 1987;258:2066.
30. Anonymous. Data do not support association between spermicides, birth defects. FDA Drug Bull 1986;16:21.
31. Mills JL. Reporting provocative results; can we publish “hot” papers without getting burned? JAMA 1987;258:3428–9.
32. Holmes LB. Vaginal spermicides and congenital disorders: the validity of a study—in reply. JAMA 1986;256:3096.