1
Effect of mothers’ occupational stress on child sex at birth
2
3
4
*K.E. Ruckstuhl, G. P. Colijn, Volodymyr Amiot and Erin Vinish, University of Calgary,
5
Department of Biological Sciences, 2500 University Drive NW, Calgary, Alberta, T2N
6
1N4, Canada. *=Corresponding author’s address: same as above. Phone: +1403 220-
7
8776, fax: +1403 289-9311, email:kruckstu@ucalgary.ca.
8
9
Running header: stress and sex ratio in humans
10
11
12
13
1
14
Abstract
15
Many women are working outside of the home, occupying a multitude of jobs with
16
varying degrees of physiological or psychological stress. Stress can lead to depression,
17
heart conditions, insomnia, immune suppression, and could affect birth sex ratios. We
18
investigated whether occupational stress in women affects child sex at birth, with the
19
prediction that women in high stress occupations would be more likely to give birth to a
20
daughter than a son. Daughters are less vulnerable to unfavourable conditions during
21
conception, pregnancy and after parturition, and are less costly to carry to term. Women
22
in high stress occupations gave more often birth to daughters, whereas women in low
23
stress occupations had a male bias in offspring. We also investigated whether maternal
24
age, or her partner’s income, could be biasing offspring sex. We found no direct effect of
25
the mother’s age, her partner’s job stress or partner income on child sex but an interaction
26
between job stress and partner income. Partner income cancelled the effect of maternal
27
stress at moderate income levels, and reversed it at high income levels. To our knowledge
28
this is the first paper reporting the effects of job stress on offspring sex ratio and the
29
mitigating effect of partner’s income.
30
31
Key words: occupational stress, offspring gender, humans, sex ratio
32
2
33
Introduction
34
Stress is omnipresent in our everyday lives. Numerous studies have shown that
35
psychological and physiological stress can increase the risk of heart attacks [1, 2] can
36
result in slower recovery from injuries [3], lead to depression and anxiety, suppressed
37
cell-mediated immunity [4], preterm deliveries [5, 6], or early foetal abortions [6-8]. For
38
example, offspring sex-ratio biases have been observed in human populations under a
39
variety of stress-inducing circumstances such as serious personal life events[7, 8],
40
economic collapse[9], earthquakes [10], caloric deprivation [11]and war [12]. On the
41
other hand, a study on Sprague Dawley lab rats (Rattus spec.) found no effects of
42
psychological stress on sex ratios, but marked effects on their offspring [13]. Offspring
43
born to behaviourally stressed mothers reacted differently to stimuli presented to them
44
later in life than those born to mothers that had not been stressed immediately before
45
conception [13]. Differences in these studies might be due to different effects of
46
psychological versus physiological stress or the timing, extent and severity of the stress,
47
and it’s not always clear if they can be considered completely separate from one another.
48
Khashan et al. (2009), for example, tried to replicate a Danish study [7], in which severe
49
life events were suggested to cause changes in sex ratio through changes in sexual
50
behaviour. However, they were not able to replicate the study when using an extended
51
data set, which spanned the original years of Hansen et al.’s (1999) study but also
52
included more recent years [8]. Khashan et al. (2009) noted that it might have been the
53
economic recession, in combination with severe life events, and not solely severe life
54
events, which could have caused the sex ratio to change. Despite mounting evidence of
55
the potential effect of stress on sex ratios in humans, very little is known about the
3
56
potential mechanisms and ultimate causes for a shift in sex ratio at birth, and some of
57
these are debated [8]. Proximate explanations for this shift in the sex ratio include the
58
following: differential sperm motility due to psychological stress in men (e.g. Kobe
59
earthquake destroyed men’s houses and killed their family members) [14, 15], Y-bearing
60
sperm may be faster but less resilient to unfavourable conditions in the reproductive tract
61
than X-bearing sperm [14, 16] and spontaneous abortions may be biased towards males
62
[16], whereby stress hormones (such as corticosteroids) could cause the abortion of more
63
male than female foetuses [17, 18]. Ultimate explanations include the Trivers-Willard
64
hypothesis, which suggests that vertebrate females subjected to physiological stress gain
65
a selective advantage by producing female offspring: male offspring are thought to be
66
more costly to produce and raise and are less likely to attain a high social status and
67
lifetime reproductive success if born to a stressed, subordinate, female [18-20]. A study
68
on women in a food-stressed rural community in southern Ethiopia, showed that, women
69
whose most recent offspring was male, had significantly higher body, fat and muscle
70
mass than other women, which led them to conclude that stronger mothers bear more
71
sons than those in worse condition [21]. Hence, physiologically stressed females would
72
hence be better off producing female offspring under conditions of stress, as daughters
73
are more likely to survive than sons under stress [21], and the reproductive success of
74
daughters is usually not dependent on social status. Therefore, under stressful conditions,
75
females should selectively choose sperm carrying an X chromosome, or abort male
76
foetuses, if possible. One mechanism that would let females bias the offspring sex in their
77
favour is the ability to abort males when the mother is in poor condition [11]. Then the
78
female would have another opportunity to conceive a daughter, or a son in better
4
79
condition, provided her own condition improved in the interim. Another assumption of
80
the Trivers-Willard hypothesis is that there is minimal parental investment by the male,
81
beyond contributing sperm. Is assumption is less applicable to humans, because males
82
often invest parental care in their young, which reduces the variance of their reproductive
83
success [18]. Many studies have been carried out trying to determine exactly how
84
applicable the Trivers-Willard hypothesis is to humans [22, 23], and what kind of factors
85
have an influence on the human offspring sex ratio[8, 9, 11, 24-28]. Many researchers
86
have found strong support for the Trivers-Willard hypothesis [28-32].
87
A study on the offspring sex in Norwegian military air pilots [26] has shown that
88
these pilots are more likely to have daughters than sons. Another study, of a British
89
population, showed that women who had a lower perceived life expectancy were more
90
likely to give birth to a daughter than a son [33]. , Despite an increasing number of
91
women opting to have a career or to hold a job and have a family, which can be
92
extremely stressful, research on female occupational stress (which can be considered a
93
form of chronic stress), and how it could impact sex ratio at birth is lacking. If stress
94
(physiological or psychological) increases the risk of spontaneous abortions of male
95
foetuses [4, 5, 16], we predict that women in high stress occupations are more likely to
96
give birth to a daughter than to a son, while women in low stress occupations should have
97
a probability of giving birth to a son or daughter that is close to the world average. The
98
world-wide proportion of male versus female births is at around 51.7% males and 48.3%
99
females, or 105-107 males per 100 females [25], and between 104 and105 males per 100
100
females in the UK, during the time of this study
101
(http://www.statistics.gov.uk/STATBASE).
5
102
Maternal age is another factor that could affect sex of offspring [34, 35]. Younger
103
women who have had at least one child, will already be more likely to produce boys [35],
104
whereas women with a lower perceived life expectancy, or older women might be more
105
likely to have daughters [33]. Younger women who are presumably in better condition
106
than those who are older, and women of higher parity have more vascularised uteri,
107
which leads to infants of higher birth weight [35]. Almond and Edlund (2007) studied the
108
effect of maternal age on offspring sex, and also found that younger women are more
109
likely than older women to give birth to sons.
110
Parental economic status and education are also useful measures of parental
111
condition or the potential to invest resources into their offspring. Therefore it can be
112
assumed that individuals who earn more are more capable of satisfying their basic needs
113
than those who earn less, and as a result will be in better condition. These assumptions
114
are supported by the results of studies that have analysed the effects of parental economic
115
status and education on offspring sex [22]. Koziel & Ulijaszek (2001) found partial
116
evidence of greater investment in female offspring at the lowest level of paternal
117
education and greater investment in males at higher levels of paternal education.
118
Similarly, a study on church rank in Mormons in the US has found that women married to
119
high ranking men where more likely to have sons than women married to lower ranking
120
men [31]. Almond & Edlund (2007), who used education as one indicator of condition,
121
found that better educated women had more sons. Finally, partner job stress could also
122
have an effect on the psychological stress experienced by women and therefore have an
123
additive effect on the likelihood of giving birth to a girl or boy. In this study, we
124
investigated how maternal job stress, her age, primiparity (yes or no) her partner’s job
6
125
stress and partner income affect the sex of their child at birth. We hypothesized that job
126
stress (psychological stress), maternal age, her partner’s income (economic status) and
127
her partner’s job stress affects the likelihood of giving birth to a particular gender. We
128
predicted that women with high stress jobs will be more likely to give birth to a daughter
129
than a son, and that maternal age affects the sex of her offspring, biasing it towards
130
daughters with older age. We also predicted that partner stress and partner income would
131
affect sex at birth, with the expectation of a bias towards daughters for families with
132
higher paternal stress and/or lower paternal income.
133
134
Methods
135
We investigated the potential effects of maternal occupational stress, age, primiparity,
136
partner’s job stress and income on offspring sex ratio in humans, using 16,384 incidences
137
of birth from a six-year (2000 to 2005 inclusive) childbirth dataset from Addenbrooke’s
138
Hospital in Cambridge, UK. Women, who were admitted to the hospital for delivery were
139
asked to fill in a form which contained questions about their age, job, partners job,
140
whether or not they already had children, etc. This form was stored and entered into a
141
central database and information on the birth (gender, weight, health etc.) of offspring
142
was added to this file. We obtained a restricted data set from Addenbrooke’s hospital
143
with: maternal age (mean= 30.94 years, SD= 5.34 years, Range= 13-53 years old,
144
N=16,345 women of known age), maternal and paternal occupations, and whether or not
145
the child was first-born, a singleton or twin (we only looked at singleton births to avoid
146
bias due to in vitro fertilization and possible hereditary effects). Consent to use and
147
publish the data was obtained from the human ethics board of the University of Calgary,
7
148
and the Addenbrooke’s hospital (and Trust) in Cambridge, UK, after the nature and
149
possible consequences of the studies were explained. All data were derived from
150
anonymous subjects, whose identities were fully protected and cannot be revealed. If data
151
were missing, about income or job description, or if there was no partner listed we
152
excluded the data from the analyses. The study was conducted in the spirit of the
153
Helsinki Declaration of 1975, as revised in 2000 (5).
154
We were unable to get a direct measure of maternal stress or paternal stress
155
hormone levels (physiological stress). Instead we categorized job stress (psychological
156
stress) by consulting several earlier studies that have assessed how stressful people
157
perceive certain types of jobs (House 1974, the website of the Centre for Occupational
158
and Health Psychology School of Psychology, Cardiff University’s report (265/2000) and
159
[36]. Jobs were categorized by lumping them into general job categories according to
160
Marchand (2007), who had categorized job types according to their perceived stress
161
levels (how stressful people perceived their jobs to be). Marchand (2007) used stress
162
levels between 1 and 10, with 1 equalling the lowest and 10 the highest amount of
163
psychological stress (see Table 1). Stay-at-home mothers and women working for the
164
armed forces were not listed in Marchand’s (2007) study. Therefore, we decided to list
165
these two professions under the “Health care” job category. The “Health care” job
166
category in Marchand’s (2007) study had a stress rating of 8 out of 10. The two
167
occupations were added to that category because of the nature of the occupations: both
168
are associated with high levels of stress, since they involve caring for dependents (similar
169
to health occupations), or dealing with conflict. So if the job were “night nurse” for
170
example, then this job would be placed in the “Health care” sector which has a stress
8
171
rating of 8. If the job was described as “clerk” in a store the job would be rated under the
172
category “Sales, services” and given a stress rating of 5. In the end, we had 10 general job
173
categories with associated stress levels in Table 1. We assumed that psychological stress
174
affects physiological stress and puts an increased physical burden on the expectant
175
mother.
The partner’s job stress was assigned using the same criteria as in table 1. The
176
177
partner’s average income was calculated using information gathered by the 2007 Annual
178
Survey of Hours and Earnings that was done by the UK government [37]. For our study,
179
we used the median gross annual income for men in 2007, according to their occupation.
180
In a second step, we then investigated how the maternal age, primiparity, job stress,
181
partner’s job stress and income affected offspring sex at birth (see statistical analyses
182
below.
183
184
Statistical analyses
185
In the first model, we simply compared the offspring sex of women in high (stress levels
186
9 and 10) versus low stress jobs (stress levels 1 and 2), using a Likelihood ratio test (Chi-
187
square), comparing numbers of occurrence. In the second model we included the
188
following predictor (independent) variables and whether or not they affect the offspring
189
sex (dependent variable): maternal occupational stress (between 1 and 10), maternal age,
190
partner stress (between 1 and 10) and income, number of children born, and whether or
191
not the child was the first born. For the second model, we used a generalized linear
192
model, with a binomial distribution and a logit link function, using R statistical software
193
(version 2.6.1; [38] and likelihood ratio tests [39] software. Subsequently, we eliminated
9
194
all non-significant variables from the model using a step-wise approach. Maternal age did
195
not affect the likelihood of giving birth to a son or daughter (Logistic regression: Wald
196
statistics= 0.42, p= 0.52, N= 16,335 births), but older mothers were more likely to occupy
197
higher stress occupations (Logistic regression: Wald statistics= 84.07, p= 0.0001). In all
198
subsequent models we left maternal age in as a covariate.
199
200
Results
201
Across the entire dataset, mothers in low stress occupations (stress levels 1 and 2) were
202
more likely to have a son than a daughter (53.68% sons), compared to mothers in high
203
(stress levels 9 and 10) stress occupations (47.05% sons, Chi-square= 7.85, p= 0.0051).
204
The same effect on offspring sex was evident when we analysed first-borns only, with
205
53.9% male births in low-stress occupations compared to 50.3% males in high-stress
206
occupations (Chi-square= 29.86, p= 0.001). Chi-square analyses were done on counts but
207
Table 1 lists total number of births and sex ratio as a percent value.
208
In the model with job stress levels ranging between 1 (lowest) and 10 (highest),
209
neither partner income (Likelihood ratio χ2 = 1.08, p= 0.2978), nor partner job stress
210
(Likelihood ratio χ2 = 14.7224, p= 0.0988) had a significant effect on the probability of
211
male offspring, but both maternal stress ((Likelihood ratio χ2 = 19.87, p= 0.0187) and the
212
interaction of maternal stress and partner income ((Likelihood ratio χ2 = 18.53, p=
213
0.0295) did have a significant effect (see Table 2 for more detailed statistics and
214
comparisons). As maternal stress increased, the probability of having male offspring
215
decreased. When plotting the interaction of maternal stress and partner income, it was
216
evident that the majority of the partners fall into either the low or moderate income
10
217
levels, with few at the higher income level (Fig. 1). When partner income was less than
218
29,991 £, increased maternal stress decreases the probability of male offspring (Fig. 1).
219
At 29,991 £, the effect of partner income cancelled the effect of maternal stress, and there
220
was an equal probability of male offspring at all maternal stress levels, which is 0.509
221
(Fig. 1). At more than 29,991 £, there was an inversion of the effect of maternal stress
222
(Fig. 1).
223
224
Discussion
225
Previous research on effects of stress on female reproduction found that it can cause
226
abortions, preterm deliveries and potentially bias the offspring sex [4, 5, 40]. Obel et al.
227
(2007) found that even moderate psychological distress in Danish women, who were
228
monitored from early pregnancy through to term, also decreased the sex ratio at birth in
229
favour of females. Our study is the first to show that job stress (which could be
230
categorized as chronic psychological stress) in women can affect the sex of their
231
offspring. As predicted, increased stress in women was associated with an increased
232
chance of having daughters.
233
Partner income itself was not a significant variable in explaining sex ratios but we
234
found an interaction of partner income and maternal job stress on sex ratio. When partner
235
income was high (specifically above 29,991 £), sex of offspring was biased towards sons,
236
as predicted. However, when partner income was less than 29,991 £, the level of maternal
237
stress had the dominant effect on the probability of male offspring. Thus it seems, that the
238
economic and social status, of males, are important in alleviating direct effects of female
239
occupational stress on offspring sex. Social status (wealth, church rank, and number of
11
240
other wives) was also reported to be an important factor in contributing to sex ratios in a
241
Mormon community in the US [28]. Women of men who were high ranking in the
242
church, rich and who had several wives were more likely to bear sons than women
243
married to lower status men.
244
The nature of the stress experienced (i.e. acute versus chronic stress) might in
245
itself significantly affect birth sex ratios. For example, s study on immune responses in
246
Sprague Dawley laboratory rats, for example, has shown that acute stress acts as a cell-
247
mediated immuno-enhancement while chronic stress can lead to immuno-suppression and
248
therefore elevated susceptibility to disease[41]. Thus, it is possible that chronic stress
249
biases sex ratios towards females, while acute stress could be irrelevant or even favour
250
male offspring. Detailed studies differentiating between types and levels of stress are
251
necessary to tease out possible mechanisms. Analyses and assessment of stress levels for
252
job categories and particularly subcategories in our study were to a large extent arbitrary
253
as we were not able to measure stress directly but had to resort to studies about perceived
254
stress levels in different occupations. Therefore, it would be interesting to measure and
255
monitor stress directly, in different women, working different jobs, rather than having to
256
resort to perceived job stress categories. Experiments done on starlings (Sturnus
257
vulgaris), for example, indicate that high stress levels can indeed bias the sex ratio at
258
hatching [23]. Females who were injected with stress hormones had a higher proportion
259
of female chicks than control females. Starling mothers had equal sex ratios in eggs laid
260
but most of the male embryos died before hatching and the ones that hatched had low
261
immune response and low survival, therefore biasing hatching sex ratio towards females.
262
While such studies show that stress can effectively bias the survival of one offspring sex
12
263
over the other, our results suggest that high occupational stress in woman has the
264
potential to bias the sex ratio at birth in humans in favour of daughters. The mechanisms
265
behind this phenomenon remain unclear. Evidence for both pre- and post-coital
266
mechanisms exists but further research in this area is needed to investigate the potential
267
pathways in more detail.
268
Understanding what can bias the human sex ratio is especially important because
269
there are already a number of existing factors, which influence sex ratio, and a number of
270
new factors, which are just beginning to be identified, such as latitude[42], floods or
271
smog [43], global warming [44], extreme life events[24], and as shown in our own study,
272
job stress, particularly, women’s job stress. Stress during conception and pregnancy are
273
therefore likely candidates responsible for the decreasing sex ratios observed in many
274
Western countries [45], in which women opt to have families, a job and a career.
275
276
Acknowledgement
277
We would like to thank Elissa Cameron, Tim Clutton-Brock, Peter Neuhaus, Angela
278
Aivaz, and Fanie Pelletier for suggestions and comments on earlier drafts, and the
279
Cambridge University Hospitals NHS Foundation Trust, Michelle Ellerbeck (Data
280
awareness manager), Addenbrooke’s hospital and all anonymous women for the
281
permission and availability of the childbirth data set. We would also like to thank the
282
University of Calgary, and the Ethics Board for permission to carry out this research, and
283
an NSERC discovery grant to KER for financial support while writing this paper.
284
285
Competing interests
13
286
The authors declare that they have no competing interests.
287
288
Authors’ contributions
289
KER conceived of and led the study, acquired the permits and data set, worked on the
290
data set, did most of the analyses and writing of the paper. GPC, VA and EV worked on
291
the data set, assigned stress categories, did part of the analyses and contributed to the
292
writing of this paper.
293
14
294
References
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
House JS: Occupational stress and coronary heart disease: a review and
theoretical integration. Journal of health and social behavio1r 1974, 15:12-27.
Wamala SP, Mittleman MA, Horsten M, Schenck-Gustafsson K, Orth-Gomer K:
Job stress and the occupational gradient in coronary heart disease risk in
women - The Stockholm Female Coronary Risk Study. Social Science &
Medicine 2000, 51:481-489.
von Thiele U, Lindfors P, Lundberg U: Self-rated recovery from work stress
and allostatic load in women. Journal Of Psychosomatic Research 2006,
61:237-242.
Dole N, Savitz DA, Hertz-Picciotto I, Siega-Riz AM, McMahon MJ, Buekens P:
Maternal stress and preterm birth. American Journal Of Epidemiology 2003,
157:14-24.
Nepomnaschy PA, Welch KB, McConnell DS, Low BS, Strassmann BI, England
BG: Cortisol levels and very early pregnancy loss in humans. PNAS 2006,
103:3938-2942.
Mulder EJH, Robles de Medina PG, Huizink AC, Van den Bergh BRH, Buitelaar
JK, Visser GHA: Prenatal maternal stress: effects on pregnancy and the
(unborn) child. Early Human Development 2002, 70:3-14.
Hansen D, Moller H, Olsen J: Severe periconceptional life events and the sex
ratio in offspring: follow up study based on five national registers. British
Medical Journal 1999, 319:548-549.
Khashan AS, Mortensen PB, McNamee R, Baker PN, Abel KM: Sex ratio at
birth following prenatal maternal exposure to severe life events: a
population-based cohort study. Human Reproduction 2009, 24:1754-1757.
Catalano RA: Sex ratios in the two Germanies: a test of the economic stress
hypothesis. Human Reproduction 2003, 18:1972-1975.
Fukuda M, Fukuda K, Shimizu T, Moller H: Decline in sex ratio at birth after
Kobe earthquake. Human Reproduction 1998, 13:2321-2322.
Williams RJ, Gloster SP: Human Sex-Ratio As It Relates To Caloric
Availability. Social Biology 1992, 39:285-291.
Zorn B, Sucur V, Stare J, Meden-Vrtovec H: Decline in sex ratio at birth after
10-day war in Slovenia. Human Reproduction 2002, 17:3173-3177.
Shachar-Dadon A, Schulkin J, Leshem M: Adversity before conception will
affect adult progeny in rats. Developmental Psychology 2009, 45:9-16.
Hingorani V, Shroff G: Natural Sex Selection For Safe Motherhood And As A
Solution For Population-Control. International Journal Of Gynecology &
Obstetrics 1995, 50:S169-S171.
Fukuda M, Fukuda K, Shimizu T, Yomura W, Shimizu S: Kobe earthquake and
reduced sperm motility. Human Reproduction 1996, 11:1244-1246.
Byrne J, Warburton D: Male Excess Among Anatomically Normal Fetuses In
Spontaneous-Abortions. American Journal Of Medical Genetics 1987, 26:605611.
Hobel CJ, Dunkel-Schetter C, Roesch SC, Castro LC, Arora CP: Maternal
plasma corticotropin-releasing hormone associated with stress at 20 week's
15
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
gestation in pregnancies ending in pretrm delivery. American Jounral of
Obstetrics and Gynecology 1999, 180:6.
Trivers RL, Willard DE: Natural selection of parental ability to vary sex ratio
offspring. Science 1973, 179:90-92.
Clutton-Brock TH, Albon SD, Guinness FE: Parental investment in male and
female offspring in polygynous mammals. Nature 1981, 289:487-489.
Clutton-Brock TH, Albon SD, Guinness FE: Great expectations: dominance,
breeding success and offspring sex ratio in red deer. Anim Behav 1986,
34:460-471.
Gibson MA, Mace R: Strong mothers bear more sons in rural Ethiopia.
Biology letters 2003, 270:S108-S109.
Koziel S, Ulijaszek SJ: Waiting for Trivers and Willard: do the rich really
favor sons? Am J Phys Anthropol 2001, 115:71-79.
Love OP, Chin EH, Wynne-Edwards KE, Williams TD: Stress hormones: A link
between maternal condition and sex-biased reproductive investment.
American Naturalist 2005, 166:751-766.
Catalano R: Exogenous shocks to the human sex ratio: the case of Sptember
11, 2001 in New York City. Human reproduction 2006, 21:5.
Hesketh T, Xing ZW: Abnormal sex ratios in human populations: Causes and
consequences. Proceedings Of The National Academy Of Sciences Of The United
States Of America 2006, 103:13271-13275.
Irgens A, Irgens LM: Male proportion in offspring of military air pilots in
Norway. Norsk Epidemiologi 1999, 9:47-49.
J. NK: Humans at tropical latitudes produce more females. Biology letters
2009.
Mackey WC: Relationships between human sex ratio and the woman's
microenvironment: 4 tests. Human Nature 1993, 4:175-198.
Gaulin SJC, Robbins CJ: Trivers-willard effect in contemporary North
American society. American Journal of Physical Anthropology 1991, 85:61-69.
Bereczkei T, Dunbar RIM: Female-biased reproductive strategies in a
Hungarian gypsy popultion. Proceedings of the Royal Society, London B 2001,
264:17-22.
Mealey L, Mackey W: Variation in offspring sex ratio in women of differing
social status. Ethology and Sociobiology 1990, 11:83-95.
Grant VJ: Maternal personality, evolution and the sex ratio. Londoon; 1998.
Johns SE: Subjective life expectancy predicts offspring sex in a contemporary
British population. Proceedings of the Royal Society, London B 2004, 271:S474S476.
Almond D, Edlund L: Trivers-Willard at birth and one year: evidence from
US natality data 1983-2001. Proceedings of the Royal Society, London B 2007,
274:5.
Braza F: Human prenatal investment affected by maternal age and parity.
Human Ecology 2004, 32:12.
Marchand A: Mental health in Canada: Are there any risky occupations and
industries? Int J Law & Psych 2007, 30:272.
16
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
37.
408
Table 1: Job category, assigned stress levels (according to Marchand, 2007), percent
409
male births, and sample sizes (N). 1-lowest stress level, 10-highest stress level. The data
410
set includes births and other information collected by Addenbrooke’s hospital in
411
Cambridge, England, between 2000 and 2005.
38.
39.
40.
41.
42.
43.
44.
45.
UK_Stats: UK stats. In Book UK stats (Editor ed.^eds.). City:
www.statistics.gov.uk; 2007.
R_Developent_Core_Team: R: A language and environment for statistical
computing. In Book R: A language and environment for statistical computing
(Editor ed.^eds.), 2.6.1 edition. City: R Foundation for statsitical computing,
Vienna, Austria; 2007.
JMP: SAS, Institute Inc. In Book SAS, Institute Inc. (Editor ed.^eds.), 8.02
edition. City: SAS Institute Inc., SAS Campus Drive, Cary, NC; 2009.
Obel C, Henriksen TB, Secher NJ, Eskenazi B, Hedegaard M: Psychological
distress during early gestation and offspring sex ratio. Human reproduction
2007, 22:3.
Dhabhar FS, McEwen BS: Acute stress enhances while chronic stress
suppresses cell-mediated immunity in vivo: A potential role for leukocyte
trafficking. Brain Behavior And Immunity 1997, 11:286-306.
Navara: Humans at tropical latitudes produce more females. Biology letters
2009.
Lyster W: Altered sex ratio after the London smog of 1952 and the Brisbane
flood of 1965. J Obstet & Gynaec Brit Comm 1974, 81:5.
Catalano R, Bruckner T, Smith KR: Ambient temperature predicts sex ratios
and male longevity. PNAS 2008, 105:5.
Allan BB, Brant R, Seidel JE, Jarrell JF: Declining sex ratios in Canada.
Canadian Medical Association Journal 1997, 156:37-41.
412
413
Table 2: Likelihood Ratio Test results on the effect of mother’s job stress (momstress),
414
partner’s job stress (partstress), partner’s income (partinc), the interaction of the mother’s
415
job stress and partner’s income (momstress*partinc) and the mother’ age (momage) on
416
the offspring sex ratio in a birth data set collected between 2000 and 2005 by
417
Addenbrooke’s hospital in Cambridge, UK. Reported are likelihood ratio estimates,
418
standard errors (SE) and the lower and upper confidence intervals (5% CI, 95% CI).
17
419
420
Figure 1: A response plane illustrating the effect of the interaction between mother’s
421
perceived job stress (x-axis) and partner income (z-axis) on the probability of producing
422
male offspring (y-axis) for parents from Cambridge, UK from 2000 to 2005. Partner
423
income is in £. The blue line illustrates the point (29,991 £) where the effect of partner
424
income cancels the effect of mother stress, so that the probability of male offspring is
425
equal across all levels of mother stress (probability = 0.509). The area between the red
426
and blue lines is where higher levels of maternal stress decrease the probability of having
427
a male. The area between the blue and green lines is where increasing partner income
428
increases the probability of having a male. The individual black dots represent individual
429
data points.
430
18
431
Job category
Stress level
% male births
N
Arts, cultural, recreational, sports
1
52.10
119
Social, legal, government, educational, religious
services
2
53.83
1257
Engineering, natural sciences, architectural, IT
3
49.02
714
Business, finance, administration
4
50.33
2492
Sales, services
5
51.47
3674
Management
6
51.61
2939
Farming, fishing and natural resources
7
44.90
49
Health, Stay-at-home, armed forces
8
52.34
4471
Trades, transport, construction
9
40.38
265
Processing, manufacturing, utilities
10
51.53
392
432
433
19
434
20