Naturally occurring birth patterns have been the subject of a variety of research efforts dating back to at least 1848.1 Although the exact nature of the periodicity of human birth is debated, both seasonal and daily rhythmic patterns have been identified.1–9 In the United states, several investigators have reported that women tend to give birth during late summer and early fall, with most conceptions occurring during colder months.1,5,6,8,9 In addition, birth after the spontaneous onset of labor has been reported to be more common in the morning or early afternoon, suggesting that the onset of labor is more likely at night.1 These studies primarily examined low-risk populations; however, detailed information about patient characteristics is often limited. Recent national birth certificate data show that births in the United States fluctuate predictably according to the day of the week, with significantly more births occurring on weekdays and fewer births occurring on weekends.9,10 The index of occurrence statistic can be used to quantify this effect. By using this technique, the average number of births per day is designated by the number 100. From 1989 to 2000, the mean index of occurrence was highest for Tuesday births at 111.66, and lowest for Sunday births at 76.08.10 This distribution could be interpreted to reflect medical interventions, such as scheduled inductions of labor and cesarean deliveries. Repeat cesarean births, for example, are typically scheduled during the workweek, with few being performed on the weekend. It has also been suggested that elective inductions of labor at term are increasing, and this may be an additional factor contributing to this pattern.10 We currently triage spontaneously laboring women with term, uncomplicated singleton pregnancies to a low-risk pregnancy labor and delivery unit. We reasoned that examination of the timing of spontaneous parturition in this unit might provide insight into the timing of naturally occurring labor and delivery.
MATERIALS AND METHODS
This is a retrospective institutional review board–approved study of women admitted to the low-risk labor unit at Parkland Hospital from January 1, 2000, to December 31, 2000. All women were between 360/7 and 416/7 weeks of gestation and in spontaneous active labor with singleton gestations. Delivery events were recorded by attending nurses, and data sheets were checked for accuracy by research nurses and entered into a computerized database. Women not in active labor or with contraindications to labor, including prior cesarean delivery, were admitted to a contiguous high-risk labor unit and were not included in this analysis. Likewise, pregnancies with significant medical complications, a known fetal anomaly, and stillbirths were not managed in the low-risk pregnancy unit. Women admitted to the low-risk labor unit and who went on to develop intrapartum complications, such as gestational hypertension, labor dystocia requiring oxytocin augmentation, operative vaginal delivery, or cesarean delivery were included in this analysis.
Labor in our low-risk pregnancy unit is managed with a standardized management protocol. Briefly, active labor is diagnosed in the presence of at least 1 of the following: 1) palpable uterine contractions every 3 minutes and cervical dilation of 3 or more centimeters, 2) uterine contractions that result in cervical dilation in 2 hours, or 3) rupture of membranes and palpable contractions every 3 minutes. Pelvic exams are performed approximately every 2 hours. If the cervix does not dilate within the first 2 hours of admission, an amniotomy is performed and labor progress is evaluated at the next 2-hour evaluation. If labor does not progress, an intrauterine pressure catheter is placed to assess uterine activity. If hypotonic contractions are present in the setting of no further cervical progress, oxytocin is administered. Uterine activity of 200 to 250 Montevideo units is expected for 2 to 4 hours before the diagnosis of dystocia. Therefore, it would not be expected for a cesarean for dystocia to be performed for at least 6 to 8 hours from admission. Staffing of the labor unit included 24-hour supervision by obstetrical faculty members from the Department of Obstetrics and Gynecology at the University of Texas Southwestern Medical School. Other staff members included 3 or 4 certified nurse-midwives per shift, directly supervised by 1 on-site second-year house officer in obstetrics and gynecology who reported to 1 on-site fourth-year house officer.
The number of births per day of the week and the number of births per month of the year were analyzed. The time of day was analyzed in 1- and 3-hour intervals with analysis performed in Central Standard Time. The effect of daylight saving on the time of day was analyzed separately. The months were grouped into 4 seasons. In Dallas, Texas, the coldest 3 months of the year are December, January, and February,11 with the warmest 3 months being June, July, and August; thus, our analysis of seasonal variation was December through February, March through May, June through August, and September through November.
The goodness of fit χ2 test was used to examine whether hours of the day, days of the week, or seasons of the year had a uniform distribution pattern. An analysis of variance was used to examine the equality of the frequency of events per time period. In the event of statistical significance, the Student–Newman–Keuls approach was used to examine the pairwise comparisons. Pearson χ2 tests were used for comparisons of categorical data. A Newman–Keuls approach was used to evaluate multiple comparisons for χ2 as well.
A total of 6,608 low-risk women met study criteria and were included in the study. The mean maternal age was 23.8 ± 5.3 years with 116 women aged 15 years or younger and 211 women aged 35 years or older. Women (n = 5,546) were Hispanic (84%), African American (n = 722; 11%), Caucasian (n = 210; 3%), and other (n = 130; 2%). Thirty-eight percent of the women were nulliparous (n = 2,522). As shown in Table 1, there was no significant difference found in the day of the week and the number of births, with a mean of 14.3% of births occurring on any given day for a frequency of 18.15 births per day. There was no effect of age, race, or parity on the number of births and day of the week (P = .42, .67, .95, respectively).
The mean number of births per hour was 1.42, with the most births occurring between 13:00 and 13:59 hours and the least number of births occurring between 22:00 and 22:59 hours, as shown in Figure 1. Table 2 shows the distribution of births in relation to the time of day broken into 3-hour intervals. Sixteen percent of women required oxytocin augmentation, and 3.8% underwent cesarean delivery. Women whose labor required these interventions were more likely to deliver between the hours of 11:00 and 13:59, whereas women who required no intervention in labor and had a vaginal delivery were more likely to deliver between the hours of 08:00 and 10:59 (P < .001). Parity also had an effect on the timing of spontaneous birth. Multiparous women were more likely to deliver between the hours of 08:00 and 10:59 (P = .02) when compared with nulliparous women. Parity did not affect the time period during which the least number of births occurred (22:00 to 22:59, both groups). There was no effect of maternal age or race on the timing of spontaneous birth (P = .93, .95, respectively). Table 3 shows the number of births in relation to the season of the year. The greatest number of deliveries occurred in the fall, September through November, and the fewest deliveries occurred in the spring, March through May.
Analysis of variance was used to examine the effect of the time of year (season) on the number of births and the day of week. Although the actual numbers of births per day of the week changes as the seasons change, these changes are randomly distributed with no trend toward any particular day (P = .34). Similarly, the day of the week had no effect on the time of birth (P = .34). Finally, daylight saving time had no effect on the time of day of birth (P = .32).
In this population of low-risk women with term pregnancies who experienced spontaneous onset of labor, births were randomly distributed throughout the week. Births were most likely to occur between the hours of 11:00 and 13:59, peaking at the 13:00 hour, and least likely to occur between 20:00 and 22:59. The peak birthing time was 13:00 hours. In multiparous women, births peaked earlier in the day, between the hours of 08:00 and 1059. There was no effect of race or age on the time of day. In addition to daily variation, we found seasonal variation in birth rates with more births occurring in the fall and fewer in the spring. Births were randomly distributed throughout the week regardless of the time of year. Indeed, we were unable to show an association with birth and the day of the week under any circumstance, including adjusting for daylight saving time.
Our finding that term births after the spontaneous onset of labor peak in the early afternoon supports previous reports of a circadian pattern of birth.1–5,8–10,12 Anderka et al,1 in a study of 242,276 low-risk term singleton births in Massachusetts, reported a 24-hour cycle in spontaneous birth, with a peak incidence between 11:00 and 13:00. Similar to our study, they reported an effect of parity with nulliparous women being more likely to deliver midday, as opposed to multiparous women who delivered in the early morning hours. Consistent with this is the observation by Cooperstock et al12 that the onset of labor is most typically at night. In their study of more than 18,000 singleton pregnancies, the onset of labor peaked at 01:45 in both term and preterm births. The mechanisms underlying the variations in the timing of birth are not well understood but may be influenced by the maternal and fetal endocrine system. Walsh et al,13 in a study of primates, showed a correlation between uterine activity and maternal and fetal steroid levels with fetal dehydroepiandrosterone sulfate and progesterone levels paralleling uterine activity in late gestation. These fetal steroid levels followed a circadian rhythm. The authors hypothesized that interaction between the maternal and fetal endocrine system may limit parturition, or at least its initiation to certain times of the day.
Seasonal patterns to human birth have been observed in virtually all populations studied, with more births occurring during certain times of the year. The peak birth rate appears to be determined primarily by location and the climatic conditions of the population studied. In northern Europe, for example, births peak in the spring with a nadir in the fall and winter months.14 In contrast, we found the fewest number of births during the spring, suggesting that women were least likely to conceive during the hottest months. This is consistent with other studies of birthing patterns in the southern United States. One explanation for this pattern of births in warmer climates may be related to male factors. Levine et al6 studied the effect of warm weather on semen quality of outdoor workers in San Antonio, Texas. They found that semen quality decreases in the summer and concluded that this correlated with decreased fertility in the hottest months of the year and a corresponding decreased number of births during the spring.
Our findings are compared with recently reported National Vital Statistics in Table 4. Nationally, there is a significant variation in the weekday and weekend birth rates not seen in our study population.10 In contrast to our report, the National Vital Statistics data are a summary of all pregnancies including those women with significant medical complications and prior cesareans. Consequently, inductions of labor and scheduled repeat cesareans would be expected to be more prevalent and likely explains the difference between our results and the national data. Our inability to find an association between the day of the week and the number of births in spontaneously laboring, uncomplicated pregnancies is not be unexpected. A given day does not in and of itself reflect a naturally occurring event. The time of day, however, as well as the season of the year is determined by naturally occurring phenomenon that appear to have effects on both conception rates and the onset of labor through circadian rhythms and biological clocks.
1. Anderka M, DeClercq ER, Smith W. A time to be born. Am J Public Health 2000;90:124–6.
2. James WH. Seasonal variation in human births. J Biosoc Sci 1990;22:113–9.
3. Glattre E, Bjerkedal T. The 24-hour rhythmicity of birth: a populational study. Acta Obstet Gynecol Scand 1983;62:31–6.
4. Cunningham FG, Gant NF, Leveno KJ, Gilstrap LC III, Hauth JC, Wenstrom KD. Williams obstetrics. 21st ed. New York (NY): McGraw-Hill; 2001.
5. Bronson FH. Seasonal variation in human reproduction: environmental factors [review]. Q Rev Biol 1995;70:141–64.
6. Levine RJ, Mathew RM, Chenault CB, Brown MH, Hurtt ME, Bentley KS, et al. Differences in the quality of semen in outdoor workers during summer and winter. N Engl J Med 1990;323:12–6.
7. Wellings K, MacDowall W, Catchpole M, Goodrich J. Seasonal variations in sexual activity and their implications for sexual health promotion. J R Soc Med 1999;92:60–4.
8. Warren CW, Gwinn ML, Rubin GL. Seasonal variation in conception and various pregnancy outcomes. Soc Biol 1986;33:116–26.
9. Martin JA, Hamilton BE, Ventura SJ, Menacker F, Park MM. Births: final data for 2000. Natl Vital Stat Rep 2002;50:1–101.
10. Curtin SC, Park MM. Trends in the attendant, place, and timing of births, and in the use of obstetric interventions: United States, 1989–97. Natl Vital Stat Rep 1999;47:1–12.
12. Cooperstock M, England JE, Wolfe RA. Circadian incidence of labor onset hour in preterm birth and chorioamnionitis. Obstet Gynecol 1987;70:852–5.
13. Walsh SW, Ducsay CA, Novy MJ. Circadian hormonal interactions among the mother, fetus, and amniotic fluid. Am J Obstet Gynecol 1984;150:745–53.
14. Lam DA, Miron JA. Global patterns of seasonal variation in human fertility. Ann NY Acad Sci 1994;709:9–28.