This document examines the question of whether the dramatic drop in infant and fetal mortality rates that occurred in the United States during the twentieth century could be a cause for some true increase in the prevalence of autism. Autism is associated with difficult births, as well as other factors, that could mean that people with autism disproportionately benefited from changes that greatly lowered the infant and fetal mortality rates.
The question of whether the prevalence of autism has risen is sometimes falsely given in an either/or form – either there is an epidemic or we have only gotten better at classifying and counting. It appears to be indisputable that there have been important cultural changes that have affected our view of autism, and these changes have affected what we have measured as the prevalence rate. Roy Grinker (2007), in Unstrange Minds, ably described cultural changes that contributed to these trends.
What might also have contributed to these changes in reported prevalence, and created a true rise in autism, is the very dramatic drop in infant and fetal mortality rates which occurred during the twentieth century. From Vital Statistic Rates in the United States 1900-1940 (Linder & Grove, 1947): “The collection of birth statistics was initiated in 1915, and since 1933 data for both births and deaths have been available for the United States as a whole.” (p. 1) In 1918, the infant mortality rate (from birth to one year) was estimated at 100.9 per 1000 live births. (p. 607). In 1933, the infant mortality rate was measured for the United States at 58.1 per 1000 live births (p. 607), and the stillbirth ratio (after about 20 weeks) was 37.0 per 1000 live births (p. 655).
Around the same time period, in 1938, Leo Kanner started noticing children with autistic characteristics. He wrote about them in 1943, and Han Asperger wrote about similar children in 1944. It is unclear what the rate of autism was at that time. For the 1950’s “autism was said to occur in only 3 in 10,000 live births” (Grinker, 2007, p. 169).
The rates grew as the concept of autism changed. Of eight studies conducted between 2000 and 2003, six found rates of between 52 per 10, 000 and 67.5 per 10,000 (that is, as high as 1 in 155). (Grinker, 2007, p. 157)
By the year 2000, the infant mortality rate had dropped to 6.9 per 1000 (Kochanek & Martin, 2005, p. 108) and the fetal mortality rate (greater than 20 weeks) had dropped to 6.61 per 1000 (MacDorman, Hoyert, Martin, Munson & Hamilton, 2007, p. 3). This gives us roughly 82 per 1030 additional infants who were born in 2000 in the United States who wouldn’t have survived if they had been born in 1933. If we use an estimate of 6 per 1000 of people on the autistic spectrum, one might expect that .5 per 1030 of those additional children would be autistic, leaving the prevalence rate unchanged.
On the other hand, there is a large amount of evidence that being autistic is associated with difficult births. Glasson, Petterson, de Klerk, Chaney, & Hallmayer (2004) found that during the birth of children later diagnosed with autism, the mother was more likely to have an induced labor (1.43 times more likely), an emergency cesarean section (1.57 times), or for the fetus to be in distress (1.59 times) (p. 621). Larsson et al. (2005) found these births were more likely to be associated with breech presentation (1.63 times more likely), the babies more likely to have a low Apgar score at 5 minutes (1.89 times), and more likely to have a gestational age at birth < 35 weeks (2.45 times) (p. 916).
One might assume from this that autism can be caused by a difficult birth, but some researchers have suggested the opposite, that autism can cause a difficult birth. Researchers Geschwind and Galaburda (1987) wrote “the same influences that alter brain development also lead to disturbance of the birth process” (p. 176). If that is true then people with autism might disproportionately have benefited and survived because of life-saving changes made throughout the twentieth century.
Beyond difficult births, though not as firmly established, autism has also been associated with not being breastfed (Schulz, et al., 2006) and using antibiotics (Fallon, 2005). It has been associated with being more likely to have injuries that need medical attention (Lee, Harrington, Chang, & Connors, 2007) and having accidents such as suffocation and drowning (Shavelle, Strauss, & Pickett, 2001). People with autism may have benefited more than most people from improvements in infant formula, medical care, and safety equipment.
If we assume that the drop in infant/fetal mortality could account for a big change in the autism rate, estimating roughly, then instead of the extra autistic children being .5 per 1030 (from the 82 per 1030 total extra children), we might assume it would be closer to 5 per 1030. This would mean that autistic children would need to have benefited from these changes 10 times more than typical children. This isn’t an impossible number, given that many smaller probabilities of mortality would multiply to a larger number, but even so, it probably is an overestimate of what would be needed for a number of reasons.
• The fetal mortality rate was probably underreported in 1933. “The real decline in fetal mortality [from 1942 to 2003] was probably larger, as reporting of fetal deaths has improved over time.” (MacDorman, et al., 2007, p. 2)
• Changes would also have benefited fetuses under twenty weeks old, and autistic children over one year old, who also could have disproportionately benefited. “In 1900, >3 in 100 children died between their first and 20th birthday; today, <2 in 1000 die.” (Guyer, Freedman, Strobino, & Sondik, 2000, p. 1307)
• Since autism is a strongly genetic disorder, autistic people can have children who have autism. That means that extra surviving people with autism could sometimes further increase the autism rate by having children who wouldn’t have otherwise been born. “The genetic basis of autism spectrum disorders is supported by clustering in families, higher concordance in monozygotic twins than in dizygotic twins, and evidence for the broader phenotype, or variable expressivity, in other family members.” (Glasson, et al., 2004, p. 618)
• Maternal mortality also dropped over the same time period. In 1933, it was 6.2 per 1000 (Linder & Grove, 1947, p. 621). “During 1982-1996, the annual maternal mortality ratio fluctuated between approximately 7 and 8 maternal deaths per 100,000 live births.” (CDC, 1998) Any maternal deaths that were prevented which would have occurred because of difficult births related to autism could then go on to have more children, who would be autistic at a higher rate than typical mothers. From a study by Lauritsen, Pedersen & Mortensen (2005): “Our study provides further evidence of the involvement of genetic factors in the etiology of autism based on the significant findings of very high relative risks of autism in the siblings of children with autism or the other PDDs studied compared to the other risk factors investigated.” (p. 966-967)
• Those with more severe autism may have benefited most from the drop in infant and fetal mortality rates, and that could make autism more visible with less rise in the total prevalence of people on the autistic spectrum. “Cases with a diagnosis of autism had more [obstetric] complications than those with pervasive developmental disorder not otherwise specified or Asperger syndrome.” (Glasson, et al., 2004, p. 618)
• Some trends, like improved sanitation, helped raise the life expectancy generally for adults in the United States. In 1900 it was estimated to be 47.3 years, and in 2000 it was 77. (Arias, 2006, p. 34-35) There have been a number of studies that suggest that autism may be associated with older parents, one or both. (Glasson, et al., 2004; Reichenberg, Bresnahan, Rabinowitz, Lubin & Davidson, 2006; Croen, Najjar, Fireman, & Grether, 2007) One explanation of this finding is to suggest that people who have traits related to autism may tend to be older when they have children. “Attributes of the parents that lead to older age at childbearing and that are related to autism risk might also explain the association we observed with advancing parental age.” (Croen, et al., 2007, p. 339) If true, then the rise in life expectancy may have been disproportionately beneficial to people who have genes related to autism, with respect to their having children.
• Falling infant mortality has contributed to the trend towards smaller families that has increased the proportion of firstborns, and there is some evidence that being firstborn is a risk factor for autism. “Cases [of autism] were more likely to be firstborn compared with control subjects. The finding is consistent with other reports of first births carrying a risk for autism.” (Glasson, et al., 2004) Total Fertility Rates fluctuated during the twentieth century, but there was a steady decline from 1959 with 3.7 children per woman to 1976 with 1.8 children per woman (National Center for Health Statistics, 2000, p. 1). “There has been remarkably little change in the fertility rate during the last quarter century.” (Guyer, et al., 2000, p. 1310)
• This analysis assumes a very large increase in prevalence, but even if the general concept is correct, the reality might have been a very small increase in true prevalence.
I believe that this document shows that there is a theoretical possibility that the dramatic drop in the infant and fetal mortality rates in the United States may have caused some true increase in autism, and that this idea could be part of a broader falling mortality hypothesis which would claim that all types of dropping mortality could contribute to rising autism rates. This argument could also be extended to other neurological conditions. Though this is only one possible factor involved in the measured prevalence of autism, it may be an important one. Based on these ideas, one prediction would be that in places in the world where the infant and fetal mortality rates remain high, true autism prevalence would remain lower.
References
Arias, E. (2006). United States life tables, 2003. National Vital Statistics Reports, 54(14), 1-40.
CDC (1998). Maternal mortality – United States, 1982-1996. Morbidity and Mortality Weekly Report (MMWR), 47(34), 705-7. http://www.cdc.gov/mmwr/preview/mmwrhtml/00054602.htm
Croen, L., Najjar, D. V., Fireman, B., & Grether, J. K. (2007). Maternal and paternal age and risk of autism spectrum disorders. Archives of Pediatrics & Adolescent Medicine, 161, 334-340.
Fallon, J. (2005). Could one of the most widely prescribed antibiotics amoxicillin/clavulanate “augmentin” be a risk factor for autism? Medical Hypotheses, 64, 312-315.
Geschwind, N., & Galaburda, A. (1987). Cerebral Lateralization: Biological Mechanisms, Associations, and Pathology. Cambridge, Massachusetts: The MIT Press.
Glasson, E. J., Petterson, B., de Klerk, N., Chaney, G., & Hallmayer, J. F. (2004). Perinatal factors and the development of autism. Archives of General Psychiatry, 61, 618-627.
Grinker, R. R. (2007). Unstrange Minds: Remapping the World of Autism. New York: Basic Books.
Guyer, B., Freedman, M. A., Strobino, D. M., & Sondik, E.J (2000). Annual summary of vital statistics: trends in the health of Americans during the 20th Century. Pediatrics, 106(6), 1307-1317.
Kochanek, K. D., & Martin, J. A. (2005). Supplemental analyses of recent trends in infant mortality. International Journal of Health Services, 35(1), 101-115.
Lauritsen, M. B., Pedersen, C. B. & Mortensen, P. B. (2005). Effects of familial risk factors and place of birth on the risk of autism: a nationwide register-based study. Journal of Child Psychology and Psychiatry, 46(9), 963-971.
Larsson, H. J., Eaton, W. W., Madsen, K. M., Vestergaard, M., Olesen, A. V., Agerbo, E., Schendel, D. Thorsen, P., & Mortensen, P. B. (2005). Risk factors for autism: perinatal factors, parental psychiatric history, and socioeconomic status. American Journal of Epidemiology, 161(10), 916-925.
Lee, L.-C., Harrington, R. A., Chang, J. J., & Connors, S. L. (2007). Increased risk of injury in children with developmental disabilities. Research in Developmental Disabilities. In Press.
Linder, F. E. & Grove, R. D. (1947). Vital Statistics Rates in the United States 1900-1940. Washington, D. C.: United States Government Printing Office.
MacDorman, M. F., Hoyert, D. L., Martin, J. A., Munson, M. L., & Hamilton, B. E. (2007). Fetal and perinatal mortality, United States, 2003. National Vital Statistics Reports, 55(6), 1-18.
National Center for Health Statistics (2000). Table 1-7. Total fertility rates and birth rates, by age of mother and race of child: United States, 1940-80. Vital Statistics of the United States, 1997, Volume I, Natality. http://www.cdc.gov/nchs/datawh/statab/unpubd/natality/natab97.htm
Reichenberg, A., Bresnahan, M., Rabinowitz, J., Lubin, G., & Davidson, M. (2006). Advancing paternal age and autism. Archives of General Psychiatry, 63, 1026-1032.
Schultz, S. T., Klonoff-Cohen, H. S., Wingard, D. L., Akshoomoff, N. A. Macera, C. A., Ji, M., & Bacher, C. (2006). Breastfeeding, infant formula supplementation, and Autistic Disorder: the results of a parent survey. International Breastfeeding Journal, 1(16).
Could Autism Increase by Infant/Fetal Mortality Dropping?
