Tobacco Documents Online

Life-Style Factors and Female Infertility ~ermaine M. Buck,z Lowell E. Sever,2 Ronald F_.. Batt,3,4 and Pauline Mendola1 We summarize the epidemiologic literature on the effect of life-style factors such as cigarette smoking, alcohol and caffeine consumption, physical exercise, body mass index, and drag use on female infertility. We identified relevant papers through MEDLINE, Index Medicus, and a manual review of reference lists. Risk factors that affect the risk of primary tubal infertility and that were corroborated in two or more studies include use of intrauterine devices (especially the Dalkon Shield) and cigarette smoking. We identified extremes in body size as a risk factor for primary ovulatory infertility. Cocaine, marijuana and alcohol use, exercise, caffeine consumption, and ever.u.se of thyroid medications were possible risk factors for various sub- types of primary infertility. Few risk factors have been assessed or identified for secondary infertility or other less common subtypes, such as cervical or endometriosis-related infertility. (Epidemiology 1997;8:435-441) Keywords: infertility, smoking, alcohol, drugs, exercise, contraception. Infertility affects approximately 2-3 million married couples in the United States3 Prevalence varies by the criteria used for operational definitionsz and choice of denominator. Although infertility is characterized by the absence of pregnancy (or a livebirth), its impact on women's health status, such as increased cancer risk,3-5 is much broader in scope, raising a number of public health concerns. Here, we review life-style factors for female infertility and identify avenues for further study. We excluded occupational factors, sexually transmitted diseases (STDs), and pelvic inflammatory disease (PID), because they were the subject of earlier reviews.6-8 We organized this paper by clinical subtype of infertility, rather than by exposure, to demonstrate the paucity of research on select subtypes and in recognition of their distinct eti- ologies. We searched the MEDLINE database, using infertility and the exposure variables as keywords, and reviewed Index Medicus and published reference lists. We selected only peer-reviewed papers published in English. We prepared this review according to the published guide- lines for epidemiologic review papers.9-12 From the Departments of ~Social and Preventive Medicine and 3Gynecology- Obstetrics, State University of New York at Buffalo, Buffalo, NY; ZBattelle/ Centers for Public Health Research and Evaluation, Seattle, WA; and 4Millard Fillmore Suburban Hospital, Buffalo, NY. Address correspondence to: Germaine M. Buck, Department of Social and Preventive Medicine, State University of New York at Buffalo, 270 Farber Hall, Buffalo, NY 14214. Submitted January 12, 1996; final version accepted December 29, 1996. © 1997 by Epidemiology Resources Inc. THZ$ ARTZCLE ZS FOR ZNDZVZDUAL USE ONLY AND NAY NOT BE FURTHER REPRODUCED OR ¢ STORED ELECTRONZCALLY ~ITHOUT NRZTTEN' PERNISSZON FRON THE COPYRZGHT , UNAUTNORZZED RE HOLDER 7 ZN FZNANOZAL u~R~OUOTZOH MAY RESULT A,,u uTHER PF-NALTZE~. ~ Terminology Infertility is typically defined as the absence of preg- nancy after 12 or more months of regular unprotected intercourse.13 This definition originally was intended to identify when couples should seek medical care,~4 but it has been varied to include a longer time period or restricted to denote the absence of tivebirths.15 Demography has distinct conceptual and method- ologic definitions for fecundity and fertility that contrast with the inconsistent use of these terms by epidemiolo- gists. Fecundity refers to the biological capacity to con- ceive, whereas fertility refers to the ability to reproduce or bear a liveborn offspring.16'z; As such, the absence of conception (infecundity) is only one of many dimen- sions of infertility. Types and Subtypes of Infertility Epidemiologic studies of infertility have addressed pri- mary infertility, or the inability to become pregnant among women who have never been pregnant, and secondary infertility, or the inability to become pregnant or carry a pregnancy to term among women who have previously been pregnant regardless of outcome.~S,~9 Both types of female infertility can be further classified by diagnostic subtype: tubal, ovulatory, uterine/peritoneal, cervical, other factor, and unexplained. The first three subtypes account for half of couple-based infertility.~s Male factor infertility is diagnosed in 40-50% of infer- tile couples; male factor alone accounts for 20-30% of couple-based infertility,z°,n Detection bias is a notewor- thy consideration for epidemiologic studies focusing on infertility, given that a large percentage of couples may have multiple ca.uses identified. The availability of di- agnostic technology can affect the identification and classification of infertile women.= 435

:7 436 BUCK ET AL Epidemiology July 1997, Volume 8 Number 4 Seeking Medical Care Selection bias is reported to threaten the validity of epidemiologic studies that rely exclusively on couples who seek medical services,z3-z5 Reasons commonly cited for women seeking medical care include more abundant reproductive services, greater public awareness and ac- ceptance, and fewer adoptable infants.26,z7 Educational attainment of >9 years, nulliparity, and younger age were associated with seeking medical care in two foreign studies.Z3.2s According to data from the National Survey of Family Growth (NSFG), 52% of women with primary infertility and 22% with secondary infertility reported seeking medical services,z4 These figures include women who were unable to become pregnant or carry a pregnancy to term. Women who sought care for primary infertility were more likely to have used contraception than women who did not seek care. Nonusers of care were older, married longer, and older at first intercourse than users of medical care. Women with secondary infertility who sought care were more likely than women who did not seek care to be white, to have higher incomes and educations, and to have used contraception,z4 The prevalence of medical care-seeking behavior ranged from 32% to 95% for women with primary in- fertility, and from 22% to 79% for women with second- ary infertility in three cross-sectional studies,z9 These ranges indicate that not all infertile couples seek care. The extent to which women who seek care differ from those who do not in relation to life-style factors is unknown, however. Prevalence The incidence of infertility is unknown; prevalence var- ies worldwide.3° The prevalence of "current" infertility ranges from 3.6% to 14.3% and "lifetime" infertility from 12.5% to 33.6%.29 In the United States, the per- centage of infertile women ages 15-44 years was re- ported to have decreased from 11.2% in 1965 to 7.9% in 1988.1 When surgically sterile women were removed from denominators, prevalence rates were 13.3% and 13.9%, respectively.13 Both race and age are associated with infertility; increasing prevalence rates for black and young (20-24 years) women have been reported.13a4'19 The percentage of infertile women ages 20-24 years increased from 4% in 1965 to 10% in 1982. Black women were 1.5 times more likely to report infertility than white women. Most prevalence data for the United States are derived from the National Fertility Study31 or Cycles II-IV of the National Survey of Family Growth.1,32 These cross-sectional surveys estimdte the prevalence of self-reported infertility among married" women and assume that current users of contraception are fecund. Age is an important factor for female infertility and is inversely related to fecundity)>3e Consideration of the underlying female age distribution is essential for com- paring rates.3°'37 Life-style factors that may account for variations in infertility rates include earlier age at first intercourse, delayed age at marriage or first birth, con- traceptive practices, and exposure to sexually transmit- ted diseases.3s-42 Several factors can increase prevalence estimates: in- clusion of all infertile women regardless of marital status, lifetime rather than current infertility, and exclusion of sexually inactive or sterile women from denominators.29 Factors that can decrease prevalence estimates entail: including women not at risk for pregnancy in denomi- nators, or restricting numerators to women who seek care, or restricting numerators to women who attempt pregnancy for more than 12 months. Life-Style Factors We listed studies that met our inclusion criteria in Table 1. Many of these studies stem from the two landmark case-control studies that assessed contraceptive practices and risk of female (tubal) infertility.43,44 These two early studies are similar with regard to their heterogeneous groups of infertile women who sought medical care and participated in personal interviews, but they differ by source of controls (birth certificates and hospitals, re- spectively). TUBAL INFERTILITY Daling et a143 conducted detailed in-home interviews with 159 women with primary tubal infertility and I59 parous control women. Cases were women residents of King County, WA, ages 20-39 years, who visited phy- sicians for infertility services between 1979 and 1981. Controls were women who had their first livebirth dur- ing the calendar year after the year in which cases reported attempting pregnancy. Controls were matched individually to cases on race, census tract of residence, and age (---5 years). They reported an increased risk of primary tubal infertility for women with a history of intrauterine device use [odds ratio (OR) = 2.6; 95% confidence interval (CI) = L3-5.2], especially if associ- ated with pelvic inflammatory disease (OR = 3.0; 95% CI = 1.2-7.3). Use of the Dalkon Shield conferred the highest risk [relative risk (RR) = 6.8; 95% CI = 1.8- 25.21. Cramer et a144 conducted a large mukicenter case- control study to assess past contraceptive use and subse- quent risk of tubal infertility. They identified cases from seven clinical centers in the United States and Canada between January 1981 and June 1983. A heterogeneous group of infertile women (N = 1,880) was identified, of which 283 women had primary and 69 had secondary tubal infertility. Women who had given birth at partic- ipating hospitals served as controls and were matched to cases on age, race, and payment status. The investigators used a life events calendar approach to collect contra- deprive information. Past intrauterine device use was associated with an increased risk of tubal infertility. The adjusted risk of tubal infertility associated with intrauter- ine device use before a livebirth was twofold (RR = 2.0; 95% CI = 1.5-2.6). Risk was greatest for users of the Dalkon Shield (RR = 3.3; 95% CI = 1.7-6.1) and

Epidemiology July 1997, Volume 8 Number 4 LIFE-STYLE AND FEMALE INFERTILITY 437 TABLE 1. Chronologic Summary of Epidemiologic Studies* Focusing on Life-Style Factors and Female Infertilityt Author, Type/Subtype of Life-Style Factor Year Design Infertility (Statistical Analysis) Life-Style Findings Daling et al,4~ 1985 • Population.based case-control I° tubal Prior use of IUD • 159 nulligravid cases with (conditional logistic tubal infertility regression) • 159 matched parous controls Cramer et al,~4 1985 Daling eta/,49 1985 Green et a/,54 1986 • Multicenter case-control 1° and 2° tubal Prior use of IUD • 283 cases with primary tubal (multivariate logistic infertility regression) • 69 cases with secondary tubal infertility • 3,833 parous controls • Population-based case-control (subset Daling et a143) • 127 women 2° tubal infertility and 395 parous controls • Population-based case-control (subset Daling et a143) • 187 1° ovulatory and 159 2° ovulatory infertility cases including those with other possible reasons for infertility • 187 controls for 1° infertility cases and 419 parous controls for 2° infertile cases Crameret a/,4s 1987 • Multicenter case-control 1° tubal • 283 nulliparous cases of tubal (primary) infertility • 3,833 parous controls (subset Cromer et 0./44) 2° tubal Induced abortion (logistic regression) 1° and 2° ovulatory Regular vigorous exercise and weight-for-height (1° conditional logistic regression, 2° unconditional logistic regression) Barrier methods and oral contraceptives (multivariate logistic regression) I. Ever- vs never-use of IUD (RR = 2.6; 95% CI = 1.3-5.2) 2. Ever-use of Dalkon Shield (RR = 6.8; 95% CI = 1.8-25.2) 1. Any IUD use relative to nonuse and 1° infertility (RR = 2.0; 95% CI = 1.5-2.6) 2. Use of Dalkon Shield and 1° infertility (RR = 3.3; 95% CI = 1.7- 6.1) Legal abortion did not increase risk of 2° tubal infertility (RR = 1.15; 95% CI = 0.70-1.89), even after :'2 abortions (RR = 1.29; 95% CI = 0.39-4.20) I. Exercise lasting ->60 minutes/day increased risk of i° ovulatory infertility (RR = 1.9; 90% CI = 0.6-5.1). Excluding women with tubal dysfunction (RR = 6.2; 90% CI = 1.00-39.8) 2. Little effect on risk of 2° ovulatory infertility (OR = 0.9; 90% CI = 0.2- 3.6) I. Ever-use of barrier methods decreased risk (RR = 0.6; 95% CI = 0.5-0.8) 2. Little effect for ever- use of oral contraceptives (RR = 1.2; 95% CI = 0.8-1.6) Phipps et al,46 1987 • Multicenter case-control 1° tubal, cervical, Cigarette smoking 1. • 901 infertile women ovulatory, endometriosis (multivariate logistic • 1,264 parous controls (subset regression) Cramer et a144) Current smoking increased risk of tubal (RR = 1.6; 95% CI = 1.1-2.2) and cervical infertility (RR = 1.7; 95% CI = 1.0- 2.7), respectively Table continues * Restricted to studies that classified infertility by type and diagnostic subtype. ~" OR = odds ratio; RR = relative risk; CI = confidence interval; IUD = intrauterine device; 1° = primary infertility; 2° = secondary infertility; BMI = body mass index.

438 BUCK ETAL Epidemiology July 1997, Volume 8 Number 4 TABLE 1. Continued Author, Type/Subtype of Life-Style Factor Year Design Infertility (Statistical Analysis) Life-Style Findings Green et al,5s 1988 Mueller et al,5° 1990 Orodstein et al,s~ 1993 Orodstein et al,ss 1993 • Population-based case-control (subset Daling et al4~) • 204 cases 1" ovulatory infertility and 172 2° ovulatory infertility • 204 parous controls for 1° and 461 parous controls for 2° infertile cases • Population-based case-control • (su~bset Daling et a143) • 150 nulligravid cases of ovulatory dysfunction and 84 nulligravid cases of tubal infertility • 150 parous controls for cases of ovulatory infertility and 84 parous controls for cases of tubal infertility • Multicenter case.control (subset Cramer er al~) • 1,050 white, primary infertile • 3,833 pamus white controls • Multicenter case-control (subset Cramer et a144) • 597 cases with I° ovulatory infertility • 3,833 parous white controls 1° and 2° ovulatory 1" tubal, ovulatory 1° tubal, ovulatory, cervical, endometriosis 1° ovulatory Body weight-for-height ( conditional logistic regression, 2° unconditional logistic regression) Recreational drag use (conditional logistic regression) Self-reported caffeine consumption (multivariate logistic regression) Self-reported prescription and nonprescription drag use (crude odds ratios, some logistic regression) 1. Body weight-for- height <-85% than ideal increased risk of 1° ovulatory infertility (RR = 4.7; 95% CI = 1.5- 14.7), also >120% of ideal weight (RR = 2.1; 95% CI = 1.0-4.3) 2. Slight effect of body weight <-85% on 2° ovulatory infertility (OR = 1.6; 95% C[ = O.7-3.4) 1. Marijuana use increased risk of ovulatory infertility (RR = 1.7; 95% CI = 1.0-3.0), especially use within year preceding attempted pregnancy (RR = 2.1; 95% CI = 1.1-4.0) 2. Cocaine use increased risk of tubal infertility (RR = 11.1; 95% CI = 1.7-70.8) 1. >7 gm of caffeine per month increased risk of tubal infertility (RR = 1.5; 95% CI = 1.1- 2.0), endometriosis- related infertility (RR = 1.6; 95% CI = 1.1-2.4), and cervical infertility (RR = 1.4; 95% CI = 0.9-2.3) I. Ever-use of thyroid replacement hormones (RR = 2.3; 95% CI = 1.5- 3.5), antidepressants (RR = 2.9; 95% CI = 0.9-8.3), asthma medication (RR = 1.7; 95% CI = 0.7- 3.5), and tranquilizers (RR =' 1.6; 95% CI = 0.7- 3.1) increased risk of ovulatory infertility Table continues 0 O~ O~ ~;

Epidemiology July 1997, Volume 8 Number 4 LIFE-STYLE AND FEMALE INFERTILITY 439 TABLE 1. Continued Author, Type/Subtype of Life-Style Factor Year Design Infertility (Statistical Analysis) Life-Style Findings i;7 Grodstein eta/,57 1994 • Multicenter case.control 1° ovulatory Self-reported BMI (crude !" "-, (subset Cramer et a/44) odds ratios and some • 597 cases with I° ovulatory multivariate logistic ., '~ infertility regression) f. • 1,695 l~rimiparous white controls Orodstein et al,s3 1994 • Multicenter case-control (subset Cromer et a/44) • 1,050 white cases with 1° infertility and 3,833 parous white controls 1° ovulatory, cervical, endometriosis, idiopathic Self-reported alcohol consumption (multivariate logistic regression) 1. BMI ---27 (obese) (RR = 3.1; 95% CI = 2.4-4.4) or BMI <17 (RR = 1.6; 95% CI = 0.7-3.9) increased risk of ovulatory infertility 2. High BMI greatest risk factor for women with ' polycystic ovary disease (RR = 6.0; 95% CI = 3.5-10.0) 1. Moderate alcohol use (-<100 gmper week) increased risk of endometriosis- related infertility (OR = 1.6; 95% CI == 1.1-2.3) 2. Heavy alcohol use (>100 gm per week) " increased risk of ovulatory infertility (OR = 1.6; 95% CI = 1.1-2.3) lowest for users of copper devices (RR = 1.6; 95% CI = 1.1-2.4). Intrauterine device use accompanied by a his- tory of infection conferred a threefold increase in risk (RR = 3~3; 95% CI = 1.8-6.1). Cramer et a/45 also evaluated the risk of tubal infertil- ity associated with past use of barrier methods or oral contraceptives. Users of barrier methods were at reduced risk of tubal infertility (RR = 0.6; 95% CI = 0.5-0.8). Past oral contraceptive use had no overall effect on tubal infertility, albeit a weak suggestion that risk increased directly with estrogen content. Phipps et a146 assessed the relative risk of primary infertility for current cigarette smokers according to four subtypes of infertility. Subjects included 901 women with primary infertility and 1,264 parous controls. Cur- rent cigarette smoking increased the risk of tubal (RR = 1.6; 95% CI = 1.1 = 2.1) and cervical infertility (RR = 1.5; 95% CI = 1.0-2.1). Comparable relative risks were observed when the analysis was restricted to women with a primary diagnosis of tubal (RR = 1.6; 95% CI = 1.1-2.2) or cervical (RR = 1.7; 95% CI = 1.0-2.7) infertility. Daling47 reported a similar increased risk for primary tubal infertility (RR = 2.7; 95% CI = 1.4-5.3) among smokers in comparison with nonsmokers. Olsen et al4s also reported that smoking increased the risk of primary (OR = 1.6; 95% CI = 1.1-2.2) and secondary (OR = 2.1; 95% CI = 1.3-3.6) subfecundity (defined as infer- tility of at least 1 year's duration). Daling et a149 evaluated the risk of secondary tubal infertility in relation to past history of an induced abor- tion. They observed only a slight elevation in risk (RR = 1.15; 95% CI = 0.70-1.89) among the 127 cases and 395 control women. Two other life-style factors have been linked to tubal. infertility. Mueller et al5° reported that recreational co- caine use increased the risk of primary tubal infertility (RR = 11.1; 95% CI = 1.7-70.8). Grodstein et alsl reported that caffeine consumption of >7 gm per month increased the risk of primary tubal infertility (RR = 1.5; 95% CI = 1.1-2.0). OVULATORY INFERTILITY Joesoef ct a~5z assessed the relation between self-reported caffeinated beverage consumption and primary infertil- ity in a study comprising 1,765 primiparous control women and 1,818 women with primary infertility. Total caffeine consumption of >7 gm per month.reflected little increase in the fecundability ratio (OR = 1.03; 95% CI = 0.92-1.16) after controlling for confounders, suggesting little effect on primary infertility. Similarly, Orodstein et al~1 observed that caffeine consumption of >7 gm per month had little effect on ovulatory infertil- ity in their case-control study (RR = 1.1; 95% CI = 0.9-1.4). Grodstein et al~3 analyzed heavy alcohol consumption and risk of ovulatory infertility. Cases were 1,050 women who sought care and were compared with 3,833 parous controls. Using average weekly intakes, they categorized women as moderate (>100 gm per week) or heavy (> 100 gm per week) drinkers. Heavy alcohol consump- tion increased the risk of ovulatory infertility (RR = 1.6; 95% CI = 1.1-2.3). Green et 0./54 assessed the role of physical exercise and risk of ovulatory infertility. They defined self-reported vigorous exercise as aerobic activity in excess of a spe- cific number of kilocalories per minute. Vigorous exer-

440 ~ BUCK ET AL cise for >--60 minutes per day was associated with an increased risk of primary ovulatory infertility (RR = 1.9; 90% CI = 0.6-5.1), but not secondary ovulatory infer- tility (RR = 0.9; 90% CI = 0.2-3.6). The effect of exercise did not appear to be confounded by body weight. Green et a155 analyzed extremes in body size in relation to ovulatory infecundity. They found a 4.7-fold (95% CI = 1.5-14.7) increase in risk of primary ovulatory infertility for women whose body weight-for-height was 85% or less than "ideal," using the Metropolitan Life Insurance tables26 They observed a smaller effect for women 120% or more over ideal weight (RR = 2.1; 95% CI = 1.0-4.3). Among women with secondary infertil- ity, body weight <-85% of ideal increased risk moder- ately (OR = 1.6; 95% CI = 0.7-3.4), and body weight >-120% of ideal (OR = 1.1; 95% CI = 0.7-1.8) slightly increased risk. Grodstein et a/57 reported that body mass indices of >-27 or <17 were associated with an increased risk of primary ovulatory infertility (RR = 3.1; 95% CI = 2.4-4.4 and RR = 1.6; 95% CI = 0.7-3.9, respectively). Mueller et also studied 150 women with primary ovu- latory dysfunction and 150 parous controls to assess the relation between recreational drug use and ovulatory infertility. History of marijuana use increased the risk of ovulatory infertility (RR = 1.7; 95% CI = 1.0-3.0); risk was greatest among women reporting recent marijuana use (RR = 2.1; 95% CI = 1.1-4.0). Grodstein et a158 assessed prescription and nonpre- scription drug use and primary ovulatory infertility. Among 597 women with primary ovulatory infertility and 3,833 parous controls, they found elevated relative risks for self-reported ever-use of thyroid hormones for >6 months (RR = 2.3; 95% CI = 1.5-3.5) or antide- pressants for >6 months (OR = 2.9; 95% CI = 0.9- 8.3 ).58 They found elevated risks for use of pain relievers (RR = 1.4), tranquilizers (RR = 1.6), and asthma drugs (RR = :.7). CERVICAL AND ENDOMETRIOSIS-RELATED INFERTILITY Few epidemiologic studies have attempted to assess risk factors specifically for cervical or endometriosis-related infertility. Phipps et a/46 reported that cigarette smoking increased the risk of cervical infertility (RR = 1.5; 95% CI = 1.0-2.1). Subsequently, Grodstein et al5~ reported that caffeine consumption of >7 gm per month slightly increased the risk of primary cervical infertility (RR = 1.4; 95% CI = 0.9-2.3). Grodstein et al~3 analyzed the relation between moderate alcohol use and cervical, endometriosis-related, and idiopathic infertility. Moder- ate alcohol (<-100 gm) use increased the risk of endo- metriosis-related infertility (OR = 1.6; 95% CI = 1.1- 2.3), even after adjusting for potential confounders. They observed a slight increase in risk for cervical in- fertility (OR = 1.3; 95% CI = 0.8-2.1) and a reduced risk for idiopathic infertility (OR = 0.9; 95% CI = 0.5-1.4). In a recent case-control study of endometriosis with two sets of controls (friend and medical), Darrow et a159 Epidemiology July 1997, Volume 8 Number 4 reported that cases were more likely than friend controls to report problems becoming pregnant (78% and 15%, respectively), despite similar intercourse and contracep- tive histories. The authors did not provide more specific information on infertility. Grodstein et al5~ identified caffeine consumption of >7 gm per month as a risk factor for endometriosis-related infertility (RR = 1.6; 9596 CI = 1.1-2.4). Conclusions The studies evaluated in this review provide evidence that life-style factors may increase the risk of various diagnostic subtypes of infertility. Much of the evidence stems from two population-based case-control studies of infertile women who sought medical services and were (presumably) accurately diagnosed and classified by type and subtype of infertility. Some risk factors were ob- served consistently across studies, despite differences in populations, control groups, and data collection proce- dures. Most published results address either tubal or ovula- tory infertility. Noticeably absent are studies of infertil- ity stemming from cervical problems, endometriosis, im- munologic causes, or other factors. The absence of such study may be one reason why life-style risk factors have been more clearly linked to tubal and ovulatory infertil- ity. Available data support cigarette smoking and past intrauterine device use as risk factors for tubal infertility and extremes in body size for ovulatory infertility. Unanswered questions pertaining to the role of life- style factors and female infertility include the timing and duration of exposures (lifetime or in the 12 months before attempting pregnancy) and the consistency of risk factors across subtypes of infertility. New epidemiologic studies of sufficient size and scope, encompassing all types and subtypes of infertility, would be instrumental in identifying' risk factors for specific diagnostic subtypes of infertility. Such research may help individuals alter or engage in behaviors that minimize the risk of infertility. We believe that additional descriptive work is needed, especially with respect to understudied aspects of infer- tility, such as whether or not the infertility is resolved, and previously unstudied groups of women (for example, unmarried women, Hispanic women). Motivational and behavioral factors that prompt women to seek care need to be examined to address concerns pertaining to selec- tion bias. Studies involving women diagnosed and treated more recently are also needed, given advances in reproductive endocrinology that may have implications for classifying women by type and subtype. Infertility is accompanied by numerous economic, psychological, and ethical considerations, all of which raise challenging public health concerns. References I. Mosher WD, Pratt WF. Fecundity and Infertility in the United States, 1965-88. Advance Data from Vital and Health Statistics. No. 192. Hyatts- ville, MD: National Center for Health Statistics, 1990. 2. Marchbanks PA, Peterson HB, Rubin GL, Wingo PA, the Cancer and o

. .4 Epidemiology July 1997, Volume 8 Number 4 Steroid Hormone Study Group. P~-.search on infertility: definition makes a difference. Am J Epidemiol 1989;130:259-267. 3. Ron E, Lunenfeld B, Menczer J, Blumsrein T, Katz L, Oelsner G, Serf D. Cancer incidence in a cohort of infertile women. Am J Epidemiol 1987;125: 780-790. 4. Brinton LA, Melton LJ 3rd, Malkasian GD Jr, Bund A, Hoover R. Cancer risk after evaluation for infertility. Am J Epidemiol 1989;129:712-722. 5. Risch HA, Marrett LD, Howe GR. Parity, contraception, infertility, and the risk of epithelial ovarian cancer. Am J Epidemiol 1994;140:585-597. 6. Rachootin P, Olsen J. The risk of infertility and delayed conception asso- ciated with exposures in the Danish workplace. J Occup Med 1983;25:394- 402. 7. Rosenberg MJ, Feldblum PJ, Marshall EG. Occupational influences on reproduction: a review of recent literature. J Occup Med 1987;29:584-59I. 8. Cates W Jr, Rolls RT Jr, Aral SO. Sexually transmitted diseases, pelvic inflammatory disease, and infertility: an epidemiologic update. Epidemiol Rev 1990;12:199-220. 9. Milne R, Chambers L. Assessing the scientific quality of review articles. J Epidemiol Community Health 1993;47:169-170. 10. Woolf SH. Review articles and disclosure of methods (Editorial). AmJ Prey Med 1991;7:53-54. 1 I. Hurchison BG. Critical appraisal of review articles. Can Fam Physician 1993;39:1097-1102. 12. Weed DL Methodologic guidelines for review papers. J Natl Cancer lnst 1997;89:6-7. 13. WesthoffCL The epidemiology of infertility, ha: Kiely M, ed. Reproductive and Perinatal Epidemiology. Boca Raton, FL: CRC Press, 1991;43-61. 14. K/lll~n B. Epidemiology of Human Reproduction. Boca Raton, FL: CRC Press, 1988;9-10. 15. Healy DL, Trounson AO, Andersen N. Female infertility: causes and treat- ment. Lancet 1994;343:1539-1544. 16. Wood JW. Dynamics of Human Reproduction: Biology,.Binmetry, Demog- raphy. New York: Aldine de Gruyter, 1994;3-22. 17. Gray R, Leridon H, Spira A. Biomedical and Demograhic Determinants of Reproduction. New York: Clarendon Press, 1993. 18. American Fertility Society. Infertility: an Overview. Consumer Protection Issues Involving h Vitro Fertilization Clinics. Hearing before the Subcom- mirtee on Regulation, Business Opportunities, and Energy of the Committee on Small Business, House of Representatives. lOlst Congress, F~t Session, Washington DC, March 9, 1989. Washington DC: U.S. Government Print- ing Office, 1989;74-81. 19. National Center for Health Statistics. Fecundity, Infertility, and Reproduc- tive Health in the United States, 1982. Vital and Health Statistics, Series 23: Data from the National Survey of Family Growth, No. 14. DHHS Pub. No. (PHS)87-1990. Washington DC: National Center for Health Statistics, 1987. 20. Zinaman MJ, Uhler ML. Clinical and laboratory assessment of male infer- tility. Female Patient 1996;21:35-47. 21. Howards SS. Treatment of male infertility. N EnglJ Med 1995;332:312-317. 22. Belsey MA. Infertility: prevalence, etiology, and natural history. In: Bracken MB, ed. Perinatal Epidemiology. New York: Oxford University Press, 1984; 255-282. 23. Rachootin P, Olsen J. Social selection in seeking medical care for reduced fecundity among women in Denmark. J Epidemiol Community Health 1981 ;35:262-264. 24. High MB, Mosher WD. Characteristics of infertile women in the United States and their use of infertility services. Fertil Steril 1987;47:618-625. 25. Schmidt L, Miinster K, Helm P. Infertility and the seeking of infertility treatment in a representative population. BMJ 1995;102:978-984. 26. Aml SO, Cates W Jr. The increasing concern with infertility: why now? JAMA 1983;250:2327-2331. 27. Daniluk JC. Infertility: intrapersonal and interpersonal impact. Ferdl Steril 1988;49:982-990. 28. Templeton A, Fraser C, Thompson B. Infertiliry--epidemiology and referral practice. Hum Reprod 1991;6:1391-1394. 29. Schmidt L, MOnster IC Infertility, involuntary infecundity, and the seeking of medical advice in industrialized countries 1970-t992: a review of con- cepts, measurements and results. Hum Reprod 1995;10:1407-1418. 30. Belsey MA. Infertility: prevalence, etiology, and natural history, ha: Bracken MB, ed. Perinatal Epidemiology. New York: Oxford University Press, 1984; 255-282. 31. Ryder NB, WestoffCF. Reproduction in the United States: 1965. Princeton, NJ: Princeton University Press, 1971. LIFE-STYLE AND FEMALE INFERTILITY 441 32. Mosher WD. Reproductive Impairments among Currently Married Couples: • United States, 1976. Advance Data from Vital and Health Statistics. Hyatts- ville, MD: National Centers for Health Statistics, 1980;55. 33. Gray RH, Doyle PE. The epidemiology of conception and fertility. In: Barron SL, Thomson AM, eds. Obstetrical Epidemiology. New York: Aca- demic Press, 1983;25-60. 34. Menken J, Tmssell J, Larsen U. Age and infertility. Science 1986;233:1389- 1394. 35. Schwartz D, Mayaux MJ. Female fecundity as a function of age. N Engl J Med 1982;306:404-406. 36. American Society for Reproductive Medicine. Age Related Infertility: Guideline for Practice. Birmingham, AL: American Society for Reproduc- tive Medicine, 1995. 37. Coale AJ, Trussell TJ. International differences in fertility. Popul Index 1974;40:185-258. 38. Battaglia AR, Graziano MR, Scafidi Fonti MG. Experimental research into the changes in the way sexuality is experienced by infertile women. Acra Eur Fertil 1983;14:67-73. 39. Khatamee MA. Infertility: a preventable epidemic, hat J Fertil 1988;33:246- 251. 40. Musher WD, Aral SO. Factors related to infertility in the United States, 1965-76. Sex Tmnsm Dis 1985;12:117-123. 41. Aral SO, Musher WD, Cares WJ Jr. Self-reported pelvic inflammatory disease in the U.S.: a common occurrence. Am J Public Health 1985;75: 1216-1218. 42. Baldwin WH, Nord CW. Delayed Childbearing in the U.S.: Facts and Fictions. Popul Bull 39(4). Washington DC: Population Reference Bureau, 1984. 43. Daling JR, Weiss NS, Metch BJ, Soderstmm RM, Moore DE, Spadoni LR, Stadel BV. Primary tubal infertility in relation to the use of an intrauterine device. N Engl J Med 1985;312:937-947. 44. Cramer DW, Schiff I, Schoenbaum SC, Gibson M, Belisle S, Albrecht B, Stillman RJ, Berger MJ. Tubal infertility and the intrauterine device. N Engl J Med 1985;312:941-947. 45. Cramer DW, Goldman MB, Schiff I, Belisle S. Albrecht B, Stadel B, Gibson M, Wilson E, Stillman R, Thompson I. The relationship of tubal infertility to barrier method and oral contraceptive use. JAMA 1987;257:2446-2450. 46. Phipps WR, Cramer DW, Schiff I, Belisle S, Stillman R, Albrecht B, Gibson M, Berger MJ, Wilson E. The association between smoking and female infertility as influenced by cause of the infertility. Fertil Steril 1987;48:377- 382. 47. Daling JR. Cigarette smoking and primary tubal infertility. In: Rosenberg MJ, ed. Smoking and Reproductive Health. Litrleton, MA: PSG Publishing, 1987. 48. Olsen J, Rachootin P, Schiodt AV, Damsbo N. Tobacco use, alcohol consumption and infertility. Int J Epidemiol 1983;12:179-184. 49. Daling JR, Weiss NS, Voigt L, Spadoni LR, Sodetstrnm R, Moore DE, Stadel BV. Tubal infertility in relation to prior induced abortion. Ferril Steril 1985;43".389-394. 50. Mueller BA, Daling JR, Weiss NS, Moore DE. Recreational drug use and the risk of primary infertility. Epidemiology 1990;1:195-200. 51. Grodstein F, Goldman MB, Ryan L, Cramer DW. Relation of female infertility to consumption of caffeinated beverages. Am J Epidemiol 1993; 137:1353-1360. 52. Joesoef MR, Beral V, Rolfs RT, Aral SO, Cramer DW. Are caffeinated beverages risk factors for delayed conception! Lancet 1990;335:136-137. 53. Grodstein F, Goldman MB, Cramer DW. Infertility in women and moderate alcohol use. Am J Public Health 1994;84:1429-1432. 54. Green BB, Daling JR, Weiss NS, Lift JM, Koepsell T. Exercise as a risk factor for infertility with ovulatory dysfunction. Am J Public Health 1986;76:14: 32-36. 55. Green BB, Weiss NS, Daling JR. Risk of ovulatory infertility in relation to body weight. Fertil Steril 1988;50:721-726". 56. Weights of Insured Persons in the United States with Lowest Mortality. Metrop Life CO Insur Star Bull 1959;40:3. 57. Grodstein F, Goldman MB, Cramer DW. Body mass index and'ovulatory infertility. Epidemiology 1994;5:247-250. 58. Grodsrein F, Goldman MB, Ryan L, Cramer DW. Self-reported use of pharmaceuticals and primary ovulatory infertility. Epidemiology 1993;4: 151-156. 59. Darrow SL, Selman S, Batt RE, Zielezny MA, Vena JE. Sexual activity, conrmception, and reproductive factors in predicting endometriosis. Am J Epidemiol 1994;140:500-509. 0