School of Public Health, University of California at Berkeley, Berkeley, CA 94720, USA
1 To whom correspondence should be addressed. Email: rayc{at}berkeley.edu
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: fetal death/population distress/September 11/sex ratio/very low birthweight
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Explanations of the reported association between population stressors and the human sex ratio typically build on the observation that males exceed females among embryonic and fetal deaths (Mizuno, 2000). The literature, however, remains vague on why the excess deaths of males in gestation apparently increase during periods of maternal stress. Speculation includes that the response of embryos and fetuses to maternal glucocorticoids determines whether or not gestation continues (Erickson et al., 2001
). These hormones perturb males in gestation more than females (Owen and Matthews, 2003
). Pregnant women produce more glucocorticoids when stressed (Hobel et al., 1999
), raising the possibility that some characteristic of embryonic or fetal response to high doses of the hormones increases the risk of death among weaker males (Erickson et al., 2001
).
The excess death theory implies that males lost to elevated maternal stress would have been at risk of premature delivery and, therefore, very low birthweight (<1500 g) had the environment remained constant. The theory explicitly posits that gestations likely to yield weak male infants have an elevated risk of spontaneous abortion. Premature, very low weight infants certainly must be considered among the weakest of newborns, given their elevated risk of death as well as of ongoing physical and behavioural deficits (Ross et al., 1991; Kiely et al., 1994
). Indeed, few would survive without considerable medical intervention. The risk of very low birthweight, moreover, increases with maternal stress and the sex ratio of very low weight infants significantly exceeds that of other live births (Hall and Carr-Hill, 1982
; Cooperstock and Campbell, 1996
; Vatten and Skjaerven, 2004
).
Although the work cited above assumes that the sex ratio drops during stressful times, controversy remains over the empirical and theoretical bases for the assumption. Empirical controversy arises from the fact that sex ratios reportedly increase during wars (MacMahon and Pugh, 1954; Graffelman and Hoekstra, 2000
), although decreases have been observed during recent conflicts (Ansari-Lari and Saadat, 2002
; Zorn et al., 2002
). The literature also reports that the sex ratio did not decline in rural Costa Rica during 19th century epidemics (Madrigal, 1996
). The chronic stress presumably induced by low socio-economic status, moreover, does not appear universally associated with low sex ratios (Marleau and Saucier, 2000
).
The theoretical literature does not agree on whether the termination of gestations likely to yield weak males is the purpose, or a side-effect, of the stress mechanism supposedly conserved by natural selection among females of sexually dimorphic species (Trivers and Willard, 1973; Brown and Silk, 2002
; Krackow, 2002
). Not all the literature, moreover, assumes that excess death of males in gestation induces the presumed decline in the secondary sex ratio during periods of environmental stress. Other explanations assume that temporal variation in the secondary sex ratio reflects earlier variation in the primary sex ratio (ratio of conceived males to females). Stress reportedly reduces sperm motility and could thereby reduce the sex ratio at conception (Fukuda et al., 1996
). Stress also reduces the frequency of coitus. Researchers have reported that coitus early in the human menstrual cycle favours the conception of males whereas that at the middle stage (i.e. ovulation) yields more females (Martin, 1997
; James, 1999
; Lazarus, 2002
). Circumstances, including cognitive stress, that decrease coital frequency should reduce the likelihood of conception early in the menstrual cycle and, therefore, reduce the primary sex ratio.
The events of September 11, 2001 as stressors of the US population
Much literature suggests that the events of September 11 induced anxiety and bereavement among Americans via the witnessing of harm (Galea et al., 2002; Schlenger et al., 2002
). Effects have been noted not only among Americans living near the attack sites but throughout the country (Schuster et al., 2004
; Whalen et al., 2004
). These reports converge with research demonstrating that witnessing harm to others induces biological responses in the witness that resemble those in the persons harmed (Singer et al., 2004
). These responses can include the production of glucocortocoids alluded to above as those implicated in the spontaneous abortion of males (Irwin et al., 1987
; Cohen et al., 1998
; Fukuda et al., 2000
).
Anxiety, fear, and their biological components could also trigger the reduced conception mechanisms. The reduction of sperm motility reported among men witnessing the effects of earthquakes could plausibly proceed from the events of September 11, 2001 (Fukuda et al., 1996). These events also could have reduced libido among bereaved males and females and thereby triggered the reduced coitus mechanism (National Center for Post Traumatic Stress Disorder, 2004
).
Hypotheses
The following hypotheses were tested.
The argument that population stressors reduce the secondary sex ratio would be supported if either or both hypotheses 1 and 2 were accepted. The fetal loss mechanism would be supported to the exclusion of the reduced conceptions mechanism if hypotheses 1, 3 and 4 were accepted but 2 rejected. The reduced conceptions mechanism would be supported to the exclusion of fetal loss if hypothesis 2 were accepted and the others rejected. The mechanisms do not exclude each other, implying that both would be supported if all the hypotheses were accepted.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
We test the first three hypotheses with time series beginning January 1989 and ending December 2002. These were the longest series with consistent definitions available to us at the time of the tests. Due to changes in definition (e.g. exclusion of therapeutic abortions), we tested hypothesis 4 with a time series that began January 1996. We end the series in December 2001 because the fetal loss theory offers no prediction for fetal deaths beyond that point.
A very low weight infant weighs <1500 g at birth (Kiely et al., 1994). Nearly all such infants are premature. A fetal death is defined by the State of California as a death prior to the complete expulsion or extraction from its mother of a product of human conception. The death is indicated by the fact that the fetus does not breathe or show any other evidence of life such as beating of the heart, pulsation of the umbilical cord, or definite movement of voluntary muscles (California Department of Health Services, 1996
).
California's Health and Safety Code requires that each fetal death, excluding induced abortions, in which the fetus has advanced to or beyond the 20th week of uterogestation shall be registered with the local registration district within 8 days following the delivery and prior to any disposition of the fetus. Only fetal deaths where the fetus has advanced to or beyond the 20th week of uterogestation shall be registered (California Department of Health Services, 1996). Our tests, therefore, do not apply to fetal deaths before the 20th week of gestation.
Any direct test of the excess death mechanism must use fetuses 20 weeks in gestation because the primary sex ratio (i.e. the ratio of male to female conceptions) cannot be known and deaths before 20 weeks cannot be accurately counted in California or elsewhere. Our test of hypothesis 4, therefore, may underestimate the association between the events of September 11 and the spontaneous abortion of males.
Monthly male and female fetal deaths averaged 151 and 139 respectively over the 72 months for which we had data. Monthly very low weight male and female births averaged 183 and 167 respectively over the 168 months of data available to us. Figures 1 and 2 show the secondary and very low weight birth sex ratios plotted over the test period. Figure 3 shows the fetal death sex ratio.
|
|
|
Time series often exhibit autocorrelation, such as trends, cycles, and the tendency to remain elevated or depressed after high or low values that complicate estimating the post-interruption values. Researchers dating at least to Fisher (1921) and his 1920 study of crop variation have solved the autocorrelation problem by decomposing time series into temporally predictable and residual components. The predictable values become those expected under the null hypothesis. Contemporary applications of Fisher's logic (e.g. Box et al., 1994
) typically decompose a time series by expressing autocorrelation as an effect of earlier values, or lags, of the series itself. If an equation includes the correct lags of the dependent variable, the residuals exhibit no autocorrelation. The analyst can then add the independent variable or variables to the equation to determine if their coefficients differ from zero in the hypothesized direction.
We used the strategy attributed to Dickey and Fuller (1979) as well as Box and Jenkins (Box et al., 1994
) to identify and model autocorrelation in male fetal deaths and very low weight births for the months January, 1996 through August 2001. The DickeyFuller test detects trends and seasonal cycles as well as their combination (e.g. upwardly or downwardly trending sine waves). The Box and Jenkins routines allow the researcher not only to model trends and cycles detected by the DickeyFuller test, but also to identify and model the tendency of a series to remain elevated or depressed, or to oscillate, after high or low values. The Box and Jenkins approach uses autoregressive and moving average parameters to model such tendencies. Autoregressive parameters best describe patterns that persist for relatively long periods, while moving average parameters parsimoniously describe less persistent patterns.
We converted the sex ratios into their natural logarithms to allow us to express our results in the familiar effect on odds format. The general form of the BoxJenkins equation applied to the natural logarithm of a sex ratio is as follows.
![]() | (1) |
We added a binary variable scored 1 for September, 2001 and 0 otherwise to the best fitting BoxJenkins equations. We specified the variable such that we could detect associations with cohorts born not only in September 2001 but, depending on the hypothesis, up to 10 months later. Testing hypotheses 1 and 2, for example, required that we specify up to 10 months because the fetal loss mechanism would predict decreased sex ratios among cohorts likely in gestation in September (i.e. those born from September, 2001 through April 2002) while the reduced conception mechanism predicts reduced sex ratios in June and July, 2002. Reduced sex ratios in May 2002 could be attributed to either mechanism.
Hypothesis 3 arises from the fetal loss mechanism and, therefore applies only to birth cohorts in gestation in September 2001. The test, therefore, requires that we specify the binary variable for 8 months in addition to September 2001. Hypothesis 4 concerns the fetal death sex ratio. Since public health surveillance systems register fetal deaths only after the 19th week of gestation, testing hypothesis 4 requires that the binary variable be specified for only 3 months after September 2001.
We estimated test equations that took the following general specification.
![]() | (2) |
We estimated final models by deleting from the initial models any variables with estimated parameters inside their 95% confidence interval (two-tailed test). We inspected the residuals of each model for autocorrelation and added BoxJenkins parameters to the equation to express any that we discovered.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
Table II shows the results for hypothesis 3 concerning the sex ratio of very low weight births. The findings support hypothesis 3 in that the sex ratio was unexpectedly low in December 2001. The antilog of the coefficient suggests that the odds of very low weight birth being male dropped by 29% in December.
|
Table III shows the results for hypothesis 4 concerning the sex ratio of fetal deaths. The results shown in Table III support hypothesis 4 in that the fetal death sex ratio exceeded expected values in October and November of 2001. The antilogs of these parameters suggest that the odds of a fetal death being male increased in October and November by 24 and 23% respectively.
|
We conducted three additional tests suggested by our primary findings. The fetal loss mechanism assumes that more male, but not female, fetuses die than expected during periods of maternal stress. To ensure that our sex ratio findings arose from excess male deaths rather than unexpectedly few deaths among females, we modelled male and female fetal deaths separately. Consistent with the fetal loss mechanism, and with the findings for the fetal death sex ratio, significantly more (i.e. =23; SE = 11) male fetuses than expected died in October 2001 and the increase remained elevated into November. Fewer female fetuses than expected died in October but the difference was not significant (i.e.
=13.1844; SE = 13.2197).
We also modelled the natural logarithm of the sex ratio of infants weighing >1499 g to determine if our findings for the secondary sex ratio resulted, as the fetal loss theory implies, primarily from losses of fetuses that, had they survived to birth, would have been very low weight. As with the secondary sex ratio, and the sex ratio for very low weight births, the sex ratio for other than very low weight births showed a lower than expected value in December, 2001. That value, however, did not fall below the 95% confidence interval (i.e. =0.0175; SE = 0.0100).
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Our findings offer no support for the decreased conception mechanism. The secondary sex ratio did not vary from expected values among cohorts conceived in September, October or November of 2001.
Our findings may not generalize beyond California. We know of no reason why Californians would have been more reactive to the events of September 11 than the remainder of the country. However, given that they live at relatively great distance from the attacks, they may have been less reactive. Additional tests should estimate any dose response associated with distance from the attacks.
The welcomed rarity of natural and man-made disasters raises the question of whether findings such as ours generalize to other, more common phenomena that perturb entire populations. As noted above, for example, most of the war literature (e.g. Graffelman and Hoekstra, 2000) is at odds with that concerning disasters. While more empirical work should be done to estimate how much of the divergence results from methodological artefacts, we also need a theory that explains the divergent findings. Could it be, for example, that wars vary in the degree to which they are anticipated, promise relief from other adversity or unite populations, and therefore in the degree to and manner in which populations habituate to them? These and similar considerations raise the question of whether fetal loss theory applies to wars.
As noted in Materials and methods, any test of the excess death mechanism must use fetuses 20 weeks in gestation. The loss of males in gestation due to maternal stress could exceed our estimates because the mechanism, at least in theory, applies to all males in gestation. Our tests of hypotheses 1 and 3, however, did not show decreased sex ratios for cohorts born between January and May 2002, suggesting that the events of September 11, 2001 did not induce significant loss of males in gestation <20 weeks.
Changes in the secondary sex ratio may not warrant public health concern, but the public health community should be concerned with mechanisms that induce such change. Increases in the incidence of male fetal deaths, for example, suggest that the biology of at least women of childbearing age can be altered by population stressors.
Those responsible for public health in the USA have responded to the attacks of September 11, 2001 by preparing for similar events (Institute of Medicine, 2002). These efforts understandably focus primarily on the populations most exposed to toxins, explosions, and other mayhem. The Institute of Medicine (2003)
, however, has urged planning for the needs of those who would be cognitively, rather than physically, exposed to such attacks. Our findings, if replicated in other states distant from the attacks, should add impetus to this planning.
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Box G, Jenkins G and Reinsel G (1994) Time Series Analysis: Forecasting and Control, 3rd edn. Prentice-Hall, London.
Brown G and Silk J (2002) Reconsidering the null hypothesis: is maternal rank associated with birth sex ratios in primate groups? Proc Natl Acad Sci USA 99, 1125211255.
California Department of Health Services, Office of Vital Records (1996) Handbook for Local Registrars of Births and Deaths. Sacramento, CA.
Catalano R (2003) Sex ratios in the two Germanies: a test of the economic stress hypothesis. Hum Reprod 9, 19721975.[CrossRef]
Catalano RA and Bruckner T (2005) Economic antecedents of the Swedish sex ratio. Soc Sci Med 60, 537543.[CrossRef][ISI][Medline]
Cohen S, Frank E, Doyle WJ, Skoner DP, Rabin BS and Gwaltney JM (1998) Types of stressors that increase susceptibility to the common cold in healthy adults. Health Psychol 17, 214223.[CrossRef][ISI][Medline]
Cooperstock M and Campbell J (1996) Excess males in preterm birth: interactions with gestational age, race, and multiple birth. Obstet Gynecol 88, 189193.
Dickey D and Fuller W (1979) Distribution of the estimators for autoregressive time series with a unit root. J Am Statist Assoc 74, 427431.[ISI]
Erickson K, Thorsen P, Chrousos G, Grigoriadis DE, Khongsaly O, McGregor J and Schulkin J (2001) Preterm birth: associated neuroendocrine, medical, and behavioral risk factors. J Clin Endocrinol Metab 86, 25442552.
Fisher RA (1921) Studies in crop variation: an examination of the yield of dressed grain from Broadbalk. J Agri Sci 11, 107135.
Fukuda M, Fukuda K, Shimizu T, Yomura W and Shimizu S (1996) Kobe earthquake and reduced sperm motility. Hum Reprod 11, 12441246.[Abstract]
Fukuda M, Fukuda K, Shimizu T and Moller H (1998) Decline in sex ratio at birth after Kobe earthquake. Hum Reprod 13, 23212322.[Abstract]
Fukuda S, Morimoto K, Mure K and Maruyama S (2000) Effect of the Hanshin-Awaji earthquake on posttraumatic stress, lifestyle changes, and cortisol levels of victims. Arch Environ Health 55, 121125.[ISI][Medline]
Galea S, Ahern J, Resnick H, Kilpatrick D, Bucuvalas M, Gold J and Vlahov D (2002) Psychological sequelae of the September 11 terrorist attacks in New York City. New Engl J Med 346, 982987.
Graffelman J and Hoekstra R (2000) A statistical analysis of the effect of warfare on the human secondary sex ratio. Hum Biol 72, 433445.[ISI][Medline]
Hall MH and Carr-Hill R (1982) Impact of sex ratio on onset and management of labour. Br Med J (Clin Res Ed) 285, 401403.[Medline]
Hansen D, Moller H and Olsen J (1999) Severe periconceptional life events and the sex ratio in offspring: follow up study based on five national registers. Br Med J 319, 548549.
Hobel C, Dunkel-Schetter C, Roesch S, Castro L and Arora C (1999) Maternal plasma corticotropin-releasing hormone associated with stress at 20 weeks' gestation in pregnancies ending in preterm delivery. Am J Obstet Gynecol 180, 257263.[ISI]
Institute of Medicine (2002) Biological threats and terrorism: assessing the science and response capabilities. In Knobler SL, Mahmoud A, and Pray LA (eds) Workshop Summary. National Academy Press, Washington, DC.
Institute of Medicine (2003) Preparing for the Psychological Consequences of Terrorism: A Public Health Strategy. National Academy Press, Washington, DC.
Irwin MR, Vale W and Britton KT (1987) Central corticotropin-releasing factor suppresses natural killer cytotoxicity. Brain Behav Immun 1, 8187.[CrossRef][Medline]
James WH (1999) The status of the hypothesis that the human sex ratio at birth is associated with the cycle day of conception. Hum Reprod 14, 21772178.
Kiely JL, Brett KM, Yu S and Rowley DL (1994) Low birth weight and intrauterine growth retardation. In Wilcox LS and Marks JS (eds) From Data to Action: CDC's Public Health Surveillance for Women, Infants, and Children, 3. US Department of Health and Human Services, Washington, DC, pp. 185202.
Krackow S (2002) Why parental sex ratio manipulation is rare in higher vertebrates (invited article). Ethology 108, 10411056.[CrossRef][ISI]
Lazarus J (2002) Human sex ratios: adaptations and mechanisms, problems and prospects. In Hardy I (ed.) Sex Ratios. Concepts and Research Methods. Cambridge University Press, Cambridge, UK, pp. 287311.
Lyster W (1974) Altered sex ratio after the London smog of 1952 and the Brisbane flood of 1965. J Obstet Gynaecol Br Commonw 81, 626631.[ISI][Medline]
Madrigal L (1996) Sex ratio in Escazu, Costa Rica, 18511901. Hum Biol 68, 427436.[ISI][Medline]
Marleau JD and Saucier JF (2000) Pregnant women's social status, stress, self-esteem, and their infants' sex ratio at birth. Percept Mot Skills 91, 697702.[ISI][Medline]
Martin JF (1997) Length of the follicular phase, time of insemination, coital rate and the sex of offspring. Hum Reprod 12, 611616.[CrossRef][ISI][Medline]
MacMahon B and Pugh TF (1954) Sex ratio of white births in the United States during the Second World War. Am J Hum Genet 6, 284292.[ISI][Medline]
Mizuno R (2000) The male/female ratio of fetal deaths and births in Japan. Lancet 356, 738739.[CrossRef][ISI][Medline]
Mocarelli P, Brambilla P, Gerthoux PM, Patterson DG and Needham LL (1996) Change in sex ratio with exposure to dioxin. Lancet 348, 409.[CrossRef][ISI][Medline]
National Center for Post Traumatic Stress Disorder (2004) Effects of Traumatic Stress in a Disaster Situation, A National Center for PTSD Fact Sheet, Normal Reactions to an Abnormal Situation (Accessed April 2, 2004 at http://www.ncptsd.org/facts/disasters/fs_effects_disaster.html?printable = yes.).
Owen D and Matthews SG (2003) Glucocorticoids and sex-dependent development of brain glucocorticoid and mineralocorticoid receptors. Endocrinology 144, 27752784.
Ross G, Lipper EG and Auld PA (1991) Educational status and school-related abilities of very low birth weight premature children. Pediatrics 88, 11251134.[Abstract]
Schlenger WE, Caddell JM, Ebert L, Jordan BK, Rourke KM, Wilson D, Thalji L, Dennis JM, Fairbank JA and Kulka RA (2002) Psychological reactions to terrorist attacks: findings from the National Study of Americans' Reactions to September 11. J Am Med Assoc 288, 581588.
Schuster MA, Stein BD, Jaycox L, Collins RL, Marshall GN, Elliott MN, Zhou AJ, Singer T, Seymour B, O'Doherty J et al. (2004) Empathy for pain involves the affective but not sensory components of pain. Science 303, 11571162.
Singer T, Seymour B, O'Doherty J et al. (2004) Empathy for pain involves the affective but not sensory components of pain. Science 303, 11571162.
Trivers R and Willard D (1973) Natural selection of parental ability to vary the sex ratio of offspring. Science 179, 9092.[ISI][Medline]
Vatten LJ and Skjaerven R (2004) Offspring sex and pregnancy outcome by length of gestation. Early Hum Dev 76, 4754.[CrossRef][ISI][Medline]
Whalen CK, Henker B, King PS, Jamner LD and Levine L (2004) Adolescents react to the events of September 11, 2001: focused versus ambient impact. J Abnorm Child Psychol 32, 111.[CrossRef][ISI][Medline]
Zorn B, Sucur V, Stare J and Meden-Vrtovec H (2002) Decline in sex ratio at birth after 10-day war in Slovenia. Hum Reprod 17, 31733177.
Submitted on May 17, 2004; resubmitted on September 9, 2004; accepted on January 7, 2005.