Effects of Cigarette Smoking on Menstrual Function and Ovulation

Gayle C. Windham, PhD
Reproductive Epidemiologist
California Department of Health Services
Oakland, CA

Introduction

Cigarette smoke is known to be toxic to the developing fetus during pregnancy. Infants of women who smoke during pregnancy typically have birth weights 150 grams less than those of non-smokers and twice the risk of low birth weight. Smoking has also been found to be associated with higher rates of infertility and less success with in-vitro fertilization (IVF). These epidemiologic findings are supported by animal studies, suggesting that smoking could exert an effect in a variety of ways from direct toxicity to interference with normal hormonal processes. Some studies have suggested increases in menstrual disorders from cigarette smoking, with an intermediate effect among ex-smokers. These studies found that smokers were more likely than non-smokers to experience menstrual cycle pain, skipped menses and other aspects of "abnormal" menstruation. Cigarette smoking also appears to have anti-estrogenic properties and several studies have reported lower levels of serum estrogens in smokers. An alteration of hormone function associated with smoking may be the cause of reported menstrual cycle perturbations, as well as infertility or sub-fecundity.

Study Description

To examine the effects of smoking on menstrual cycle function, we conducted a study using new techniques of biologic measurements, rather than relying on reporting of somewhat subjective symptoms. In population-based studies it is difficult to expect women to cooperate and give multiple blood samples in order to measure the variation in hormone levels that occurs during a normal menstrual cycle. This study is rare in the use of daily measures of urinary hormone metabolite levels to profile menstrual cycles for a fairly large group of women.

The development of laboratory assays for these hormone metabolites in urine allows us to study more women efficiently and cost-effectively, rather than relying on clinic populations that may be experiencing health problems. Women for this study were recruited from the membership of a large HMO in Northern California. We were interested in studying women of reproductive age who were generally healthy and able to become pregnant (e.g. no known reason they could not, such as sterilizing surgery, use of oral contraceptives, lack of a recent menstrual period, or trying to get pregnant unsuccessfully). If eligible, women were asked to collect a first morning urine sample every day and to store a small amount in their freezers, for up to 6 months. During the urine collection period, the 400 women who participated kept a daily diary about a number of symptoms and some habits, including daily smoking.

The samples were used to measure daily levels of metabolites of estrogen and progesterone that are excreted in urine to look at the pattern of hormone and menstrual cycle function. We categorized the information to look at cycle length and cycle variability. Based on the pattern of hormones, we estimated ovulatory status and the day of ovulation (release of the egg from the ovary). Using this date, the cycle was split into the follicular and luteal phases (approximately the first and second half, respectively). In an article published in Obstetrics and Gynecology (1999;93:59-65), we reported that heavy smokers who smoked an average of 20 or more cigarettes per day had shorter menstrual cycles (from the start of one bleed to the start of the next), than non-smokers, by an average of 2.6 days. Looked at another way, they were four times as likely to have short cycles (< 25 days) than non-smokers, even after taking other factors into account, like age, race, pregnancy history, and caffeine and alcohol intake. Most of this shortening appeared to occur during the follicular phase of the cycle, when an egg is maturing before ovulation. Nevertheless, these heavy smokers were also about three times as likely as non-smokers to have a short luteal phase (<11 days). The number of days of menstrual bleeding did not vary much by smoking status.

Looking at the variability of the cycle and phase lengths showed that women who smoked an average of a half-pack or more per day were 3-4 times as likely as non-smokers to have irregular cycle lengths (e.g. they varied from cycle to cycle). They also experienced more variable phase lengths and menses length, across their cycles. Smokers were more likely to have prolonged episodes of not ovulating than non-smokers, but the number of affected cycles were too few for making strong conclusions.

We also examined history of past smoking as reported on an initial interview. Interestingly, women who had either smoked for a fairly long time, or a larger amount, also had their average cycle length shortened, although not as dramatically as the current smokers. Similarly, they were about three times as likely as never smokers to have short cycles of less than 25 days long. There was also a suggestion of more anovulatory and irregular cycles among these past smokers, but again numbers were small for firm conclusions.

Other Studies and Conclusions

A few other studies have examined menstrual function in relation to smoking, but mostly related to menstrual cycle pain or other self-reported symptoms. A community survey in Los Angeles found that heavier smokers (> 15 cigarettes/day) were more likely to have menstrual disorders that led to consulting a doctor than non-smokers. A study that was conducted decades ago reported that women 30 or older who smoked were more likely to experience "irregular" menstrual periods than non-smokers, consistent with our findings. Similarly a mail survey in England found six of seven aspects of "abnormal" menstruation to be more frequently reported by smokers, including frequent (or short) and irregular cycles. But that study also found prolonged and heavy bleeding during periods among smokers, which is not consistent with our study. A more recent study of slightly older women (37-39) who also kept a menstrual diary found an increased likelihood of pain and greater amount of bleeding, as well as shorter length of bleeding, among smokers. We did not find a decrease in average menses length among smokers, nor have some other studies. That study also suggested that smokers may have had more variable cycle lengths, but not as strongly as ours. A very small study of hormone levels noted that heavy smokers had cycles on average 1.6 days shorter than non-smokers, with a mean follicular phase 1.4 days shorter, also consistent with our results. Differences in steroid hormone levels and metabolism among smokers have been suggested, which may explain some of the findings with menstrual cycle function. Further work is necessary to clarify the role of smoking in hormone regulation. We will examine urinary hormone metabolite levels from this data in a future report, as well as the potential effects of passive smoking.

The importance of these alterations in menstrual function on other health endpoints, such as fertility, were not examined in our study, but there may be several effects. A short luteal phase may indicate a progesterone response that is not adequate for implantation and maintenance of the fertilized egg; such a deficiency has been implicated as a cause of infertility, as well as recurring miscarriage. Women with variable cycle lengths may have difficulty trying to conceive as timing of ovulation will be less predictable. Anovulation has obvious relevance for fertility and how long it may take to get pregnant. Studies of women undergoing IVF suggest that smoking may have an effect on the development and viability of the immature egg during the follicular phase. These effects are consistent with evidence showing that smoking is associated with decreased fertility. Smoking may also accelerate the loss of ovarian function that occurs naturally with age. Shorter, or more frequent cycles, may contribute to this and lead to earlier menopause. Smokers have been shown to reach menopause on average two years earlier than non-smokers and early menopause may be associated with other health problems. Women with short cycles may also be at higher risk of breast cancer.

In conclusion, smoking appears to affect menstrual function, using the most objective data available thus far. Because of the limitations of most prior studies on this subject, these findings should be confirmed in other prospective studies. If confirmed, they would indicate effects of smoking on endpoints that affect many women and which have other ramifications for reproductive health, providing additional reasons for women to reduce or quit smoking.