During Prolonged Postnatal Abstinence, Risky Behavior Rises for Many African Men

D. Hollander

First published online:

In Côte d'Ivoire, where couples typically refrain from having intercourse for a year or more after a woman has given birth, men have elevated odds of engaging in risky sexual behavior during this period, according to analyses based on data from the 1994 Demographic and Health Survey (DHS).1 The odds that a monogamous married man will have extramarital sex, and that he will do so without using a condom, are about doubled if his wife is practicing postnatal abstinence. Prolonged postnatal abstinence is customary in much of West Africa, largely because of the belief that sperm poisons breast milk. The practice could be expected to provide protection against infection with sexually transmitted diseases; but as the analysts point out, this protective effect is "diluted" if men substitute unsafe sexual practices for conjugal relations after their wives give birth.

The analysis was motivated by results of an earlier study showing elevated odds of extramarital sex among husbands observing postnatal marital abstinence in Benin. The analysts sought to determine whether a similar pattern would be found elsewhere in the region; they studied Côte d'Ivoire because after a birth, married couples there abstain from intercourse for an average of 15 months (median, 12 months), and extramarital sex is common. Contraceptive prevalence is low in Côte d'Ivoire (15%, according to the DHS), and the most commonly used methods are the pill and periodic abstinence. An estimated 11% of adults were infected with HIV at the end of 1999.

A nationally representative sample of 8,099 women aged 15-49 and 2,552 men aged 15-59 participated in the DHS. To eliminate concerns that polygynous men might not differentiate between marital and extramarital partners, the analysts limited their sample to monogamous men whose wives were also interviewed for the survey. They excluded a small number of couples because the wife's report of postnatal abstinence was inconsistent with the husband's report of the timing of most recent marital intercourse, leaving a final sample size of 709.

According to the men's reports, two-thirds were younger than 40, and a similar proportion lived in rural areas. Nearly half had no education, and slightly more than half were farmers or agricultural workers. The sample was roughly evenly divided by level of household wealth. It also was about evenly divided among Christians, Muslims and adherents to traditional religions or no religion; two-thirds of the men reported the same religion as their wife. Most of the men lived in the country's Center or South region; one-fifth lived in the North. Eighteen percent were married to a woman who reported that she had observed postnatal abstinence at some point during the two months leading up to the survey. The analysts used logistic regression to assess the effect of each of these factors on men's extramarital sexual behavior.

When all other factors were controlled for, the odds of extramarital sexual activity were nearly twice as high among men whose wife reported postnatal abstinence as among others (odds ratio, 1.7). Men who lived in urban areas, residents of the Center region and those who reported a different religion than their wife also had significantly elevated odds of engaging in extramarital sex (odds ratios, 1.8-2.7).

Postnatal abstinence also was associated with a doubling of the odds that in the previous two months, a man had had extramarital sex and had not used a condom on the last occasion (odds ratio, 2.0). Notably, only one other factor was associated with this behavior: Residents of the South and Center regions had elevated odds of having engaged in unprotected extramarital sex (2.2-2.7).

In an analysis based only on men who reported having had extramarital sex, the odds of condom use at the last extramarital sexual encounter were significantly lower among Muslims than among Christians (odds ratio, 0.3). As the analysts note, the number of men in this subgroup--134--may have been too small to identify other significant associations.

The analysts observe that the findings for Côte d'Ivoire confirm those for Benin, and they suggest that "these two sets of results likely can be generalized to West Africa and perhaps to other countries on the continent." Assuming that HIV will continue to spread in the region, the researchers stress the importance of efforts to increase marital sex while decreasing extramarital sex. Such efforts could take place both at the program level, where health and family planning staff could encourage "earlier rather than later resumption of conjugal sex...together with contraceptive use to ensure adequate birth spacing," and at the policy level, where campaigns could be undertaken to change the widespread belief that sperm poisons breast milk.--D. Hollander


1. Ali MM and Cleland JG, The link between postnatal abstinence and extramarital sex in Côte d'Ivoire, Studies in Family Planning, 2001, 32(3):214-219.