American Journal of Epidemiology Copyright © 1992 by The Johns Hopkins University School of Hygiene and Public Health All rights reserved
Vol 136, No. 3 Printed in USA.
Infant Resuscitation Is Associated with an Increased Risk of Left-handedness
Christianna S. Williams,1 Kimberly A. Buss,1 and Brenda Eskenazi12
The etiology of left lateral preference is not well understood, but some studies have suggested that it can be caused by complications at birth. The authors used data from the Child Health and Development Study, a large prospective study of pregnancy and child development conducted 1959-1966 in the San Francisco Bay Area of California, to examine the association between specific birth stressors and hand and foot preference. The study population consisted of 6,968 5-year-olds with no severe congenital abnormalities, and the authors controlled for potential demographic confounders and familial left-handedness. Infants who required resuscitation after delivery or who were twins or triplets were about twice as likely to demonstrate left hand preference at age 5 years (odds ratio (OR) = 1.8, 95% confidence interval (Cl) 1.3-2.5, and OR = 2.2, 95% Cl 1.2-4.0, respectively). Left-footedness was also significantly associated with the same stressors. No other individual stresses were significantly associated with leftlateral preference, and a composite measure indicated only a weak association. Although males, blacks, and those with left-handed siblings are more likely to show left lateral preference, these variables do not confound the association between birth stress and left lateral preference. These results indicate that specific types of birth stress are strongly associated with left hand and foot preference; however, much of the left laterality in non-clinical populations remains unexplained. Am J Epidemiol 1992;136: 277-86. birth order; labor complications; laterality; maternal age; resuscitation
autism, and mental retardation (3). Perhaps the same pathology underlies both sinistrality and these neurologic problems. If so, investigating the etiology of lateralization could lead to a better understanding of the neurologic outcomes as well. One proposal for the cause of left-handedness was put forth by Bakan (4) in the early 1970s. Bakan et al. (5) and others (6) postulated that the left hemisphere is more susceptible than the right to early environmental insult, and that this susceptibility is also greater in males. Bakan (4) suggested that oxygen deficiency associated with birth complications would preferentially affect the left hemisphere, thereby interfering with its usual contralateral control of the right side; those who were naturally right-sided would thus show left or mixed preference. Bakan's initial findings (7) of an association between
The etiology of left-handedness has long intrigued people, and this interest has been heightened by the recent controversial suggestion that left-handers die at younger ages than their dextral counterparts (1, 2). It has been observed that left-handedness is more common among those with certain neurologic disoiueia, such as speech impairments, Received for publication July 19, 1991, and in final form January 23, 1992. 1 Epidemiology Program, School of Public Health, University of California, Berkeley, CA. 2 Maternal and Child Health Program, School of Public Hearth, University of California, Berkeley, CA. Reprint requests to Dr. Brenda Eskenaa, 312 Warren Hall, School of Public Health, University of California, Berkeley, CA 94720. The authors are grateful to Robert Sholtz, Paul English, and Roberta Chnshanson. They would also acknowledge that the opportunity to carry out this work was provided by the course Biomedfcal and Environmental Health Sciences 265.
277
278
Williams et al.
left-handedness and "high risk" birth order (first or later than third) retrospectively reported by college students stimulated further examination of the association between birth complications and lateral dominance. A refinement of Bakan's hypothesis was put forth by Satz (8), who claimed that pathologic handedness is not necessarily lefthandedness; rather, it is handedness that is the opposite of that dictated by heredity. Therefore, the vast majority of the population would be right-handed, and a pathologic shift would be to the left. For the minority who would have naturally been left-handed, a right-handed outcome would in fact be pathologic. For this reason, in a study of the association between birthrelated factors and hand preference, it is important to consider familial handedness. Although there have been mixed results, several studies have found positive associations between various measures of birth stress and left lateral preference (reviewed by Searleman et al. (9)). However, as Searleman et al. detail in their review, these studies have had several limitations. Many have considered only imprecise correlates of birth stress, such as birth order or maternal age (7, 10-13), while others are incomplete because they are based on questionnaire data obtained from the adult participant or their mother in regard to the birth long after the event took place (13-15). Further, small sample sizes may have hindered the detection of an association because both lefthandedness and birth stress are relatively rare (9). Most studies have not included foot or eye preference, which may be more sensitive indicators of lateral dominance, because they are less strongly influenced by social factors than hand preference (9). Ehrlichman et al. (16) corrected many of these methodological problems in a prospective study on 5,899 births drawn from the Collaborative Perinatal Project. They found little evidence of an association between birth stress and left hand or left eye preference at age 7 years. Similarly, McManus (17) analyzed a large data base containing birth and handedness records and concluded that left-handedness and birth stress were
not associated. However, no previous study has adequately controlled for a family history of left lateral preference, an important predictor of left-handedness. Several aspects of the present investigation permit the authors to overcome most of the limitations outlined by Searleman et al. (9) in their critique of the current research into the relation between laterality and birth complications. The data are from follow-up of a large cohort of women who delivered between 1959 and 1966. Detailed information on individual birth complications and performance measures of three aspects of laterality (hand, foot, and eye) at age 5 years were collected prospectively. In addition, information was gathered on handedness in older siblings. MATERIALS AND METHODS Study population
Data for this analysis were obtained from the Child Health and Development Study, a longitudinal study to examine the relation between medical and demographic characteristics of pregnant women, their pregnancy outcomes, and growth and development of their offspring. Information was collected on nearly all consecutive pregnancies at Kaiser Permanente of Oakland, California, between June 1959 and September 1966, providing data on 20,754 pregnancies and 19,044 live births. Two groups of infants were part of follow-up studies on physical and cognitive development of children at age 5 years. One group included 4,931 (or 65 percent) of the infants born between June 1960 and January 1963, and the second group included 3,412 (or 71 percent) of the infants born between April 1964 and April 1966. The distribution of demographic characteristics, handedness, and birth stress is very similar for the two groups of 5-yearolds, so the groups are combined in all analyses. Children were eligible to participate in this follow-up if they were still living in the Greater San Francisco Bay Area and if their mother had agreed to participate in the initial part of the study. For the purpose of the present analysis, children with potentially
Infant Resuscitation and Hand Preference
life-threatening or disabling congenital anomalies (encompassing structural defects, functional abnormalities, metabolic disorders, and chromosomal aberrations) were excluded (n = 274). Only one sibling was chosen at random from each family (thus eliminating 1,101), leaving 6,968 participants. Measurement of variables
Information on socioeconomic status and reproductive history was usually collected by interview with the gravida early in her pregnancy. For a small number of women (11 percent), information was collected by interview after the pregnancy or from secondary sources. Information on complications of delivery was abstracted from hospital records and delivery room dictations. These complications were dichotomized to reflect the presence or absence of a significant biologic stress to the infant. Birth stressors included low birth weight (1 minute or >2 minutes, was not associated with left-handedness. Although those taking a long time to cry were more likely to be resuscitated, this correlation was not strong enough to produce an observable association between time to crying and laterality. Much of the controversy surrounding the association between birth stress and lateral preference has centered around maternal age and birth order, which have been assumed to be indicators of birth stress (e.g., 7, 1013). Though these measures are associated with birth stress and with resuscitation in our study population, they do not predict left lateral preference. Possibly the association between birth order or maternal age and birth stress has been stronger in other study populations. Or, alternatively, the association between birth order or maternal
Infant Resuscitation and Hand Preference
age and left lateral preference that has been found in other studies may not be mediated by birth stress. Stratified analyses showed that the association between resuscitation and left preference was slightly stronger among whites than blacks, although blacks are more likely to be left-handed. This interaction suggests that resuscitation might have been administered for less severe hypoxia for blacks than for other races. Males are also more likely to show left lateral preference, but logistic regression provided no evidence that they were differentially affected by resuscitation or twinning. Since there was no observable interaction between birth stress and familial lefthandedness in our study population, our data provide no support for Satz' theory (8) that the association between stress and lefthandedness will be strongest in those with no family history. However, it must be noted that our measure of familial left-handedness is incomplete, since there is no information on parental handedness. Although we found a strong association between infant resuscitation and left lateral preference, it remains plausible that both of these factors are themselves outcomes of an earlier developmental event. The situation may be analogous to what has recently been demonstrated for cerebral palsy and seizure disorders, conditions once thought to be caused by birth stress. It was found that for many instances of cerebral palsy (25) or seizure disorders (26) there was evidence that the fetus was neurologically damaged prior to the onset of labor. In summary, resuscitation, multiple birth, race, sex, and family history all independently help to predict lateral preference in this population of 5-year-olds without severe congenital anomalies. However, a relatively small proportion of left-handedness can be explained by birth stress: the low prevalence of resuscitation and multiple birth, coupled with the moderate relative risks observed yields an estimate of the attributable risk of only 3.8 percent for resuscitation and 1.4 percent for twinning. Thus, although those
285
who are twins or who require resuscitation at birth are more likely to be left-handed, the etiology of lateral preference in most "lefties" remains elusive.
REFERENCES 1. Halpern DF, Coren S. Handedness and lifespan. (Letter). N Engl J Med 1991 ;324:998. 2. Left-handedness and life expectancy. (Letters from Harrell HL Jr, Morens DM, Katz AR; Rothman KJ; Strang J; Marks JS, Williamson JF; Wolf PA, D'Agostino RB, Cobb J; Halpern DF, Coren S.) N Engl J Med 1991 ;325:1041-3. 3. Kinsbourne M. Sinistrality, brain organization, and cognitive deficits. In: Molfese DL, Segalowitz SJ, eds. Brain lateralization in children: developmental implications. New York: Guilford Press, 1988: 259-79. 4. Bakan P. Are left-handers brain damaged? New Scientist 1975;67:20O-2. 5. Bakan P, Dibb G, Reed P. Handedness and birth stress. Neuropsychologia 1973; 11:363-6. 6. Geschwind N, Galaburda AM. Cerebral lateralization: biological mechanisms, associations and pathology. Cambridge, MA: MIT Press, 1987. 7. Bakan P. Handedness and birth order. (Letter.) Nature 1971;229:195. 8. Satz P. Left-handedness and early brain insult: an explanation. Neuropsychologia 1973;11:115-17. 9. Searleman A, Porac C, Coren S. Relationship between birth order, birth stress, and lateral preferences: a critical review. Psychol Bull 1989;105: 397^08. 10. Hubbard JI. Handedness not a function of birth order. (Letter.) Nature 1971;232:276-7. 11. Leviton A, Kilty T. Birth order and left-handedness. (Utter.) Arch Neurol 1976,33:664. 12. Coren S. Left-handedness in offspring as a function of maternal age at parturition. (Letter.) N Engl J Med 1990:322:1673. 13. Schwartz M. Left-handedness and high risk pregnancy. Neuropsychologia 1977;15:341-4. 14. Hicks RA, Dusek C, Larsen F, et al. Birth complications and the distribution of handedness. Cortex 1980; 16:483-6. 15. Coren S, Porac C. Birth factors and laterality: effects of birth order, parental age, and birth stress on four indices of lateral preference. Behav Genet 1980;10:123-38. 16. Ehrlichman H, Zoccolotti P, Owen D. Perinatal factors in hand and eye preference: data from The Collaborative Perinatal Project. Int J Neurosci 1982;17:17-22. 17. McManus IC. Handedness and birth stress. Psychol Med 1981:11:485-96. 18. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: John Wiley and Sons, 1981. 19. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research: principles and quantitative
286
20. 21. 22. 23.
Williams et al.
methods. Belmont, CA: Lifetime Learning Publications, 1982. Coren S, Searleman A, Porac C. The effects of specific birth stressors on four indexes of lateral preference. Can J Psychol 1982;36:478-87. Hecaen H, de Ajuriaguerra J. Left-handedness; manual superiority and cerebral dominance. New York: Grunc and Stratton, 1964. Tan LE, Nettleton NC. Left-handedness, birth order and birth stress. Cortex 1980; 16:363-73. Barnes F. Temperament, adaptability and left-
handers. New Scientist 1975^4:202-3. 24. Smart JL, Jeffery C, Richards B. A retrospective study of the relationship between birth history and handedness at six yean. Early Hum Dev 1980;4: 79-88. 25. Torfs CP, van den Berg BJ, Oechsli FW, et al. Prenatal and perinatal factors in the etiology of cerebral palsy. J Pediatr 1990; 1166:15-19. 26. Nelson KB, Ellenberg JH. Antecedents of seizure disorders in early childhood. Am J Dis Child 1986; 140:1053-61.
Vol 136, No. 3 Printed in USA.
Infant Resuscitation Is Associated with an Increased Risk of Left-handedness
Christianna S. Williams,1 Kimberly A. Buss,1 and Brenda Eskenazi12
The etiology of left lateral preference is not well understood, but some studies have suggested that it can be caused by complications at birth. The authors used data from the Child Health and Development Study, a large prospective study of pregnancy and child development conducted 1959-1966 in the San Francisco Bay Area of California, to examine the association between specific birth stressors and hand and foot preference. The study population consisted of 6,968 5-year-olds with no severe congenital abnormalities, and the authors controlled for potential demographic confounders and familial left-handedness. Infants who required resuscitation after delivery or who were twins or triplets were about twice as likely to demonstrate left hand preference at age 5 years (odds ratio (OR) = 1.8, 95% confidence interval (Cl) 1.3-2.5, and OR = 2.2, 95% Cl 1.2-4.0, respectively). Left-footedness was also significantly associated with the same stressors. No other individual stresses were significantly associated with leftlateral preference, and a composite measure indicated only a weak association. Although males, blacks, and those with left-handed siblings are more likely to show left lateral preference, these variables do not confound the association between birth stress and left lateral preference. These results indicate that specific types of birth stress are strongly associated with left hand and foot preference; however, much of the left laterality in non-clinical populations remains unexplained. Am J Epidemiol 1992;136: 277-86. birth order; labor complications; laterality; maternal age; resuscitation
autism, and mental retardation (3). Perhaps the same pathology underlies both sinistrality and these neurologic problems. If so, investigating the etiology of lateralization could lead to a better understanding of the neurologic outcomes as well. One proposal for the cause of left-handedness was put forth by Bakan (4) in the early 1970s. Bakan et al. (5) and others (6) postulated that the left hemisphere is more susceptible than the right to early environmental insult, and that this susceptibility is also greater in males. Bakan (4) suggested that oxygen deficiency associated with birth complications would preferentially affect the left hemisphere, thereby interfering with its usual contralateral control of the right side; those who were naturally right-sided would thus show left or mixed preference. Bakan's initial findings (7) of an association between
The etiology of left-handedness has long intrigued people, and this interest has been heightened by the recent controversial suggestion that left-handers die at younger ages than their dextral counterparts (1, 2). It has been observed that left-handedness is more common among those with certain neurologic disoiueia, such as speech impairments, Received for publication July 19, 1991, and in final form January 23, 1992. 1 Epidemiology Program, School of Public Health, University of California, Berkeley, CA. 2 Maternal and Child Health Program, School of Public Hearth, University of California, Berkeley, CA. Reprint requests to Dr. Brenda Eskenaa, 312 Warren Hall, School of Public Health, University of California, Berkeley, CA 94720. The authors are grateful to Robert Sholtz, Paul English, and Roberta Chnshanson. They would also acknowledge that the opportunity to carry out this work was provided by the course Biomedfcal and Environmental Health Sciences 265.
277
278
Williams et al.
left-handedness and "high risk" birth order (first or later than third) retrospectively reported by college students stimulated further examination of the association between birth complications and lateral dominance. A refinement of Bakan's hypothesis was put forth by Satz (8), who claimed that pathologic handedness is not necessarily lefthandedness; rather, it is handedness that is the opposite of that dictated by heredity. Therefore, the vast majority of the population would be right-handed, and a pathologic shift would be to the left. For the minority who would have naturally been left-handed, a right-handed outcome would in fact be pathologic. For this reason, in a study of the association between birthrelated factors and hand preference, it is important to consider familial handedness. Although there have been mixed results, several studies have found positive associations between various measures of birth stress and left lateral preference (reviewed by Searleman et al. (9)). However, as Searleman et al. detail in their review, these studies have had several limitations. Many have considered only imprecise correlates of birth stress, such as birth order or maternal age (7, 10-13), while others are incomplete because they are based on questionnaire data obtained from the adult participant or their mother in regard to the birth long after the event took place (13-15). Further, small sample sizes may have hindered the detection of an association because both lefthandedness and birth stress are relatively rare (9). Most studies have not included foot or eye preference, which may be more sensitive indicators of lateral dominance, because they are less strongly influenced by social factors than hand preference (9). Ehrlichman et al. (16) corrected many of these methodological problems in a prospective study on 5,899 births drawn from the Collaborative Perinatal Project. They found little evidence of an association between birth stress and left hand or left eye preference at age 7 years. Similarly, McManus (17) analyzed a large data base containing birth and handedness records and concluded that left-handedness and birth stress were
not associated. However, no previous study has adequately controlled for a family history of left lateral preference, an important predictor of left-handedness. Several aspects of the present investigation permit the authors to overcome most of the limitations outlined by Searleman et al. (9) in their critique of the current research into the relation between laterality and birth complications. The data are from follow-up of a large cohort of women who delivered between 1959 and 1966. Detailed information on individual birth complications and performance measures of three aspects of laterality (hand, foot, and eye) at age 5 years were collected prospectively. In addition, information was gathered on handedness in older siblings. MATERIALS AND METHODS Study population
Data for this analysis were obtained from the Child Health and Development Study, a longitudinal study to examine the relation between medical and demographic characteristics of pregnant women, their pregnancy outcomes, and growth and development of their offspring. Information was collected on nearly all consecutive pregnancies at Kaiser Permanente of Oakland, California, between June 1959 and September 1966, providing data on 20,754 pregnancies and 19,044 live births. Two groups of infants were part of follow-up studies on physical and cognitive development of children at age 5 years. One group included 4,931 (or 65 percent) of the infants born between June 1960 and January 1963, and the second group included 3,412 (or 71 percent) of the infants born between April 1964 and April 1966. The distribution of demographic characteristics, handedness, and birth stress is very similar for the two groups of 5-yearolds, so the groups are combined in all analyses. Children were eligible to participate in this follow-up if they were still living in the Greater San Francisco Bay Area and if their mother had agreed to participate in the initial part of the study. For the purpose of the present analysis, children with potentially
Infant Resuscitation and Hand Preference
life-threatening or disabling congenital anomalies (encompassing structural defects, functional abnormalities, metabolic disorders, and chromosomal aberrations) were excluded (n = 274). Only one sibling was chosen at random from each family (thus eliminating 1,101), leaving 6,968 participants. Measurement of variables
Information on socioeconomic status and reproductive history was usually collected by interview with the gravida early in her pregnancy. For a small number of women (11 percent), information was collected by interview after the pregnancy or from secondary sources. Information on complications of delivery was abstracted from hospital records and delivery room dictations. These complications were dichotomized to reflect the presence or absence of a significant biologic stress to the infant. Birth stressors included low birth weight (1 minute or >2 minutes, was not associated with left-handedness. Although those taking a long time to cry were more likely to be resuscitated, this correlation was not strong enough to produce an observable association between time to crying and laterality. Much of the controversy surrounding the association between birth stress and lateral preference has centered around maternal age and birth order, which have been assumed to be indicators of birth stress (e.g., 7, 1013). Though these measures are associated with birth stress and with resuscitation in our study population, they do not predict left lateral preference. Possibly the association between birth order or maternal age and birth stress has been stronger in other study populations. Or, alternatively, the association between birth order or maternal
Infant Resuscitation and Hand Preference
age and left lateral preference that has been found in other studies may not be mediated by birth stress. Stratified analyses showed that the association between resuscitation and left preference was slightly stronger among whites than blacks, although blacks are more likely to be left-handed. This interaction suggests that resuscitation might have been administered for less severe hypoxia for blacks than for other races. Males are also more likely to show left lateral preference, but logistic regression provided no evidence that they were differentially affected by resuscitation or twinning. Since there was no observable interaction between birth stress and familial lefthandedness in our study population, our data provide no support for Satz' theory (8) that the association between stress and lefthandedness will be strongest in those with no family history. However, it must be noted that our measure of familial left-handedness is incomplete, since there is no information on parental handedness. Although we found a strong association between infant resuscitation and left lateral preference, it remains plausible that both of these factors are themselves outcomes of an earlier developmental event. The situation may be analogous to what has recently been demonstrated for cerebral palsy and seizure disorders, conditions once thought to be caused by birth stress. It was found that for many instances of cerebral palsy (25) or seizure disorders (26) there was evidence that the fetus was neurologically damaged prior to the onset of labor. In summary, resuscitation, multiple birth, race, sex, and family history all independently help to predict lateral preference in this population of 5-year-olds without severe congenital anomalies. However, a relatively small proportion of left-handedness can be explained by birth stress: the low prevalence of resuscitation and multiple birth, coupled with the moderate relative risks observed yields an estimate of the attributable risk of only 3.8 percent for resuscitation and 1.4 percent for twinning. Thus, although those
285
who are twins or who require resuscitation at birth are more likely to be left-handed, the etiology of lateral preference in most "lefties" remains elusive.
REFERENCES 1. Halpern DF, Coren S. Handedness and lifespan. (Letter). N Engl J Med 1991 ;324:998. 2. Left-handedness and life expectancy. (Letters from Harrell HL Jr, Morens DM, Katz AR; Rothman KJ; Strang J; Marks JS, Williamson JF; Wolf PA, D'Agostino RB, Cobb J; Halpern DF, Coren S.) N Engl J Med 1991 ;325:1041-3. 3. Kinsbourne M. Sinistrality, brain organization, and cognitive deficits. In: Molfese DL, Segalowitz SJ, eds. Brain lateralization in children: developmental implications. New York: Guilford Press, 1988: 259-79. 4. Bakan P. Are left-handers brain damaged? New Scientist 1975;67:20O-2. 5. Bakan P, Dibb G, Reed P. Handedness and birth stress. Neuropsychologia 1973; 11:363-6. 6. Geschwind N, Galaburda AM. Cerebral lateralization: biological mechanisms, associations and pathology. Cambridge, MA: MIT Press, 1987. 7. Bakan P. Handedness and birth order. (Letter.) Nature 1971;229:195. 8. Satz P. Left-handedness and early brain insult: an explanation. Neuropsychologia 1973;11:115-17. 9. Searleman A, Porac C, Coren S. Relationship between birth order, birth stress, and lateral preferences: a critical review. Psychol Bull 1989;105: 397^08. 10. Hubbard JI. Handedness not a function of birth order. (Letter.) Nature 1971;232:276-7. 11. Leviton A, Kilty T. Birth order and left-handedness. (Utter.) Arch Neurol 1976,33:664. 12. Coren S. Left-handedness in offspring as a function of maternal age at parturition. (Letter.) N Engl J Med 1990:322:1673. 13. Schwartz M. Left-handedness and high risk pregnancy. Neuropsychologia 1977;15:341-4. 14. Hicks RA, Dusek C, Larsen F, et al. Birth complications and the distribution of handedness. Cortex 1980; 16:483-6. 15. Coren S, Porac C. Birth factors and laterality: effects of birth order, parental age, and birth stress on four indices of lateral preference. Behav Genet 1980;10:123-38. 16. Ehrlichman H, Zoccolotti P, Owen D. Perinatal factors in hand and eye preference: data from The Collaborative Perinatal Project. Int J Neurosci 1982;17:17-22. 17. McManus IC. Handedness and birth stress. Psychol Med 1981:11:485-96. 18. Fleiss JL. Statistical methods for rates and proportions. 2nd ed. New York: John Wiley and Sons, 1981. 19. Kleinbaum DG, Kupper LL, Morgenstern H. Epidemiologic research: principles and quantitative
286
20. 21. 22. 23.
Williams et al.
methods. Belmont, CA: Lifetime Learning Publications, 1982. Coren S, Searleman A, Porac C. The effects of specific birth stressors on four indexes of lateral preference. Can J Psychol 1982;36:478-87. Hecaen H, de Ajuriaguerra J. Left-handedness; manual superiority and cerebral dominance. New York: Grunc and Stratton, 1964. Tan LE, Nettleton NC. Left-handedness, birth order and birth stress. Cortex 1980; 16:363-73. Barnes F. Temperament, adaptability and left-
handers. New Scientist 1975^4:202-3. 24. Smart JL, Jeffery C, Richards B. A retrospective study of the relationship between birth history and handedness at six yean. Early Hum Dev 1980;4: 79-88. 25. Torfs CP, van den Berg BJ, Oechsli FW, et al. Prenatal and perinatal factors in the etiology of cerebral palsy. J Pediatr 1990; 1166:15-19. 26. Nelson KB, Ellenberg JH. Antecedents of seizure disorders in early childhood. Am J Dis Child 1986; 140:1053-61.
