Intergenerational predisposition to operative delivery.

OBJECTIVE: To determine the risk of cesarean delivery for women who themselves were born via operative delivery. METHODS: A linked data base was constructed between the birth certificates of individuals born in Utah during 1947-1957 (parental cohort) and who subsequently became a parent of offspring born in Utah between 1970-1991 (offspring cohort). Parental cohort women (cases) who had been delivered operatively (cesarean delivery, mid- or high forceps) as well as women who had a sibling delivered by an operative procedure were matched (1:2) with parental-cohort women born by spontaneous vaginal delivery (controls). Both cases and controls were selected based on having a record of at least one delivery in Utah during 1970-1991. RESULTS: Women who were delivered by cesarean were at increased risk of subsequently delivering their children by cesarean (odds ratio [OR] 1.41, 95% confidence interval [CI] 1.18-1.70; P < .001). Progressive risk was associated with parental delivery by mid- or high forceps (OR 1.72, 95% CI 1.20-2.47; P = .004), parental cesarean because of cephalopelvic disproportion alone (OR 1.83, 95% CI 1.16-2.88; P = .01), or parental cesarean for dysfunctional labor (OR 5.97, 95% CI 1.5-23.6; P < .001). The attributable risk for cesarean delivery to the contemporary population is 3.5%. CONCLUSION: An intergenerational predisposition to cesarean delivery exists.

B TETRI ~~ E Volume 87 June 1996 Number 6 The Intergenerational Predisposition to Operative Delivery MICHAEL w. VARNER, MD, ALlSON M. FRASER, CHERI Y. HUNTER, PATRICE S. CORNELl, AND RYKH. WARD,PhD Objective: To determine the risk of cesarean delivery for women who themselves were born via operative delivery. Methods: A linked data base was constructed between the birth certificates of individuals born in Utah during 1947- 1957 (parental cohort) and who subsequently became a parent of offspring born in Utah between 1970-1991 (off­spring cohort). Parental cohort women (cases) who had been delivered operatively (cesarean delivery, mid- or high for­ceps) as well as women who had a sibling delivered by an operative procedure were matched (1:2) with parental-cohort women born by spontaneous vaginal delivery (controls). Both cases and controls were selected based on having a record of at least one delivery in Utah during 1970-1991. Results: Women who were delivered by cesarean were at increased risk of subsequently delivering their children by cesarean (odds ratio [OR] 1.41, 95% confidence interval [CI] 1.18-1.70; P < .001). Progressive risk was associated with parental delivery by mid- or high forceps (OR 1.72, 95% CI 1.20-2.47; P = .004), parental cesarean because of cephalo­pelvic disproportion alone (OR 1.83, 95% CI 1.16-2.88; P ::: .01), or parental cesarean for dysfunctional labor (OR 5.97, 95% CI 1.5-23.6; P < .001). The attributable risk for cesarean delivery to the contemporary population is 3.5%. Frol1l tllc Departmcnts of Obstetrics lIud Gynecology and Human Genetics, Univcrsifll of Utall School of Medicine, Salt Lake City, Utah. Supported ill part blf Natiollal Institlltes of Health grant 1 R01 HD23492 (RHW) and II generalis gift from the Geary fiwlily. VOL. 87, NO.6, )U1':E 19% Conclusion: An intergenerational predisposition to cesar­ean delivery exists. (Obstet Gynecol1996;87:905-11) Cesarean delivery rates have consistently increased in the past decades in the United States,l The most com­mon indication for primary cesarean in this country has been dystocia, a term that includes failure to progress in labor (prolonged or dysfunctional labor) as well as cephalopelvic disproportion. This has been attributed to a more cautious approach to the management of dys­tocia, increased utilization of regional anesthetics, and increasing medical indications for induction of labor. In addition, it is well recognized that these factors are often interrelated. Another possible explanation for this increase in the primary cesarean rate may be a familial tendency toward inefficient labor. Previously, familial associa­tions have been identified for numerous other obstetric and gynecologic conditions, including preeclampsia,2 endometriosis,3 postdates pregnancy,4 low birth weight,S and ovarian cancer.6 In addition, social, cultural, or economic pressures may also result in familial cluster­ing. If such a familial tendency toward inefficient labor 0029-7844/96/$15.00 905 PH 50029-7844(96)00064-6 exists, then improved maternal and perinatal survival rates combined with more liberal indications for cesar­ean delivery and a less aggressive approach to opera­tive vaginal delivery might be an explanation for the recent increase in dystocia-related cesarean delivery rates. Thus, the purpose of this study was to test the null hypothesis that there is no difference in the risk of cesarean delivery for women who were born via oper­ative delivery relative to women themselves born via spontaneous vaginal delivery. The alternative hypoth­esis was that a woman who was born via an operative delivery is at an increased risk of requiring an operative delivery. Materials and Methods A linked data base of Utah birth certificates was estab­lished that encompasses two distinct cohorts: an off­spring cohort comprised of births occurring between 1970-1991 and a parental cohort comprised of births occurring between 1947-1957. This protocol was ap­proved by the University of Utah Institutional Review Board in agreement with the Utah State Department of Health. The offspring cohort initially consisted of all live births occurring between 1970-1986. The original suit­ability of these records for linkage was compromised because the middle name information and state of birth were not consistently available. As a result, all birth certificates were manually keyed from microfilm records. This original offspring cohort was then extended to include computerized birth certificates for 1987-1991. These latter data were merged with the original off­spring cohort data, maintaining a similar data format across the entire offspring cohort time period. The parental cohort was defined as all birth certifi­cates retrievable from microfilm in the state of Utah for the period 1947-1957. By confirmation with the original Utah State Department of Health copies, over 90% of the 1947-1954 birth records from this period are in­cluded. The microfilm birth certificates from 1955-1957 con­tained only index information, and the bottom portion of the record containing medical information relating to that birth was not recorded. Although 92% of the 1957 birth information was eventually recovered and en­tered, only about 27% of the 1955-1956 data could be recovered. Therefore, the 1955 and 1956 births were not included in the current data base. Thus, the parental cohort consists of 197,466 birth records. For the time period 1947-1957, 99,801 mother­offspring links were formed by linking the birth certif­icates of mothers to the birth certificates of one or more 906 Varner et al Heredity and Cesarean of their children. After adjusting for mISSIng name information in the parental cohort, the current data base of mother-offspring links represents approximately 58% of all possible links between mothers and children for this time period. A one-to-two case-to-control cohort design was se­lected to control for those demographic variables re­corded on the Utah birth certificate known to be asso­ciated with operative delivery and also to maximize statistical power. Using power analysis procedures for case-control studies,7 we determined that 500 cases with 1000 controls would provide 73% power to detect an odds ratio (OR) of 1.5 and 99% power to detect an OR of 2.0. During the 45-year study period, seven different birth certificate forms have been used by the Utah State Department of Health. The format used from 1989 onward contains a set of 19 specific criteria for medical risk factors for the current pregnancy, nine attributes that define the method of delivery, and 20 attributes that identify complications of labor and! or delivery. To create a uniform data system, these specific criteria were applied to the entire study period as follows. For birth certificates dated 1947-1957, information on pregnancy and labor complications as well as type of delivery was initially recorded as free text on the original document. Each birth certificate was reviewed and coded using the 1989-1993 birth certificate format. For birth certificates dated 1970-1977, the method of delivery was the only computerized obstetric informa­tion available during the data base development. Con­sequently, the birth certificates of offspring of all iden­tified cases and controls during this period were reviewed from original records and coded by the 1989- 1993 criteria. From 1978 through 1988, medical information con­cerning complications was available as a four-digit ICD code, whereas the method of delivery was coded into one of five categories (spontaneous, low forceps, mid­forceps, primary cesarean, repeat cesarean). These codes and records were reviewed individually and entered into the data base. In all cases, the diagnoses were restricted to those that were recorded on the Utah birth certificate by the attending physician. Patients selected as "cases" were female children of the parental cohort who were known to have had at least one live birth in the offspring cohort. All cases were either delivered by cesarean, mid forceps, or high forceps or had sisters who were delivered by cesarean, midforceps, or high forceps. The classifications as mid­forceps or high forceps were taken directly from the individual birth certificates. Patients selected as "controls" were female children of the parental cohort who were known to have had at Obstetrics & Gynecology least one live birth in the offspring cohort and who delivered via spontaneous vaginal or low forceps deliv­ery. In addition, women selected as controls had no record of any sisters who were delivered by cesarean, midforceps or high forceps. Entry criteria were the following: 1) either the present birth or a previous birth was a cesarean; 2) there was a sibling born by cesarean; 3) the birth was by mid- or high forceps from a cephalic presentation, and 4) there was a sibling of a mid- or high forceps birth from a cephalic presentation. To avoid duplication of cases, available birth certificate records were assigned to cat­egories in a stepwise fashion, beginning with entry criterion 1. The diagnosis of cephalopelvic disproportion was limited to pregnancies with vertex presentation. Dysto­cia was not a coded indication for operative delivery in either cohort. In an attempt to apply current practice standards to this era, any cesarean performed for ceph­alopelvic disproportion or prolonged labor or dysfunc­tional labor was included in a category subsequently referred to as dystocia. There were no exclusion criteria. Female infants from the parental cohort who were delivered by spontaneous vaginal delivery and who subsequently had at least one delivery in Utah during 1970-1991 were identified as controls on a one-to-two basis as outlined previously. Control patients were matched by birth certificate data for maternal age, county in which the delivery occurred, multiplicity, and parity. Entry Criterion (Cases) Parental Cohort (1947-1957) Offspring Cohort (1970·1991 2 4 Mothers Offspring Matched Controls (2:1) 1 2 3 4 Figure 1. Schematic diagram of the study design. The entry criteria are: 1) mother delivered by cesarean, 2) maternal sibling with cesarean, 3) mother delivered by mid- or high forceps, 4) maternal sibling delivered by mid- or high forceps. The dark portion of the four entry criterion groups (method of maternal deliverv) represent the percent­age of each group that fulfilled the definition of "cases." The dark portions under matched controls represent the 2:1 matched "controls." Both cases and controls were then linked to their respective pregnan­cies in the offspring cohort. VOL. 87, NO.6, JUJ\;E 1')96 Table 1. Distribution of Cases and Controls Entry criterion 1) Cesarean delivery Cases Controls 2) Sibling born via cesarean delivery Total eligible birth certificates* 4010 Cases 1661 Controls 3) Mid- or high forceps Cases 944 Controls 4) Sibling born via mid- or high forceps Cases Controls 8% Linked female birth certificates* (mothers) 509 1018 235 470 114 228 131 262 * Born during parental cohort (1947-1957). I Born during offspring cohort (1970-1991). Linked birth certificates t (offspring) 1708 3461 859 1642 387 768 440 838 Figure 1 represents the development of the parental and offspring cohort populations, the selection of case and control mothers (parental cohort), and the linkage to case and control offspring (offspring cohort). The method of delivery of the offspring mothers was cross-classified by that of the parental mothers. Log­linear analysis with likelihood ratio test statistics8 was used to determine the significance of association be­tween delivery methods of the two generations.9 Odds ratios and 95% confidence intervals (Cl) were calculated from the estimated model parameters and the standard errors. The same method was used to determine sepa­rate ORs for the offspring mothers who had mothers with cephalopelvic disproportion and those whose mothers did not have cephalopelvic disproportion. Three-way log-linear analysis was used to determine whether the ORs differed between the two groups. Likelihood ratio test statistics and parameter estimates were generated using generalized linear interactive modeling.'! Results There were 5671 birth certificates (female and male newborns) which satisfied one of the first two criteria (ie, born by cesarean or had a sibling born by cesarean). These included 4010 cesarean births (entry criterion 1) and 1661 who had a sibling delivered by cesarean but were not themselves delivered by cesarean (entry crite­rion 2). The 5671 newborns were delivered from 3226 women (grandmothers). There were 1840 birth certificates (female and male Varner et al Heredity and Cesarean 907 Percent Cesarean Deliveries 20,----------------------------------------, 15~--------------------------~ 10~------------------ 5~------------------ Calendar Year Figure 2. Schematic diagram of cesarean section rates during the course of the study. Dark bars = primary cesarean rates; opm bars = repeat cesarean rates. newborns) that satisfied one of the last two criteria, ie, born by mid- or high forceps or had a sibling born by mid- or high forceps. These included 944 mid- or high forceps births (entry criterion 3) and 896 newborns who had a sibling delivered by mid- or high forceps but were not themselves delivered by cesarean or forceps (entry criterion 4). The 1840 newborns were delivered from 928 women (grandmothers). Of the 4010 newborns delivered by cesarean (entry criterion 1), 509 females were linked to one or more births (total 1708 male and female) in the offspring cohort. Of the 1661 newborns who had a sibling deliv­ered by cesarean but were not themselves delivered by cesarean (entry criterion 2), 235 females were linked to one or more births (total 859 male and female) in the offspring cohort. Of the 944 newborns delivered by mid- or high forceps (entry criterion 3), 114 were linked to one or more births (total 387 male and female) in the offspring cohort. Of the 896 newborns who had a sibling deliv­ered by mid- or high forceps but were not themselves delivered by mid- or high forceps or cesarean (entry criterion 4), 131 were linked to one or more births (total 440 male and female) in the offspring cohort. The total distribution of cases and controls is summarized in Table 1. Of all the controls, 54% matched exactly on age, 60% matched exactly on parity, 59% matched exactly on county of birth. The remainder were matched within 2 years of maternal age, within two births for parity, and matched to a similar county in terms of urban, rural, or frontier classification. The distribution in this data base of key demographic attributes, including race, county of birth, multiplicity, 908 Varner et al Heredity and Cesar('(ln maternal age and parity, marital status, and adequacy of prenatal care is similar to the entire Utah population. Figure 2 presents the distribution of cesarean deliveries in both cohorts. Table 2 presents representative labor and/ or delivery complications by cohort population. The maternal cesarean entry criterion group was compared with the subset that was selected as cases. There were no significant differences noted between year of delivery, maternal or paternal age, multiplicity, parity, birth weight or malpresentation. In addition, there were no significant differences in the diagnosis of cephalopelvic disproportion, dysfunctional labor, or prolonged labor. The offspring of case and control women were com­pared for average maternal and paternal education, maternal and paternal race, and marital status. There were no significant differences. Table 3 presents a cross-tabulation of the parental groups and their controls versus the methods and frequency of delivery for female offspring. A woman who was herself delivered by cesarean (entry criterion 1) had a significantly increased likelihood of delivering by cesarean (OR 1.41, 95% CI 1.18-1.70; P < .001). In addition, a woman who was herself delivered by mid­or high forceps (entry criterion 3) also had a signifi­cantly increased likelihood of delivering by cesarean (OR 1.72, 95% CI 1.20-2.47; P = .004). When these groups were combined, a maternal history of operative delivery (cesarean, mid- or high forceps) was highly associated with an increased probability of cesarean in the next generation (OR 1.45, 95% CI 1.25-1.72; P < .001). These results are shown in Figure 3. The subpopulation of women (mothers) who them­selves were delivered by cesarean was further exam­ined (Table 4). The subgroup delivered because of cephalopelvic disproportion was significantly more likely to require cesarean delivery when they had their children (OR 1.83, 95% CI 1.16-2.88; P = .01). The women delivered by cesarean because of indica­tions other than cephalopelvic disproportion were com­pared with their matched controls. A significant inter­generational association was also identified (OR 1.29, Table 2. Distribution of Representative Complications of Labor and / or Delivery by Cohort Prolonged labor (>20 h) Dysfunctional labor Cephalopelvic disproportion Parental cohort, 1947-1957 (N = 197,466) 300 (0.14'1c) 447 (0.22 Ck) 758 (0.37) * Data not available for 1970-1977. Offspring cohort, 1978-1991 (N = 359,069) 1624 (0.29%) 8216 (1.49%) 14,048 (2.54%) Obstetrics & Gynecology Table 3. Maternal Method of Delivery as a Predictor of Female Offspring's Delivery Group SVD Pri CD RCD LF MF Other Unknown Total A) Maternal CD (entry criterion 1) Case 1368 103 126 12 75 7 17 1708 Control 2874 179 161 38 128 22 59 3461 B) Maternal sister with CD (entry criterion 2) Case 711 40 60 5 29 5 9 859 Control 1418 64 66 12 61 7 14 1642 C) Maternal forceps (entry criterion 3) Case 297 28 34 4 21 2 387 Control 635 40 37 6 32 6 9 768 D) Maternal sister with forceps delivery (entry criterion 4) Case 367 27 18 3 20 0 4 440 Control 698 44 36 12 30 3 13 838 SVD = spontaneous vaginal delivery; Pri CD primary cesarean delivery; R CD = repeat cesarean delivery; LF low forceps; MF = mid forceps; Other = vaginal birth after cesarean, vacuum extraction; CD = cesarean delivery. 95% CI 1.06-1.58; P = .01). The ORs between the cephalopelvic disproportion and noncephalopelvic dis­proportion groups are not significantly different (P = .22) (Table 4). The women (mothers) who themselves were deliv­ered by cesarean because of prolonged labor or dys­functional labor or cephalopelvic disproportion (dysto­cia) had an even higher likelihood of requiring a cesarean delivery when they had their children (OR 2.11, 95% CI 1.38-3.24; P < .001). After taking into account the three-way interactions between modes of delivery in mother and offspring, this OR of 2.11 was higher than the OR of 1.29 for mothers whose cesarean birth was not associated with dystocia (P = .04). This Percent Cesarean Delivery # 2 3 Entry Criterion 4 Figure 3. Percent cesarean delivery in women who themselves were delivered by operative delivery (dark bars) compared with matched controls. Group 1 = mother delivered by cesarean (*odds ratio [OR] 1.41, 950/0 confidence interval ICI] 1.2-1.7; P < .0(1); group 2 = maternal sibling with cesarean ("OR 153, 95'/r CI 1.2-2.0; P = .003); group 3 = mother delivered by mid- or high forceps ('OR 1.72, 95'7< CI 1.2-2.5; P < .(01); group 4 = maternal sibling delivered by mid- or high forceps (not significant). VOL. H7, NO.6, JUNE 1996 increased risk is partly attributable to the fact that, for mothers who have offspring by cesarean delivery, there is a significant association between maternal dystocia and maternal cesarean delivery (OR 1.62, 95% CI 1.04- 2.50; P = ,032). To confirm a familial predisposition to operative delivery, we studied the reproductive outcomes of female infants whose maternal aunts had been deliv- Table 4. Odds Ratios and 95% Confidence Intervals for Risk of Cesarean Delivery Among Case Versus Control Offspring Odds ratio 95% CI P A) Stratification by entry criterion 1 (maternal cesarean) 1.41 (1.2-1.7) <.001 2 (maternal aunt-cesarean) 1.53 0.2-2.0) .003 3 (maternal forceps) 1.72 0.2-2.5) .004 4 (maternal aunt-forceps) 1.07 (0.7-1.6) NS lor 3 1.47 (1.3-1.7) <.001 B) Stratification by selected subgroups of entry criterion 1 * Maternal cesarean with 2.11 0.4-3.2) <.001 dystocia" Maternal cesarean without 1.30 0.1-1.6) .01 dystocia" Maternal cesarean with 1.83 (1.2-2.9) .01 cephalopelvic disproportion* \1aternal cesarean without 1.35 (1.1-1.6) .003 cephalopelvic disproportion+ \1aternal cesarean with 5.18 (0.5-56.1) NS prolonged labor Maternal cesarean with 5.97 0.5-23.6) <.001 dysfunctional labor CI = confidence interval; NS = not significant. * Compared with their previously assigned 1:2 controls. I Dystocia: cephalopelvic disproportion or prolonged labor or dys­functional labor. Odds ratios for the dystocia group were significantly higher than that for the non-dystocia group (P = .04). , No evidence that the odds ratios differed between the cephalopel­vic disproportion and non-cephalopelvic disproportion groups (P = .22). Varner et al Heredity and Cesarean 909 ered by cesarean (entry criterion 2) and mid- or high­forceps (entry criterion 4). These data are also presented in a cross-tabulated form in Table 4. Women whose maternal aunt delivered by cesarean had a significantly increased risk for cesarean delivery (OR 1.53, 95% CI 1.16-2.02; P = .003). However, a history of a maternal aunt being delivered by mid- or high forceps was not associated with an increased risk of a woman requiring a cesarean delivery (P = .73) (Figure 3). With an overall cesarean rate of approximately 2.5% in the parental cohort, the relative risk of 1.41 for these women results in an attributable fraction of only 3.5%. In other words, 96.5% of the women experiencing a cesarean in the offspring cohort were themselves deliv­ered vaginally. Discussion The creation of this linked multigenerational data base containing specific obstetric indices has allowed the first systematic study of the intergenerational predisposition to operative delivery. Our data show that women who were born via operative delivery have a significantly increased likelihood of requiring a cesarean delivery during their own pregnancies, particularly if the mother was born by mid- or high forceps or if the maternal cesarean was needed because of cephalopelvic dispro­portion or dystocia. The case-control design of the study has minimized ascertainment bias in this popu­lation. We recognize that clinical obstetrics during the pa­rental cohort era (1947-1957) was different from current practice. Improved knowledge and medical-legal con­cerns have led to an increased frequency of diagnosis of most labor and/or delivery complications (Table 2). Specifically, obstetricians are now more aware of the potential risks of prolonged labor.lO As shown in Table 1, the diagnosis of cephalopelvic disproportion has increased 6.9-fold, and the diagnosis of dysfunctional labor has also increased 9.9-fold. However, the diagno­sis of prolonged labor has increased only twofold, suggesting that physicians no longer tolerate prolonged labor and are more likely to intervene operatively with a clinical diagnosis of dysfunctional labor or cephalo­pelvic disproportion. Previously, the possibility of a familial reproductive factor predisposing to operative delivery was consid­ered. X-ray pelvimetry series suggest that 98% of pelviC configurations are normal growth variants, probably on a hereditary basis.]] Preconceptional maternal nutrition also plays an important part in maternal pelvic de vel­opment. 12 More recently, Harrison13 predicted that, in Africa, improved nutrition of pregnant women will initially 910 Varner et al Heredity <llid Cesarean result in a higher cesarean rate because of cephalopelvic disproportion. However, this initial increase will de­cline as more women, who were optimally nourished both in utero and in childhood, reach their reproductive epoch. In part, these predictions confirm the earlier observations of Ounsted and Ounsted,14 who suggested that the maternal mechanisms constraining fetal growth are quantitatively set as each generation of mothers pass through their intrauterine life. Although these considerations are important in any discussion of longitudinal population trends, we be­lieve that the historic socioeconomic homogeneity of the Utah population should minimize any such effects in these data (Report to the Governor: Poverty in Utah 1993, State Community Services Office, Utah Depart­ment of Community and Economic Development, Salt Lake City, Utah, December 1993). Likewise, dietary patterns in Utah are similar to those of the United States as a whole.1s The population of Utah is ethnically, educationally, and economically homogeneous, but not genetically different from other white American popu­lations. 16 Utah still maintains the highest fecundity rate (2.6 births per woman), has the lowest median age (26.2 years), and has the highest dependency ratio (82 depen­dents per 100 working persons) of any state.17 Utah thus provides an optimum background for evaluation of the null hypothesis of this study. Although this may limit the applicability of these findings to other populations, the low-risk nature of the Utah population may be more likely to reveal genetic effects, because environmental pressures such as malnutrition have been minimal over the course of this study. Demographic variables also change with time and any consideration of intergenera­tiona I effects should control for maternal age and parity, as in this study. Marital status is also correlated with risk for cesarean delivery. The offspring cohort's lower percentage of married women is consistent with national trends of births to unmarried women, especially teenagers. IS There were no significant differences between any of our case and control groups for marital status. Our attributable risk calculations suggest that the majority of the recent increase in cesarean rates is due to changes in practice management19 - 21 rather than to the survival of women with functional reproductive inad­equacy. Nonetheless, our data identify a real biologic phenomenon. The precise mechanism(s) of this biologic imperative remain unknown. The likelihood of a verti­cally transmissible absolute cephalopelvic dispropor­tion trait is highly improbable. However, these data are compatible with an inheritable tendency toward poor uterine contraction or soft tissue relaxation. Several previous cross-sectional studies have provided sugges­tive evidence for such a mechanism. Rechberger et al22 Obstetrics & Gynecology showed that primigravidas whose cervices had high levels of collagen and hyaluronic acid had longer cer­vical dilation times. Granstrom et aID demonstrated that, compared with controls, women delivered by cesarean because of protracted labor had a significantly higher concentration of collagen in the uterine isthmus and cervix as well as a decrease in solubility of the collagen. Both Granstrom et al23 and Uldbjerg et af4 concluded that insufficient remodeling of uterine tissue during pregnancy may contribute to protracted labor, ie, operative delivery. These alterations in collagen mechanics may be due to changes in the ratio of collagen types or to isomorphisms/5 either of which may be hereditary.26 Our data demonstrate an intergenerational predispo­sition to operative delivery, particularly if the parental cesarean was performed for cephalopelvic dispropor­tion or dysfunctional labor. Although the current attrib­utable risk is low, continued amplification of this trend may have a substantial impact on future obstetric practice. References 1. Notzon FC, Placek PJ, Taffel SM. Comparisons of national cesare­an- section rates. N Engl J Med 1987;316:386-9. 2. Chesley LC, Cooper DW. Genetics of hypertension in pregnancy: Possible single gene control of pre-eclampsia and eclampsia in the descendants of eclamptic women. Br J Obstet Gynaecol 1986;93: 898-908. 3. Lamb K, Hoffmann RG, Nichols TR. Family trait analysis: A case-control study of 43 women with endometriosis and their best friends. Am J Obstet Gynecol 1986;154:596-601. 4. Turnbull A. Prolonged pregnancy. In: Turnbull A. Chamberlain G, eds. Obstetrics. Edinburgh, Scotland: Churchill Livingstone, 1989: 774. 5. Johnstone F, Inglis L. Familial trends in low birth weight. BMJ 1974;3:659-61. 6. Parazzini F, Negre E, LaVecchia C, Restelli C, Franceschi S. Family history of reproductive cancers and ovarian cancer risk: An Italian case-control study. Am J Epidemiol 1992;135:35-40. 7. Schlesselmann Jj. Case control studies: Design, conduct and anal­ysis. New York: Oxford University Press, 1982. S. Lawley DN. A general method for approximating to the distribu­tion of likelihood-ratio criteria. Biometrika 1956;43:295-303. 9. Numerical Algorithms Group, Inc. Generalized linear interactive modeling 3.77. London: Numerial Algorithms Group, 1985. 10. Friedman EA. Labor: Clinical evaluation and management. 2nd ed. New York: Appleton-Century-Crofts, 1978. 11. Steer CM. Moloy's evaluation of the pelvis in obstetrics. 2nd ed. Philadelphia and London: WB Saunders Co., 1959:13. VOL. 87, NO.6, JUNE 19% 12. Thoms H. Pelvimetry. New York: Harper, 1956:47. 13. Harrison KA. Predicting trends in operative delivery for cephalo­pelvic disproportion in Africa. Lancet 1990;335:861-2. 14. Ounsted M, Ounsted C. On fetal growth rate (its variations and their consequences). Clin Dev Med 1973;46:72-82. 15. Lyon JL, Sorenson AW. Colon cancer in a low risk population. Am J Clin Nutr 1978;31:5227-30. 16. McLelland T, Jorde LB, Skolnick MH. Genetic distances between Utah Mormons and related populations. Am J Hum Genet 1984; 36:836-40. 17. Utah Data Guide, Utah State Data Center. Vo!' 13, no. 1. January 1994. Salt Lake City, Utah: Governor's Office of Planning and Budget, 1994. lR. Hein HA, Burmeister LF, Papke KR. The relationship of unwed status to infant mortality. Obstet Gynecol 1990;76:763-8. 19. Taffel SM, Placek PJ. 'v1oien M. 1988 U.s. cesarean-section rate at 24.7 per 100 births-A plateau? \J Engl J Med 1990;323:199-200. 20. Shiono PH, McNellis D, Rhoads GG. Reasons for the rising cesarean delivery rates: 1978-1984. Obstet Gynecol 1987;69:696- 700. 21. Taffel SM. Cesarean section in America: Dramatic trends, 1970- 1987. Stat Bull Metrop lnsur Co 1989;70:2-11. 22. Rechberger T, Uldbjerg 1\, Oxlund H. Connective tissue changes in the cervix during normal pregnancy and pregnancy complicated by cervical incompetence. Obstet Gynecol 1988;7:563-6. 23. Granstrom L, Ekman G, Glmsten U, Malmstrom A. Changes in the connective tissue of corpus and cervix uteri during ripening and labour in term pregnancy. Br J Obstet Gynaecol 1989;96:1198-202. 24. Uldbjerg N, Ekman G, Malmstrom A, Olsson K, Ulmsten U. Ripening of the human uterine cervix related to changes in collagen, glycosaminoglycans, and collagenolytic activity. Am J Obstet GynecoI1983;147:662-6. 25. Ox lund H. Relationship between the biomechanical properties, composition, and molecular structure of connective tissue. Con­nect Tissue Res 1986;15:65-72. 26. Prockop D, Kivirikko K. Tuderman L, Guzman NA. The biosyn­thesis of collagen and its disorders (first of two parts). N Engl J Med 1979;301 :13-23. Address reprint requests to: Michael W. Varner, MD Department of Obstetrics and Gynecology University of Utah School of Medicine 50 North Medical Drive, #28200 Salt Lake City, UT 84132 Received AugHst 21, 1995. Received in ITl'ised form December 18, 1995. Accepted December 19, 1995. Copyright © 1996 by The American College of Obstetricians and Gynecologists. Published by Elsevier Science Inc. Varner et al Heredity and Cesarean 911