Share this article on:

Do Women Residents Delay Childbearing Due to Perceived Career Threats?

Willett, Lisa L. MD; Wellons, Melissa F. MD; Hartig, Jason R. MD; Roenigk, Lindsey MD; Panda, Mukta MD; Dearinger, Angela T. MD; Allison, Jeroan MD; Houston, Thomas K. MD, MPH

doi: 10.1097/ACM.0b013e3181d2cb5b
Gender Issues

Purpose To assess gender differences among residents regarding their plans to have children during residency and determine the most influential reasons for these differences.

Method Using the Health Belief Model as a framework, the authors created an instrument to survey 424 residents from 11 residency programs at three academic medical institutions about their intentions to have children during residency. The authors developed a scale to assess the perceived career threats of having children during residency, evaluated its psychometric properties, and calculated the effect of the mediators.

Results The response rate was 77% (328/424). Forty-one percent of men versus 27% of women planned to have children during residency (P = .01). The instrument measured four career threats—extended training, loss of fellowship positions, pregnancy complications, and interference with career plans—on a five-point Likert scale. The scale had a Cronbach alpha of 0.84 and an eigenvalue of 2.2. Compared with men, women had higher scores for each item and a higher mean score (2.9 versus 2.1, P = .001), signifying greater belief in the potential of pregnancy to threaten careers. After adjusting for age, institution, postgraduate year, and knowledge of parental leave policies, women were less likely to plan to have children during residency (odds ratio 0.46 [95% confidence interval 0.25–0.84]). In mediation analysis, threats to career explained 67% of the gender variance.

Conclusions Women residents intentionally postpone pregnancy because of perceived threats to their careers. Medical educators should be aware of these findings when counseling female trainees.

Dr. Willett is associate professor of medicine and associate director of the internal medicine residency program, Division of General Internal Medicine, University of Alabama at Birmingham, Birmingham, Alabama.

Dr. Wellons is a fellow in endocrinology, Division of Endocrinology, Diabetes and Metabolism, University of Alabama at Birmingham, Birmingham, Alabama.

Dr. Hartig is assistant professor of medicine and director of the medicine–pediatrics residency program, Division of General Internal Medicine and Division of General Pediatrics, University of Alabama at Birmingham, Birmingham, Alabama.

Dr. Roenigk is a fellow in pulmonary–critical care, Division of Pulmonary and Critical Care, University of Alabama at Birmingham, Birmingham, Alabama.

Dr. Panda is professor of medicine and chair, Department of Internal Medicine, University of Tennessee, College of Medicine, Chattanooga, Chattanooga, Tennessee.

Dr. Dearinger is assistant professor, Division of General Internal Medicine, University of Kentucky, Lexington, Kentucky.

Dr. Allison is vice chair, Department of Quantitative Health Sciences, associate vice provost for health disparities, and professor of quantitative health sciences, University of Massachusetts Medical School, Worcester, Massachusetts.

Dr. Houston is scientist, Center for Health Quality, Outcomes & Economic Research (CHQOER), Bedford VAMC, Bedford, Massachusetts; and professor of quantitative health sciences and medicine chief, Division of Health Informatics and Implementation Science, and assistant dean for continuing medical education/medical education research, University of Massachusetts Medical School, Worcester, Massachusetts.

Please see the end of this article for information about the authors.

Correspondence should be addressed to Dr. Willett, BDB 341, 1530 Third Avenue South, Birmingham, AL 35294-0012; telephone: (205) 934-2494; e-mail:

Supplemental digital content is available for this article. Direct URL citations appear in the printed text; simply type the URL address into any Web browser to access this content. Clickable links to the material are provided in the HTML text and PDF of this article on the journal's Web site

The percentage and number of women in medical residencies is on the rise. In the past decade, an almost 10% increase in women residents has occurred. From 1998 to 2009, the percentage of women in U.S. residencies rose from 36% to 45%.1

As the number of women residents has increased, work duty hour limitations allow residents—both men and women—more personal time and, therefore, more time to devote to the role of parenting. But having a child during residency may have perceived negative career repercussions.3–6 Clinical duty coverage, board certification requirements, and possible extended training are work-related concerns that impact the decision to begin childbearing during residency.3,7,8 A leave of absence for parental leave is typically longer for women residents than men, so the decision to become pregnant during residency may be more difficult for women who, as a result of pregnancy, may have to extend their training and alter their career plans more than men would.3,4

Women must consider not only the career implications but also the biological timing of a pregnancy. In 2007, the average medical school graduate was 27 years old, and 15% of graduates were 30 or older.9 In observational studies, a slight decrease in female fertility occurs in the late 20s, and a marked decrease occurs after age 35.10 Thus, for the average female medical school graduate, residency training coincides with crucial childbearing years.7,11 Infertility and pregnancy complications are clearly associated with advancing maternal age,12–14 but evidence that residency training increases pregnancy complications is inconclusive.4,15,16 Therefore, postponing pregnancy until residency training is complete may result in pregnancy complications, unintentional childlessness, or having fewer children than desired.

Notably, highly educated women have the highest rates of permanent childlessness and are most at risk of involuntary childlessness related to delayed motherhood.17 In a classic report on the effect of female age on fertility, the percentage of non-contraceptive-using women who remained childless rose steadily according to the women's ages at marriage: 9% at age 25 to 29, 15% at age 30 to 34, 30% at age 35 to 39, and 64% at age 40 to 44.18 Hence, the risk of subfertility in women residents who delay childbearing until completion of training is considerable. If a woman resident delays childbearing during her residency (i.e., typically, in her late 20s or early 30s), she will be at high risk for subfertility and for having fewer children than she desires. She may be unknowingly sacrificing her goals regarding family for those of her career.

To better understand the factors that influence residents' decisions to have children, we surveyed a multiinstitutional group of men and women in multiple specialties and assessed their plans to have children during residency. Our goal was to assess gender differences and to understand the mediators for the differences. We hypothesized that women residents struggle with the decision to have children and would be more likely than men to intentionally postpone having children. We further hypothesized both that perceived career threats would be associated with postponing childbearing and that these threats would mediate the differences between men and women residents. In our review of the literature, we found no validated instrument to measure perceived career threats. Thus, as part of our study, we evaluated the psychometric properties of the scale we developed, which assesses the perceived career threats of having children during residency.

Back to Top | Article Outline


Study design, sample, and setting

We developed a cross-sectional survey and administered it to 424 residents from 11 residency programs at three academic medical institutions: the University of Tennessee, Chattanooga (UTC; 150 residents), the University of Alabama at Birmingham (UAB; 172 residents), and the University of Kentucky (UK; 102 residents) in 2006 and 2007. We included residents from internal medicine, family practice, pediatrics, medicine–pediatrics, surgery, and obstetrics–gynecology. All of the family practice, surgery, and obstetrics–gynecology residents were from UTC. All three institutions' institutional review boards granted exempt ethical approval, and resident participation was both anonymous and voluntary. This study was completely funded with internal sources.

Back to Top | Article Outline

Conceptual framework

We used the Health Belief Model (HBM)19,20 as a framework to conceptualize factors that might influence residents' plans to have children. The HBM was developed to explain an individual's behavior toward a health-related outcome by focusing on the beliefs of the individual. It has been used in a variety of medical behavioral programs to predict an individual's likelihood to action.19,21,22 According to the HBM, the likelihood of an individual taking action is dependent both on the individual's perceived susceptibility to the condition and on the influence of modifying factors. The perceived susceptibility is the amount of concern the individual has for the likelihood, or chance, of getting the condition. This is then modified by the individual's knowledge of the condition and his or her perception of threat (the negative impact or consequences of the condition). This is further influenced by cues to action (an internal or external event that increases one's motivation to do the behavior).19 For example, in a smoker who is contemplating cessation, his likelihood to quit depends on his perceived susceptibility for lung cancer modified by his knowledge of lung cancer and the consequences of possible death from it. This is further influenced by cues to action, such as developing hemoptysis or dyspnea, limited finances to buy cigarettes, or a family member encouraging cessation. We viewed planning to have children as a health behavior and felt this model was applicable to our study. We used a modified version of the HBM and applied the constructs (knowledge, threats, cues) within our survey questionnaire (Figure 1).

Figure 1

Figure 1

Back to Top | Article Outline

Survey development, content, and perceived threat scale

Two of us (L.L.W., M.F.W.) developed the survey, incorporating elements of fertility awareness and intention to have children during residency. Through a thorough literature search, we found one published, nonvalidated survey related to our interests.23 We adapted this survey, of nonmedical university students in Sweden, for our study. We piloted the survey with five volunteer residents at UAB and refined the survey for clarity. We then asked a convenience sample of faculty in the Division of Preventive Medicine in the Department of Medicine at UAB to review the survey for key content, and we further modified the survey through an iterative process. The final survey, available as supplemental digital content at, is a two-page document with 35 questions.

To assess perceived career threats, we developed a series of 21 questions related to reasons for deferring childbearing during residency. Because we could find no previously published, validated scale to assess perceived career threats, we developed such a scale and assessed its psychometric properties. Our five-point Likert scale measured four potential career threats. We asked residents to rate their levels of agreement (1 = completely disagree; 5 = completely agree) with the following reasons for personally postponing having children during residency: (1) extension of residency training, (2) loss of fellowship positions, (3) pregnancy complications, and (4) interference with career plans. One example statement is, “I would defer a pregnancy in myself or my spouse/partner during residency because I might have to extend my residency training.” We included pregnancy complications as a potential career threat because, in addition to the personal and emotional stress, bed rest or extended maternity leave can further delay completion of residency and/or impact career goals.

Back to Top | Article Outline

Main variables

Our main dependent variable was self-report of a plan to have children during residency as measured by the question: Do you (or female partner/spouse) plan on becoming pregnant during residency? Our main independent variable was gender.

Back to Top | Article Outline

Mediation analysis: Mediating variable and covariates

We applied mediation analysis to better understand the effect of gender on plans for having children during residency. Mediation analysis is a set of statistical techniques which quantifies the mechanisms through which observed effects operate.24 It accounts for the effect of mediating variables and confounders on outcomes. Mediating variables are behavioral, biological, psychological, or social constructs that transmit the effect of an independent variable to a dependent variable.24 From our modified HBM (Figure 1), the hypothesized mediating variable was perceived threats to career. We measured other modifying factors as covariates that might influence a resident's plans to have children including knowledge of subfertility (the ages at which a slight and a marked decrease in women's ability to become pregnant occur, and the ideal physiologic age for pregnancy) and cues to action (awareness of the relevant certifying board's requirements and of the training program's parental leave policies). We also collected the following data: each resident's age, postgraduate level of training, current parental status, institution, and training program.

Back to Top | Article Outline

Statistical analysis

We conducted all analyses using Stata SE, version 9 (College Station, Texas). First, we compared the proportion of men and women residents who reported plans to have children during residency.

Second, we analyzed the measures of perceived career threats. Using t tests, we assessed gender differences in scores for each of the four items on our career threats scale. We calculated Cronbach alpha to evaluate the internal consistency of the four items and then explored the factor structure using a principal component factor analysis with varimax rotation. We created a mean score for the perceived threat scale by summing the scores of the four items and dividing by four. We calculated the P for trend both for the association of increasing score on the perceived threats scale and for a plan to have children during residency using Mantel–Haenzel odds.

Third, we investigated the covariates from the HBM. We performed bivariate comparisons of gender with each covariate, and we assessed the bivariate association of each factor with the main dependent variable. We included those factors significantly associated with the independent or dependent variable in the mediation analyses.

Finally, we conducted the mediation analysis using principles from Baron and Kenny.25 We adjusted all regression models for age, current status as parents, postgraduate year, and institution. In the multivariable models, we limited the sample to married residents and those living with partners, because only four residents who were single planned to have children during residency. For each logistic model, we examined discrimination using the c-statistic (explained below). We standardized all regression coefficients before entering them into the mediation analysis. Standardization was necessary because the outcome for pathway c and c′ (explained below) was dichotomous, and we used logistic regression for two of the models.24

Figure 1 depicts the mediation model. Mediation analysis is accomplished using three regression models and decomposes the total effect of the independent variable into a direct component and an indirect, or mediated, component. The indirect component occurs through the mediated pathway (pathways a and b), and the direct effect (pathway c′) occurs independently from the mediated pathway. The analysis to estimate the mediation effect was based on the methods of MacKinnon and Dwyer24 and requires a series of two models: (1) a main effects model (pathway c), obtained by regressing the dependent variable on the main independent variable, and (2) a final mediation model, where the outcome is regressed on the main independent variable while adjusting for the mediator. From these regression models, the main effect (c) and direct effect (c′) were taken as the standardized coefficient for the main independent variable.24

The mediation effect represents the proportional influence the mediator (career threat) has on the effect (the plan for having children). The mediation effect was calculated as (c − c′)/c, and the precision of the mediated effect was assessed with 95% bias-corrected and accelerated confidence intervals (CIs) from bootstrapped resampling with replacement.26

Back to Top | Article Outline


The overall response rate was 328/424 (77%). The response rates for each school were as follows:

  • 77% (115/150) for UTC,
  • 85% (147/172) for UAB, and
  • 65% (66/102) for UK.

Fifty-seven percent of respondents were men (n = 187), and the response rates by specialty (308 total specialty responses) were as follows:

  • 47% from internal medicine (n = 145),
  • 21% from pediatrics (n = 66),
  • 10% from combined medicine–pediatrics (n = 32),
  • 5% from family medicine (n = 15),
  • 4% from obstetrics–gynecology (n = 11), and
  • 13% from surgical specialties (n = 39) (See Table 1).
Table 1

Table 1

We observed no differences by institution in gender or marital status; of those residents who identified their age, UTC had more residents over the age of 30 years (40%, 50/126) than did UAB and UK (10%, 16/155 and 20%, 14/69, P < .05). The mean number of children desired for each gender was 2.8.

Back to Top | Article Outline

Plan to have children during residency

Overall, 35% (115/328) of the residents reported a plan to have children during residency. Of 326 residents who identified their marital status, 51% (108/212) of the married residents, 14% (3/21) of the unmarried residents living with a partner, and 4% (4/93) of the single or divorced residents reported a plan to have children during residency. A significant gender difference existed in this result: 41% of men (77/188) versus 27% of female (38/140) residents reported a plan to have children during residency (P = .01).

Back to Top | Article Outline

Perceived career threat: Scale performance and factor analysis

Compared with men, women had higher scores on each of the four items of perceived threats, signifying a greater perception of threat (Table 2). The percentage of women who scored >3 on each items was higher than for men:

Table 2

Table 2

  • concern for extension of residency training, 47% of women (66/139) versus 19% of men (35/185),
  • loss of fellowship positions, 21% of women (29/139) versus 10% of men (19/184),
  • pregnancy complications, 38% of women (53/139) versus 9% of men (17/184), and
  • interference with career plans, 43% of women (59/138) versus 25% of men (45/183).

The four items had high internal consistency (Cronbach alpha = 0.84). In factor analysis, the four items loaded onto a single factor with an eigenvalue of 2.2 and factor loadings of 0.8 and above for each item (Table 2), demonstrating that the four items measure a similar construct of perceived threat. Women had higher mean scores on the perceived threat scale (2.9; standard deviation [SD] = 1.1) compared with men (2.1; SD = 1.0), P = .001 (pathway a, Figure 1). As the score on the perceived risk scale increased, the proportion of residents planning to have children during residency declined (Mantel–Haenzel odds for trend, P = .001) (see Figure 2 and pathway b, Figure 1).

Figure 2

Figure 2

Back to Top | Article Outline

Covariates from the HBM

Both genders overestimated the age of female fertility decline; women reported a younger age than men for the age of slight decline (women reported 31.9 years [SD = 4.1]; men reported 32.9 years [SD = 4.1]; P = .04) and marked decline (women reported 37.6 years [SD = 4.3]; men reported 38.8 years [SD = 4.1]; P = .03). Women reported an older ideal physiologic age to begin pregnancy than did men, although the difference was not significant (women reported 24.5 years [SD = 3.1]; men reported 23.8 years [SD = 3.5]; P = .08). Reported mean ages of fertility decline were similar for those planning to have children during residency and those not (P = .7)

One cue to action, awareness of the program's parental leave policy, was positively associated with planning to have children during residency. Those who were aware of their program's policy more frequently reported planning to have children during residency (41%, 85/207) than did those who were unaware of the policy (25%, 30/121), P = .003). The majority of residents (62%, 203/328) were aware of their training program's policies for parental leave, but only 33% (108/328) were aware of their certifying board requirements; there were no differences in either of these cues to action by gender.

Back to Top | Article Outline

Mediation analysis

Two covariates (knowledge of the age of marked female fertility decline and awareness of the training program's parental leave policies) were associated with the dependent and independent variables and were included in the mediation analysis. In the main effects multivariable logistic regression analysis (Table 3), women were again less likely to report a plan for having children during residency (pathway c) (adjusted odds ratio [OR] = 0.46 [95% CI 0.25–0.84]). After further adjustment for career threats, the association of gender and plan for having children during residency (pathway c′) was strongly attenuated (OR 0.74 [95% CI 0.37–1.5]) and no longer significant, P = .4.

Table 3

Table 3

Using standardized coefficients, we calculated the mediation effect ([c − c′]/c) as 67% (95% CI = 19%–100%), suggesting that the majority of the difference between men and women residents was due to the perceived threats to career.

Back to Top | Article Outline

Discussion and Conclusions

Despite improvements in cultural acceptance of women in medicine, improved work hours, and transparent parental leave policies,2–4,27,28 our study demonstrates that women residents plan to postpone having children during residency more than their male colleagues. Our evaluation tool was able to show that the majority of the gender difference (67%) was mediated by the perception of career threat for female residents. Women were most concerned with, of the four threats we studied, extending their residency training.

By postponing pregnancy until after residency is complete, female residents are risking subfertility.29 They are, in essence, placing the timing of their careers before the normal biologic timing of pregnancy. In a cohort of Swedish university students, half of the female students intended to have children after the age of 35 but were unaware of age-related fertility decline.23 In contrast, the women residents who participated in our study had more accurate knowledge of age-related fertility decline than their male counterparts, yet still planned to postpone having children. According to the HBM, people often underestimate their own susceptibility.19 Likewise, the female residents in our study seemed to underestimate their risks of subfertility and unintentional childlessness. We speculate that a perception of “it won't happen to me” may be contributing to this apparent conflict of rationale.

In addition to subfertility, advanced maternal age increases pregnancy complications.12–14 In a prospective multicenter analysis of over 36,000 women,12 women over 35 showed an increase in miscarriage (OR 2.0), chromosomal abnormalities (OR 4.0), congenital anomalies (OR 1.4), and cesarean delivery (OR 1.6). Complication rates were even higher for women over age 40. Women residents across specialties are concerned about pregnancy complications.4,15,16,29 In a recent national survey of medical residents, 85% of female respondents believed residency increased pregnancy complications,15 and in our study, 38% of women acknowledged this as a reason to defer childbearing during residency. Yet, studies show no strong association between residency training and pregnancy complications other than an increased risk of preterm labor.4,15,16 Therefore, women who choose to delay pregnancy beyond residency, and certainly beyond fellowship, are putting themselves at greater risk of complications than are women who choose to be pregnant during residency.

Previous studies report several career concerns that female residents have toward pregnancy,3–5,7,8 but no studies included a validated instrument that identifies and measures specific concerns. The perceived threat scale we developed had high internal consistency and demonstrated that the four items measure a similar construct of perceived threat. The scale meets criteria for mediation of the gender difference in plans for having children: Perceived threats differed significantly by gender, perceived threats were strongly associated with plans for having children during residency, and when perceived threats were introduced into the multivariable model, the prior strong association of gender and the outcome was fully attenuated and no longer significant. Researchers can use our scale in future investigations to further study this important topic.

The mediation analysis demonstrates that, although other covariates may be important, the perceived threat to one's career accounts for the majority of the gender differences of residents and is the major reason women residents plan to postpone pregnancy during residency. Of the career threats, women are most concerned about extended training. Interestingly, the concern for loss of fellowship had the lowest scores. Why are women so concerned about extending their residency? There are many potential explanations, including the personal desire to remain on schedule, perceived cultural pressure from peers or supervisors to return to clinical duties, and financial issues related to debt burden.4,15 In the 1980s, Sayres and colleagues30 found that 80% of institutions lacked a maternity leave policy and were unprepared for residents' pregnancies. In her review, Finch4 states, “the establishment of written, well-defined parental leave policies is important to support residents and their families.” The federal government,27 the Accreditation Council for Graduate Medical Education,28 and the American Board of Medical Specialties require written policies that detail requirements for parental leave and extended training time by specialty.3 Yet, we found that the existence of established parental leave policies is not a cue to action. Although a resident's awareness of parental leave policies was positively associated with planning children, not all residents were aware of the policies, and awareness of the policies did not attenuate the gender difference.

The career implications of childbearing on women in medicine do not end with residency training.31–35 In a recent survey of faculty at one academic medical center, Shollen and colleagues34 found significant gender differences in the perception of family-related obstacles. Significantly more women than men thought their parental leave resulted in additional clinical duties for their colleagues, and women indicated more family-related obstacles to their career success (e.g., lack of access to child care, inadequate parental leave policy, lack of part-time promotion tracks). In a survey of 142 medical school deans from the United States and Canada, family-friendly policies specifically designed to increase gender equity were available at fewer than 14% of institutions; only 3 of the 15 policies (i.e., benefits for part-time faculty, paid maternity leave, paid paternity leave) were offered at more than two-thirds of the schools.35 Therefore, waiting to have children until after residency may create even greater career obstacles for women physicians in academic medicine.

Our study has important limitations. The decision to have children is complex and affected by many additional factors beyond the scope of our study, such as finances, child care provisions, and partner desires. Our survey was cross-sectional, measured knowledge and intent—not outcomes—and may not have included all career-related concerns. Further, variations may exist across medical disciplines. For example, women residents in surgical specialties may have greater or differing concerns and may perceive less cultural acceptance than those in primary-care-based specialties (which are generally perceived as more family friendly). The training programs surveyed are all in the southern region of the United States; our results may not be representative of other regions. We did, however, accomplish broad representation. Our 11 programs and three institutions provided diversity in resident age, program size, and specialty training. Although only one of the academic institutions included surgical and obstetrics–gynecology residents, the patterns of associations (including gender and threats to career) were similar among subgroups of trainees by specialty (primary care versus surgical).

Our study is an important contribution to the literature. We demonstrate significant gender differences for residents in their plans for having children during residency. Women residents are intentionally postponing pregnancy because of perceived career threats—specifically, extended training. Our findings support the concepts of the HBM that perceived threats are the most powerful of the dimensions that determine likely action.36 Female residents should be cognizant of the choices and potential sacrifices they are making for their career or family goals. Those involved in medical education need to be aware of these important findings, especially when counseling female students and residents.

Back to Top | Article Outline



Back to Top | Article Outline

Other disclosures:


Back to Top | Article Outline

Ethical approval:

This study was approved by the institutional review board of the University of Alabama at Birmingham; the University of Tennessee, Chattanooga; and the University of Kentucky.

Back to Top | Article Outline

Previous presentations:

The University of Alabama at Birmingham results of this study were presented in a poster presentation at the Society of General Internal Medicine National Meeting; April 27, 2007; Toronto, Ontario, Canada.

Back to Top | Article Outline


1 Leadly J. Women in U.S. Academic Medicine: Statistics and Benchmarking Report 2008–2009. Washington, DC: Association of American Medical Colleges; 2009. Available at: Accessed January 4, 2010.
2 Accreditation Council for Graduate Medical Education. Duty Hours Language. Available at: Accessed January 4, 2010.
3 Jagsi R, Tarbell NJ, Weinstein DF. Becoming a doctor, starting a family—Leaves of absence from graduate medical education. N Engl J Med. 2007;357:1889–1891.
4 Finch SJ. Pregnancy during residency: A literature review. Acad Med. 2003;78:418–428.
5 Phalen ST. Pregnancy during residency. I. The decision ‘to be or not to be.’ Obstet Gynecol. 1988;72:425–431.
6 Gough HG, Hall WB. Number of children wanted and expected by American physicians. J Psychol. 1977;96:45–53.
7 van Dis J. Residency training and pregnancy. JAMA. 2004;291:636.
8 Rose SH, Burkle CM, Elliott BA, Koenig LF. The impact of parental leave on extending training and entering the board certification examination process: A specialty-based comparison. Mayo Clin Proc. 2006;81:1449–1453.
9 Association of American Medical Colleges. Division of Medical Education. 2007 Medical School Graduation Questionnaire All Schools Summary Report FINAL All Schools Report. Washington, DC: Association of American Medical Colleges; 2007. Available at: Accessed January 4, 2010.
10 Homad GF, Davies M, Norman R. The impact of lifestyle factors on reproductive performance in the general population and those undergoing infertility treatment: A review. Hum Reprod Update. 2007;13:209–223.
11 Verlander G. Female physicians: Balancing career and family. Acad Psychiatry. 2004;28:331–336.
12 Cleary-Goldman J, Malone FD, Vidaver J, et al. Impact of maternal age on obstetric outcome. Obstet Gynecol. 2005;105(5 pt 1):983–990.
13 Nybo Andersen AM, Wohlfahrt J, Christens P, Olsen J, Melbye M. Maternal age and fetal loss: Population based register linkage study. BMJ. 2000;320:1708–1712.
14 Odibo AO, Nelson D, Stamilio DM, Sehdev HM, Macones GA. Advanced maternal age is an independent risk factor for intrauterine growth restriction. Am J Perinatol. 2006;23:325–328.
15 Collier VU, McCue JD, Markus A, Smith L. Stress in medical residency: Status quo after a decade of reform? Ann Intern Med. 2002;136:384–390.
16 Klebanoff MA, Shiono PH, Rhoads GG. Outcomes of pregnancy in a national sample of resident physicians. N Engl J Med. 1990;323:1040–1045.
17 Sobotka T. In pursuit of higher education, do we postpone parenthood too long? Gend Med. 2006;3:183–186.
18 Menken J, Trussell J, Larsen U. Age and infertility. Science. 1986;233:1389–1394.
19 Redding CA, Rossi JS, Rossi SR, et al. Health behavior models. Int Electonic J Health Educ. 2000;3(special issue):180–193.
20 Glanz K, Lewis FM, Rimer BK, eds. Health Behavior and Health Education: Theory, Research, and Practice. 2nd ed. San Francisco, Calif: Jossey-Bass, Inc; 1997.
21 Sadler GR, Ko CM, Cohn JA, White M, Weldon RN, Wu P. Breast cancer knowledge, attitudes, and screening behaviors among African American women: The Black cosmetologists promoting health program. BMC Public Health. 2007;7:57.
22 Venmans LM, Gorter KJ, Hak E, Rutten GE. Short-term effects of an educational program on health-seeking behavior for infections in patients with type 2 diabetes: A randomized controlled intervention trial in primary care. Diabetes Care. 2008;31:402–407.
23 Lampic C, Svanberg AS, Karlström P, Tydén T. Fertility awareness, intentions concerning childbearing, and attitudes towards parenthood among female and male academics. Human Reprod. 2006;21:558–564.
24 MacKinnon DP, Dwyer JH. Estimating mediated effects in prevention studies. Eval Rev. 1993;17:144–158.
25 Baron RM, Kenny DA. The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. J Pers Soc Psychol. 1986;51:1173–1182.
26 Davison AC, Hinkley DV. Bootstrap Methods and Their Application. Cambridge, UK: Cambridge University Press; 1997.
27 United States Department of Labor. Available at: Accessed April 15, 2009.
28 Accreditation Council for Graduate Medical Education. ACGME Institutional Requirements. Available at: Accessed January 4, 2010.
29 The Practice Committee of the American Society for Reproductive Medicine. Definition of “infertility.” Fertil Steril. 2006;86(5 suppl):S228.
30 Sayres M, Wyshak G, Denterlein G, Apfel R, Shore E, Federman D. Pregnancy during residency. N Engl J Med. 1986;314:418–423.
31 Yedidia MJ, Bickel J. Why aren't there more women leaders in academic medicine? The views of clinical department chairs. Acad Med. 2001;76:453–465.
32 Hamel MB, Ingelfinger JR, Phimister E, Solomon CG. Women in academic medicine: Progress and challenges. N Engl J Med. 2006;355:310–312.
33 Carr PL, Ash AS, Friedman RH, et al. Relation of family responsibilities and gender to the productivity and career satisfaction of medical faculty. Ann Intern Med. 1998;129:532–538.
34 Shollen SL, Bland CJ, Finstad DA, Taylor AL. Organizational climate and family life: How these factors affect the status of women faculty at one medical school. Acad Med. 2009;84:87–94.
35 Dannels S, McLaughlin J, Gleason KA, McDade SA, Richman R, Morahan PS. Medical school deans' perceptions of organizational climate: Useful indicators for advancement of women faculty and evaluations of a leadership program's impact. Acad Med. 2009;84:67–79.
36 Janz NK, Becker MH. The Health Belief Model: A decade later. Health Educ Behav. 1984;11:1–47.
© 2010 Association of American Medical Colleges