Unreliable first day of the last menstrual period (LMP) is common. It has been reported with varied frequencies (10–44.7%) depending on the population studied and criteria used.1–4 The LMP is normally considered unreliable if a woman reports irregular menstrual cycles, conception within 3 months of stopping oral contraceptives (OCs), or uncertainty about the LMP.3 Previous studies found that an unreliable LMP was associated with high rates of low birth weight (LBW), low socioeconomic status, and high rates of preterm delivery.5–7 In the present study, we tested the hypothesis that women with unreliable LMPs might have increased risks of late abortion, stillbirth, postnatal death within 1 year, preterm birth, and LBW.
Materials and Methods
The study was a population-based follow-up study. During the period 1986–1996, obstetric ultrasound data from 21,936 unselected singleton pregnancies in a geographically well-defined area were recorded prospectively as part of routine examination in maternity care at Herlev University Hospital. Ultrasound scanning and data registration were done by trained midwives. The database was linked to registered data from the Danish National Board of Health to collect information about pregnancy outcomes. To validate the data from that link, we studied 3893 patient records, the total number of women referred for scanning from January 1, 1993 to March 31, 1994. Furthermore, we studied the files of all 487 singleton pregnancies that resulted in spontaneous, missed, and induced abortions and fetal and postnatal death within 1 year that had ultrasounds in the second trimester and were registered in the ultrasound database. Relevant data from patient records were added to the database, including demographics (race, age, height, and weight), reproductive histories, LMPs, ultrasound findings, and pregnancy outcomes, which made it possible to control the validity of data in the linked registries. The validation study showed a high degree of completeness and correctness of raw data such as date of birth, gender, birth weight, and Apgar scores reported from local hospitals to the national registry.
Women were excluded from the study if they had pregnancies with only one biparietal diameter (BPD) measurement either before 12 weeks (20 mm) or after 22 weeks of gestation (55 mm) (1152, 5.3%), were missing registration of LMP (298, 1.4%), had induced abortions after 12 weeks of gestation (185, 0.8%), had missed abortions diagnosed at the first ultrasound scan (52, 0.2%), or had no follow-up data (168, 0.8%) (moved abroad or delivered at other hospitals). Some cases had multiple reasons for exclusion. In total, 1692 singleton pregnancies (7.7%) were excluded, leaving 20,244 for study; 19,971 infants were still alive 1 year after birth and 273 died between 12 weeks' gestation and 1 year after birth (112 spontaneous or missed abortions between 12 and 28 weeks' gestation, 79 stillbirths, 48 postnatal deaths within the first month, and 34 between 1 month and 1 year).
The LMP was registered as reliable (81.4%) or unreliable (18.6%) according to women's recall of their LMPs and obstetricians' assessments of their medical histories at the first visit (between 12 and 22 weeks' gestation). The guidelines of the obstetric department for registration of LMP as unreliable were irregular cycles, uncertainty about the first day of LMP, conception within 3 months after stopping OCs, or conception within 3 months of an abortion or delivery.
Gestational age was calculated by BPD according to Persson's equation.8 Date zero, corresponding to the first day of the LMP, was calculated by subtracting gestational age, based on BPD, from the actual date of ultrasound examination. Term by BPD was predicted by adding 280 days to date zero. Preterm birth was defined as delivery more than 21 days before predicted term by BPD.
The sex-specific expected mean birth weight was calculated according to a Danish-Swedish equation obtained from longitudinal intrauterine ultrasound–estimated fetal weight from 86 normal pregnancies.9 Actual weight deviation (%) was calculated as: Birth weight −sex-specific expected birth weight for gestation/estimated weight. Weight deviation below −22% was used as a cutoff for defining small for gestational age (SGA), which corresponded to −2 standard deviations (SD) of the expected birth weight for gestational age.
The following adverse pregnancy outcomes were considered: death (spontaneous or missed abortion between 12 and 28 weeks' gestation, fetal death after 28 weeks' gestation, or postnatal death within 365 days), preterm birth, LBW (less than 2500 g), and LBW for gestational age (weight deviation less than −22%).
Measurement of Maternal Serum Alpha-Fetoprotein (AFP)
Each pregnant woman was offered AFP measurement between 12 and 22 weeks' gestation as a screening procedure for neural tube defects. In total, 19,150 of 21,936 women (87.3%) had at least one measurement. Alpha-fetoprotein was measured in kU/L. To obtain the reference median of AFP in relation to gestational age, we performed linear regression analysis using ultrasound gestational age (GABPD in days) as the independent variable and concentration of AFP as the dependent variable. The regression line was:
Multiples of the median (MoM) of AFP were calculated as the concentration of AFP divided by the regressed mean concentration of AFP for the corresponding GABPD.
Independent Variables and Statistical Analysis
Logistic regression analysis was used to estimate the risk of adverse pregnancy outcomes in relation to unreliable LMPs and to potential confounding variables. Numerous covariates from the linked registries were studied initially as potential confounders, including demographics (maternal age, height, weight), reproductive history (number of pregnancies, pregnancy losses, elective abortions), prenatal diagnostic tests (amniocentesis or chorionic villus sampling), hospital admissions during pregnancy (number of women who bled early in pregnancy, number of hospital admissions due to preeclampsia), and malformations confirmed at birth. The risk of adverse pregnancy outcome was analyzed first with unreliable LMP alone in a model to calculate univariate estimates of effect. The adjusted model was presented to include only covariates that were significantly associated with adverse pregnancy outcomes.
Because abnormal values of AFP are associated with adverse pregnancy outcomes,10–15 the risk was analyzed in three groups of AFP (MoM less than 0.5, between 0.5 and 2, and greater than 2). The risk of adverse pregnancy outcome was not significantly different for pregnancies stratified with normal and low AFP MoM. These data were therefore pooled and compared with pregnancies with AFP MoM greater than 2.
The Kaplan-Meier method was used to estimate survival curves for fetuses from pregnancies with reliable and unreliable LMPs. The start time of pregnancy was set to the BPD-calculated date zero (corresponding to the first day of the LMP). The risk of death was studied from 12 weeks to 280 + 365 days after date zero. All analyses were done using the statistical software package SAS version 6.12 (SAS Institute Inc, Cary, NC).
Women with unreliable LMPs tended to be younger, to have had more previous deliveries, to come earlier to routine scanning in the second trimester, and to have more scans per pregnancy (Tables 1 and 2).
The chance of survival during the first year of life was highest in pregnancies with reliable LMPs (odds ratio [OR] 2.0; 95% confidence interval [CI] 1.5, 2.6; P = .001) (Figure 1). The risks of fetal and postnatal death were doubled in pregnancies with unreliable LMPs compared with those with reliable LMPs, and this risk was highest for stillbirth (OR 2.7; 95% CI 1.7, 4.3) (Table 3). The risks of other adverse pregnancy outcomes, including preterm birth, LBW, and LBW for gestation, were also significantly higher in pregnancies with unreliable LMPs (Table 4). Even after adjustment for significant confounding covariates, the risk of adverse pregnancy outcome was still robust and significantly higher in the group of women with unreliable LMPs (Table 4).
The frequency of AFP MoM greater than 2 was 2.6% overall (490 of 18,615 with measured AFP), and this was borderline significantly high in women with unreliable LMPs (3.1%) compared with those with reliable LMPs (2.5%) (OR 1.2; 95% CI 0.99, 1.5). High AFP was associated with significantly increased risks of death (OR 6.1; 95% CI 4.1, 8.9), preterm birth (OR 4.0; 95% CI 3.2, 5.0), LBW (OR 4.2; 95% CI 3.3, 5.5), and LBW for gestation (OR 3.0; 95% CI 2.3, 3.8). In the adjusted model in which LMP and AFP MoM were included as covariates, the risk of adverse pregnancy outcome was increased more than three times when AFP MoM was greater than 2 (Table 5). Women with unreliable LMPs and AFP MoM greater than 2 had an almost 12 (1.9 × 6) times higher risk of fetal or infant death compared with women with reliable LMPs and AFP MoM no more than 2 (Table 5).
The main finding of this study was that an unreliable LMP was associated with increased risk of death, preterm birth, LBW, and LBW for gestation. High AFP (AFP MoM greater than 2) increased the risk by more than three times.
The data for classification of LMP were not standardized when registered as either reliable or unreliable, which limits the possibility of analyzing the cause and effect relation between different components of menstrual history and adverse pregnancy outcomes. The study period of 10 years, combined with the lack of standardized data for LMP, might result in heterogeneity of the collected data. However, the analysis of stability of risk factors over time showed a tendency toward a higher incidence among women with unreliable LMPs during the study period, and the variation in relation to LMP registered as reliable or unreliable was apparently narrow (data not shown). The risk of recall bias was negligible because data were collected prospectively. The number of patients was large, so the calculated OR for adverse pregnancy outcomes between unreliable and reliable LMPs was significantly different, but the CI was narrow, so chance was an unlikely explanation for our observations.
The cause of the association between unreliable LMP and death is not obvious. In a study of 286,430 single births in Norway from 1975 to 1979, Herman et al16 reported that about 13% of women were “unsure” of the exact day of LMP, and these women did not have increased or decreased rates of fetal, neonatal, or infant death. However, information about cycle length, regularity, and contraception were not available, which makes it difficult to compare these findings with ours.
The cause of the associations between unreliable LMP and increased frequency of preterm birth, LBW, LBW for gestation, and abnormal AFP was also unclear. Results from several studies agree with our observations.5–7,17,18 However, the data in the literature tell little about whether the association was causal and has a biologic explanation, or merely reflects the confounding effect of socioeconomic circumstances. An unreliable first day of the LMP could be related to a short interval between pregnancies because short intervals between pregnancies have been associated repeatedly with adverse perinatal outcomes (eg, LBW and preterm birth). It is unknown whether this association is due to confounding by other factors, such as maternal age, socioeconomic status, and reproductive history.19–22 Some authors reported an increased incidence of vaginal infections among pregnant women with lower socioeconomic status and poor pregnancy outcomes.23–26 Unreliable LMP might be associated with increased risk of vaginal infection, making it a confounding variable. However, we were unable to find any study that supported this hypothesis.
A plausible explanation for the association between unreliable LMP and increased risk of adverse pregnancy outcome might be that there is a complex interaction between dysfunction of the endocrine system and unfavorable socioeconomic conditions. Buekens et al5 studied 22,404 pregnancies and found that unknown menstrual period (16%) was associated with high rates of LBW, low socioeconomic and sociodemographic status, and high rates of preterm deliveries. Computerized birth files compiled by the state of North Carolina from 1975 to 1977 showed that reportedly missing or inaccurate gestational age data were found in women who came from a sociodemographically high-risk subpopulation.6 Hakala7 reported that among 29,061 pregnancies, “unmarried women from lower social classes more often had ‘behavioral’ risk factors: they were more often heavy smokers, were unreliable about the dates, had had two or more abortions and neglected maternity care.” Chimbira17 reported that a small population of 214 women with unreliable LMPs had significantly increased risks of adverse pregnancy outcomes such as cesarean and forceps delivery, vacuum extraction, stillbirth, and neonatal mortality compared with 786 women with reliable LMPs. Xu et al18 studied 5291 women and investigated the relation between maternal menstrual history and SGA fetuses. They found a more than twofold increase in the risk of SGA among underweight and normal-weight women when their menstrual cycles were at least 31 days compared with those whose cycles were 28 days or less, but no associated increase in risk for overweight women. The risk of SGA was increased among women with ages at menarche of 15 years or older compared with those with ages at menarche of 14 years or younger.18 Concentrations of fetal and cord serum insulin-like growth factor-1 were positively correlated to fetal size,27–30 and the concentrations were increased in the early pubertal stages and decreased in the late stages.31 Therefore, Xu et al18 speculated that the effect of insulin-like growth factor-1 on fetal growth could be a possible link between fetal growth restriction and maternal menstrual variables. Increased AFP in women with unreliable LMPs might indicate abnormalities of placental structure or function early in pregnancy.32 Data from cancer epidemiology showed that women who reported their menstrual cycles as irregular at 20 years old had a reduced risk of breast cancer mortality.33,34
This exploratory study found an association between uncertain LMP and adverse pregnancy outcomes, which gives inspiration to further studies and clinical implications. The registration of LMP needs to be more detailed in the classification of various types of unreliable LMPs, which will allow further investigations of associations.
1. Geirsson RT, Busby-Earle RM. Certain dates may not provide a reliable estimate of gestational age. Br J Obstet Gynaecol 1991;98:108–9.
2. Geirsson RT. Ultrasound instead of last menstrual period as the basis for gestational age assignment. Ultrasound Obstet Gynecol 1991;1:212–9.
3. Campbell S, Warsof SL, Little D, Cooper DJ. Routine ultrasound screening for the prediction of gestational age. Obstet Gynecol 1985;65:613–20.
4. Hall MH, Carr-Hill RA, Fraser C, Campbell D, Samphier ML. The extent and antecedents of uncertain gestation. Br J Obstet Gynaecol 1985;92:445–51.
5. Buekens P, Delvoye P, Wollast E, Robyn C. Epidemiology of pregnancies with unknown last menstrual period. J Epidemiol Community Health 1984;38:79–80.
6. David RJ. The quality and completeness of birthweight and gestational age data in computerized birth files. Am J Public Health 1980;70:964–73.
7. Hakala T. Obstetric care, pregnancy risk factors and perinatal outcome in the province of Uusimaa, Finland, in 1980–1981. Ann Chir Gynaecol Suppl 1987;203:1–83.
8. Persson PH, Weldner BM. Normal range growth curves for fetal biparietal diameter, occipito frontal diameter, mean abdominal diameters and femur length. Acta Obstet Gynecol Scand 1986;65:759–61.
9. Marsal K, Persson PH, Larsen T, Lilja H, Selbing A, Sultan B. Intrauterine growth curves based on ultrasonically estimated foetal weights. Acta Paediatr 1996;85:843–8.
10. Brazerol WF, Grover S, Donnenfeld AE. Unexplained elevated maternal serum alpha-fetoprotein levels and perinatal outcome in an urban clinic population. Am J Obstet Gynecol 1994;171:1030–5.
11. Maher JE, Davis RO, Goldenberg RL, Boots LR, DuBard MB. Unexplained elevation in maternal serum alpha-fetoprotein and subsequent fetal loss. Obstet Gynecol 1994;83:138–41.
12. Waller DK, Lustig LS, Smith AH, Hook EB. Alpha-fetoprotein: A biomarker for pregnancy outcome. Epidemiology 1993;4:471–6.
13. Crandall BF, Robinson L, Grau P. Risks associated with an elevated maternal serum alpha-fetoprotein level. Am J Obstet Gynecol 1991;165:581–6.
14. Killam WP, Miller RC, Seeds JW. Extremely high maternal serum alpha-fetoprotein levels at second-trimester screening. Obstet Gynecol 1991;78:257–61.
15. Robinson L, Grau P, Crandall BF. Pregnancy outcomes after increasing maternal serum alpha-fetoprotein levels. Obstet Gynecol 1989;74:17–20.
16. Herman AA, Yu KF, Hoffman HJ, Krulewitch CJ, Bakketeig LS. Birth weight, gestational age and perinatal mortality: Biological heterogeneity and measurement error. Early Hum Dev 1993;33:29–44.
17. Chimbira TH. Uncertain gestation and pregnancy outcome. Cent Afr J Med 1989;35:329–33.
18. Xu B, Jarvelin MR, Xu X, Wang Z, Qin H, Rimpela A. Maternal menstrual history and small-for-gestational-age births in a population-based Chinese birth cohort. Early Hum Dev 1997;49:183–92.
19. Zhu BP, Rolfs RT, Nangle BE, Horan JM. Effect of the interval between pregnancies on perinatal outcomes. N Engl J Med 1999; 340:589–94.
20. Rawlings JS, Rawlings VB, Read JA. Prevalence of low birth weight and preterm delivery in relation to the interval between pregnancies among white and black women. N Engl J Med 1995;332:69–74.
21. Basso O, Olsen J, Knudsen LB, Christensen K. Low birth weight and preterm birth after short interpregnancy intervals. Am J Obstet Gynecol 1998;178:259–63.
22. Klerman LV, Cliver SP, Goldenberg RL. The impact of short interpregnancy intervals on pregnancy outcomes in a low-income population. Am J Public Health 1998;88:1182–5.
23. Cotch MF, Pastorek JG II, Nugent RP, Yerg DE, Martin DH, Eschenbach DA. Demographic and behavioral predictors of Trichomonas vaginalis infection among pregnant women. The Vaginal Infections and Prematurity Study Group. Obstet Gynecol 1991;78:1087–92.
24. Zhang ZF. Epidemiology of trichomonas vaginalis. A prospective study in China. Sex Transm Dis 1996;23:415–24.
25. Cotch MF, Pastorek JG II, Nugent RP, Hillier SL, Gibbs RS, Martin DH, et al. Trichomonas vaginalis associated with low birth weight and preterm delivery. The Vaginal Infections and Prematurity Study Group. Sex Transm Dis 1997;24:353–60.
26. Gratacos E, Figueras F, Barranco M, Vila J, Cararach V, Alonso PL, et al. Spontaneous recovery of bacterial vaginosis during pregnancy is not associated with an improved perinatal outcome. Acta Obstet Gynecol Scand 1998;77:37–40.
27. Woods KA, Camacho-Hubner C, Savage MO, Clark AJ. Intrauterine growth retardation and postnatal growth failure associated with deletion of the insulin-like growth factor I gene. N Engl J Med 1996;335:1363–7.
28. Gluckman PD. Clinical review 68: The endocrine regulation of fetal growth in late gestation: The role of insulin-like growth factors. J Clin Endocrinol Metab 1995;80:1047–50.
29. Delmis J, Drazancic A, Ivanisevic M, Suchanek E. Glucose, insulin, HGH and IGF-I levels in maternal serum, amniotic fluid and umbilical venous serum: A comparison between late normal pregnancy and pregnancies complicated with diabetes and fetal growth retardation. J Perinat Med 1992;20:47–56.
30. Giudice LC, de Zegher F, Gargosky SE, Dsupin BA, de las Fuentes L, Crystal RA, et al. Insulin-like growth factors and their binding proteins in the term and preterm human fetus and neonate with normal and extremes of intrauterine growth. J Clin Endocrinol Metab 1995;80:1548–55.
31. Juul A, Bang P, Hertel NT, Main K, Dalgaard P, Jorgensen K, et al. Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: Relation to age, sex, stage of puberty, testicular size, and body mass index. J Clin Endocrinol Metab 1994;78:744–52.
32. Burton BK. Elevated maternal serum alpha-fetoprotein (MSAFP): Interpretation and follow-up. Clin Obstet Gynecol 1988;31:293–305.
33. Michels-Blanck H, Byers T, Mokdad AH, Will JC, Calle EE. Menstrual patterns and breast cancer mortality in a large U.S. cohort. Epidemiology 1996;7:543–6.
© 2000 The American College of Obstetricians and Gynecologists
34. Parazzini F, La Vecchia C, Negri E, Franceschi S, Tozzi L. Lifelong menstrual pattern and risk of breast cancer. Oncology 1993;50:222–5.