CanJPsychiatry 2015;60(12):587–590
Perspective
Misattributions and Potential Consequences: The Case of Child Mental Health Problems and Fetal Alcohol Spectrum Disorders John D McLennan, MD, MPH, PhD, FRCPC1 1
Associate Professor, Departments of Pediatrics, Psychiatry, and Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Alberta. Correspondence: Teaching, Research and Wellness Building, 3rd Floor, 3280 Hospital Drive NW, Calgary, AB T2N 4Z6; [email protected].
Key Words: fetal alcohol spectrum disorder, mental disorders, bias, comorbidity, child Received January 2015, revised, and accepted June 2015.
N
umerous Canadian agencies have prioritized services for people diagnosed with FAS and its broader construct FASD. This prioritization extends to prevention interventions aimed at reducing or eliminating PAE. The many difficulties identified as associated with FAS, FASD, and PAE is one of the justifications for this prioritization. Mental health symptoms and disorders are among the most commonly highlighted challenges experienced by people labelled with FAS or FASD or exposed to alcohol in utero. However, the extent of the relation between the FAS, FASD, and PAE cluster and mental health symptom and disorder clusters may be inflated secondary to at least 3 factors: diagnostic criteria overlap and etiologic assumptions, referral bias, and failure to control for confounding variables when assessing associations. Lack of awareness of these factors may lead to dissemination of misinformation, which could adversely distort the development and provision of mental health services. Diagnostic criteria for conditions falling under the FASD umbrella have been operationalized in several different guidelines. The 2005 Canadian guideline1 aimed, in part, to harmonize aspects of the 2 leading approaches at the time, that is, those of the Institute of Medicine2 and the Washington 4-digit diagnostic code.3 Guidelines typically include the complete syndrome, FAS, and require positive findings in 4 domains: problematic patterns of alcohol exposure in utero (for example, from maternal binge drinking), growth abnormalities (for example, low birth weight for gestational age), facial dysmorphology (for example, short palpebral fissures), and CNS neurodevelopment abnormalities (for example, microcephaly at birth).2 Guidelines then typically go on to describe various partial syndromes. In the case of the Canadian guideline, the following partial syndromes are included: FAS (without confirmed alcohol exposure), partial FAS, and ARND.1 Difficulties identified within children with high PAE, but who do not have classical dysmorphic manifestations, is used to support the inclusion of partial syndromes.4 However, this broadening likely contributes to problematic overlap with those children with mental health difficulties for whom PAE may be present but for whom it is not etiologic. This is particularly problematic as there is no consensus on a pathognomonic behavioural manifestation of PAE or FASD. Although some propose a unique mental health profile linked to FASD,5 such profiles are based on small clinical samples and do not appear to have been independently replicated using a nonreferred population. Nevertheless, a resulting 10-item screening tool appears to be receiving national promotion in Canada.6 Concerns that weaknesses in the operationalization of partial FASD syndromes may lead to misattribution of PAE as causal for various difficulties (for example, behavioural problems) has been raised in critiques of the 2 dominant diagnostic approaches used in the field,2,3 that is, the US sources for the Canadian guidelines.7 The ARND diagnosis, within the Canadian guidelines, requires “evidence of impairment in three or more of the following CNS domains: hard and soft neurologic signs; brain structure; cognition; communication; academic achievement; memory; executive functioning and abstract reasoning; attention deficit/hyperactivity; adaptive behavior, social skills, and social communication.”1, p S12 However, abnormalities in 3 of the listed domains would also
www.TheCJP.ca
The Canadian Journal of Psychiatry, Vol 60, No 12, December 2015 W 587
Perspective
be commonly found in many children with various mental health disorders. PAE may not be uncommon in children with such problem clusters. However, the fraction for whom PAE is primarily etiologic is unknown, and to assume it is typically the leading etiology is highly problematic.8 Complicating the picture is the inclusion of an etiologic variable, in this case PAE, in the diagnostic criteria for FASD. This is at odds with a key direction taken in psychiatry as reflected in contemporary versions of the DSM, that is, to avoid etiologic assumptions within diagnostic criteria.9 That the DSM-5 has included neurobehavioural disorders associated with prenatal alcohol exposure under the section “Conditions for Further Study”10 is perhaps a new exception. Unfortunately, as currently written, this DSM5 provisional criteria may contribute to further diagnostic overlap with mental disorders not caused by PAE. This is in part due to the low threshold set for diagnostic criteria for this provisional disorder. A leading epidemiologist has noted that the inclusion of proposed etiological variables in diagnostic criteria may undermine the validity of epidemiologic evaluations of such defined disorders.11 Further complicating considerations of diagnostic overlap and (or) potential comorbidity is a direction taken by some in the FASD community to label mental disorder manifestations in children with PAE or FASD as being secondary disabilities.12 Although this may be valid in some situations, it is concerning if there is a general assumption that manifestations of mental disorders in the context of PAE or FASD are typically secondary phenomena. Relying primarily on the temporal relation is inadequate to make this linkage. Various mental disorders have different times of manifestation, regardless of PAE, and always manifest postpartum. It is not clear how a clinician could determine that the emergence of a mental disorder is secondary to PAE or FASD and not a function of genetic or other environmental factors that may underlie timing of mental disorder emergence. Leaving the problem of diagnostic criteria overlap and etiologic assumptions aside, referral bias (also known as Berksonian or selection bias) may also lead to misattribution. Referral bias may occur when a factor of interest in the clinic population (for example, psychiatric symptoms) also influences whether a patient is seen in the clinic.13 This may lead to spurious correlations and generalizations. Apparent associations identified from studies using clinic Abbreviations ADHD
attention-deficit hyperactivity disorder
ARND
alcohol-related neurodevelopmental disorder
CNS
central nervous system
DSM
Diagnostic and Statistical Manual of Mental Disorders
FAS
fetal alcohol syndrome
FASD
fetal alcohol spectrum disorders
PAE
prenatal alcohol exposure
588 W La Revue canadienne de psychiatrie, vol 60, no 12, décembre 2015
Clinical Implications •
Clinicians and researchers should attend to diagnostic criteria overlap and etiologic assumptions, referral bias, and potential influence of confounders when considering associations between different disorders, symptom clusters, and purported risk factors.
•
Some mental health problems in children diagnosed with FASD may be inaccurately attributed to in utero alcohol exposure and (or) FASD.
•
In some cases, children with mental health disorders may be mismanaged secondary to misattributing most or all their difficulties to FASD.
Limitations •
This paper is not based on a systematic literature review.
•
The number of high-quality source documents to evaluate the relation between FASD and child mental health disorders is sparse.
samples that have not been verified in nonreferred samples warrant considerable skepticism. Although it is possible that a child may be sent to an FASD clinic based solely on unique aspects of facial dysmorphology, factors such as disruptive behaviours are much more likely to drive referrals. Consequently, specialized FASD assessment clinics are likely receiving referrals of children and youth with a disproportionally high occurrence of disruptive behaviours. Clinicians then experience high rates of cooccurrence, and the researchers who rely on studies of clinically referred samples may report strong but potentially spurious associations. Unfortunately, most FAS or FASD studies appear to rely on clinic samples. Fortunately, there are a few population-based samples that include an examination of psychiatric morbidity among children assessed for PAE and (or) FAS or FASD and that may help identify whether referral bias underlies some of the high comorbidity noted in clinical populations. ADHD may be the most studied psychiatric comorbidity in FASD. One review found an overall ADHD prevalence of 48% in children with FASD14; however, most source studies in that review used clinic-referred samples. In contrast, a nonreferred, school-sampled study found a combined-type ADHD prevalence of 8.7% among children with FASD based on teacher ratings, compared with 5.3% among children without FASD, a nonsignificant difference.15 The marked differences in ADHD prevalence found in children with FASD in this population-based study, compared with the review of clinic-based studies, may lead to the hypothesis that a substantial portion of ADHD found in children with FASD in referred clinic samples have ADHD as a function of a wider range of possible etiologies.16 Unfortunately, prevalence of other psychiatric disorders have not been reported from population-based studies of FASD; however, referral bias is also likely for other conditions that are apt to precipitate referrals (for example, aggression, conduct disorder, and developmental delays). www.LaRCP.ca
Misattributions and Potential Consequences: The Case of Child Mental Health Problems and Fetal Alcohol Spectrum Disorders
Whether such distortions extend to internalizing problems and disorders is unknown. Confounding variables may also contribute to an overestimation of the strength of relations between the FAS, FASD, and PAE cluster and mental health problems. Children referred to FASD clinics are often found to have had many risk exposures in addition to PAE, such as being victims of physical or sexual abuse and exposure to parental substance use after birth.17 It is unclear how clinicians can attribute mental health presentations to the PAE, compared with one or more of the other risk exposures. In some assessment and advocacy contexts, mental health symptoms and disorders may be framed as primarily or entirely explained by PAE. Such assumptions do not seem to hold up as suggested by findings from numerous studies aimed at examining the extent to which PAE independently explains adverse child outcomes relative to other known risk factors for child mental health problems. For example, in a FASD study within a school-based sample, measures of PAE explained only a modest portion of cognitive and behavioural outcomes and these association were no longer significant once other risk factors (for example, lower maternal education) were included in the multivariate model.18 Another study, based on a sample derived from a maternity clinic, found PAE was related to child behavioural problems at follow-up but explained only 0.6% to 1.7% of outcome variance; in contrast, a measure of maternal psychopathology explained 13.0% to 29.1% of the variance.19 In a population-based study examining the impact of lowto-moderate PAE, the association with ADHD symptoms became statistically insignificant once the multivariate model factored in measures of social adversity.20 Finally, another large population-based study attempted to consider genetic and other environmental contributions beyond PAE by including siblings and cousins with variable PAE.21 They found that the significant differences in attentional and impulsivity problems between children with and without PAE for unrelated families was not found within related family.21 However, this study did find some evidence for a significant relation between PAE and conduct problems after controlling for family membership.21 Although measurement issues and epidemiologic-based findings may, at times, seem disconnected from clinical practice, in this case, failure to appreciate these underlying patterns may have distorting effects on services and policies. For example, at a service level, if a comorbid psychiatric disorder is inaccurately attributed to FASD, this may lead to an inappropriate diversion to a newly designed FASD specialty intervention that lacks evidence of effectiveness for the comorbid psychiatric problem rather than being directed to an evidence-based intervention specific to the identified psychiatric comorbidity. For example, use of evidence-based behavioural interventions or medications for ADHD may be discouraged if the presenting attentional problem is attributed to FASD. A second example may include the setting of eligibility for certain services and supports based on the presence of an FASD diagnosis www.TheCJP.ca
rather than a given identified need that may not be specific to FASD (for example, a social skills deficit). This may reinforce unsubstantiated assumptions made about children with FASD (for example, they all have social skill deficits) and compromise access to a service for a child who may have a need for additional services (for example, social skills training) but who had, say, prenatal polysubstance exposure rather than PAE. A further concern is a potential for harm. It has been proposed that flawed causal models that may be employed within some FASD-focused clinics may violate the ethical principle of nonmaleficence.8 More specifically, assumptions made about causes may inadvertently revive past psychiatric frameworks that emphasized maternal behaviour as etiologic in child psychiatric disorders.8 Finally, the failure to consider an epidemiologic perspective may also distort prevention efforts which may, for example, overemphasize a single risk exposure (for example, elimination of drinking during pregnancy) in efforts to improve child mental health in the population. This may result in relative neglect of other important risk factors contributing to child mental health problems, such as poverty.22 None of the preceding argument proposes that PAE is not an important risk exposure that can adversely impact child development or that FAS is not an important medical disorder. However, as legitimately concerned as advocates, clinicians, and researchers may be about PAE, learnings from epidemiology have the potential to improve the evidence base of our understanding of child mental health and, by extension, improving intervention services and health policies. To ignore such evidence raises the risk for misattribution, lost opportunities, and potential harm. It is also critical to retain attention on the relative impacts of multiple etiological factors contributing to mental health and to be wary of approaches emphasizing single risk factor models.
Acknowledgements
Thanks to Dr Peter Braunberger, Dr Kathy Georgiades, and Dr Michael Boyle for critical feedback on earlier drafts of this paper, as well as recommendations by the anonymous reviewers that helped improve the final paper. In addition, thanks to several people who have previously raised the issue of potential misattribution in the field of FASD, such as Dr Elizabeth M Armstrong, Dr Ben Gibbard, and Dr Barry Zuckerman. No funding was provided for this research and the author has no conflicts of interest to declare.
References
1. Chudley AE, Conry J, Cook JL, et al. Fetal alcohol spectrum disorder: Canadian guidelines for diagnosis. CMAJ. 2005;172(5 Suppl):S1–S21. 2. Stratton KR, Howe CJ, Battaglia FC. Fetal alcohol syndrome: diagnosis, epidemiology, prevention, and treatment. Washington (DC): National Academy Press; 1996. The Canadian Journal of Psychiatry, Vol 60, No 12, December 2015 W 589
Perspective 3. Astley SJ, Clarren SK. Diagnosing the full spectrum of fetal alcohol-exposed individuals: introducing the 4-digit diagnostic code. Alcohol Alcohol. 2000;35(4):400–410. 4. Mattson SN, Riley EP, Gramling L, et al. Neuropsychological comparison of alcohol-exposed children with or without physical features of fetal alcohol syndrome. Neuropsychology. 1998;12(1):146–153. 5. Nash K, Rovet J, Greenbaum R, et al. Identifying the behavioural phenotype in fetal alcohol spectrum disorder: sensitivity, specificity and screening potential. Arch Womens Ment Health. 2006;9(4):181–186. 6. Canadian Association of Paediatric Health Centres. National screening tool kit for children and youth identified and potentially affected by FASD [Internet]. Ottawa (ON): Canadian Association of Paediatric Health Centres, Public Health Agency of Canada; 2012 [cited 2015 May 14]. Available from: http://ken.caphc.org/xwiki/bin/ view/FASDScreeningToolkit/National+Screening+Tool+Kit+for+ Children+and+Youth+Identified+and+Potentially+Affected+ by+FASD. 7. Hoyme HE, May PA, Kalberg WO, et al. A practical clinical approach to diagnosis of fetal alcohol spectrum disorders: clarification of the 1996 institute of medicine criteria. Pediatrics. 2005;115(1):39–47. 8. Price KJ, Miskelly KJ. Why ask why? logical fallacies in the diagnosis of fetal alcohol spectrum disorder. Ethics Behav. 2015:25(5):418–426. 9. Rogler LH. Making sense of historical changes in the diagnostic and statistical manual of mental disorders: five propositions. J Health Soc Behav. 1997;38(1):9–20. 10. American Psychiatric Association (APA). Diagnostic and statistical manual of mental disorders. 5th ed. Washington (DC): APA; 2013. 11. Rothman KJ. Epidemiology an introduction. New York (NY): Oxford University Press; 2002.
590 W La Revue canadienne de psychiatrie, vol 60, no 12, décembre 2015
12. Streissguth A. Fetal alcohol syndrome: a guide for families and communities. Baltimore (MD): Paul H Brookes Publishing Co, Inc; 1997. 13. Berkson J. Limitations of the application of fourfold table analysis to hospital data. Biometrics. 1946;2(3):47–53. 14. Burd L, Carlson C, Kerbeshian J. Fetal alcohol spectrum disorders and mental illness. Int J Disabil Hum Dev. 2007;6(4):383–396. 15. Aragon AS, Coriale G, Fiorentino D, et al. Neuropsychological characteristics of Italian children with fetal alcohol spectrum disorders. Alcohol Clin Exp Res. 2008;32(11):1909–1919. 16. Waldman ID, Gizer IR. The genetics of attention deficit hyperactivity disorder. Clin Psychol Rev. 2006;26(4):396–432. 17. Streissguth AP, Bookstein FL, Barr HM, et al. Risk factors for adverse life outcomes in fetal alcohol syndrome and fetal alcohol effects. J Dev Behav Pediatr. 2004;25(4):228–238. 18. May PA, Tabachnick BG, Gossage JP, et al. Maternal factors predicting cognitive and behavioral characteristics of children with fetal alcohol spectrum disorders. J Dev Behav Pediatr. 2013;34(5):314–325. 19. Sood B, Delaney-Black V, Covington C, et al. Prenatal alcohol exposure and childhood behavior at age 6 to 7 years: I. dose-response effect. Pediatrics. 2001;108(2):E34. 20. Rodriguez A, Olsen J, Kotimaa AJ, et al. Is prenatal alcohol exposure related to inattention and hyperactivity symptoms in children? Disentangling the effects of social adversity. J Child Psychol Psychiatry. 2009;50(9):1073–1083. 21. D’Onofrio BM, Van Hulle CA, Waldman ID, et al. Causal inferences regarding prenatal alcohol exposure and childhood externalizing problems. Arch Gen Psychiatry. 2007;64(11):1296–1304. 22. Costello EJ, Compton SN, Keeler G, et al. Relationships between poverty and psychopathology: a natural experiment. JAMA. 2003;290(15):2023–2029.
www.LaRCP.ca
Perspective
Misattributions and Potential Consequences: The Case of Child Mental Health Problems and Fetal Alcohol Spectrum Disorders John D McLennan, MD, MPH, PhD, FRCPC1 1
Associate Professor, Departments of Pediatrics, Psychiatry, and Community Health Sciences, Cumming School of Medicine, University of Calgary, Calgary, Alberta. Correspondence: Teaching, Research and Wellness Building, 3rd Floor, 3280 Hospital Drive NW, Calgary, AB T2N 4Z6; [email protected].
Key Words: fetal alcohol spectrum disorder, mental disorders, bias, comorbidity, child Received January 2015, revised, and accepted June 2015.
N
umerous Canadian agencies have prioritized services for people diagnosed with FAS and its broader construct FASD. This prioritization extends to prevention interventions aimed at reducing or eliminating PAE. The many difficulties identified as associated with FAS, FASD, and PAE is one of the justifications for this prioritization. Mental health symptoms and disorders are among the most commonly highlighted challenges experienced by people labelled with FAS or FASD or exposed to alcohol in utero. However, the extent of the relation between the FAS, FASD, and PAE cluster and mental health symptom and disorder clusters may be inflated secondary to at least 3 factors: diagnostic criteria overlap and etiologic assumptions, referral bias, and failure to control for confounding variables when assessing associations. Lack of awareness of these factors may lead to dissemination of misinformation, which could adversely distort the development and provision of mental health services. Diagnostic criteria for conditions falling under the FASD umbrella have been operationalized in several different guidelines. The 2005 Canadian guideline1 aimed, in part, to harmonize aspects of the 2 leading approaches at the time, that is, those of the Institute of Medicine2 and the Washington 4-digit diagnostic code.3 Guidelines typically include the complete syndrome, FAS, and require positive findings in 4 domains: problematic patterns of alcohol exposure in utero (for example, from maternal binge drinking), growth abnormalities (for example, low birth weight for gestational age), facial dysmorphology (for example, short palpebral fissures), and CNS neurodevelopment abnormalities (for example, microcephaly at birth).2 Guidelines then typically go on to describe various partial syndromes. In the case of the Canadian guideline, the following partial syndromes are included: FAS (without confirmed alcohol exposure), partial FAS, and ARND.1 Difficulties identified within children with high PAE, but who do not have classical dysmorphic manifestations, is used to support the inclusion of partial syndromes.4 However, this broadening likely contributes to problematic overlap with those children with mental health difficulties for whom PAE may be present but for whom it is not etiologic. This is particularly problematic as there is no consensus on a pathognomonic behavioural manifestation of PAE or FASD. Although some propose a unique mental health profile linked to FASD,5 such profiles are based on small clinical samples and do not appear to have been independently replicated using a nonreferred population. Nevertheless, a resulting 10-item screening tool appears to be receiving national promotion in Canada.6 Concerns that weaknesses in the operationalization of partial FASD syndromes may lead to misattribution of PAE as causal for various difficulties (for example, behavioural problems) has been raised in critiques of the 2 dominant diagnostic approaches used in the field,2,3 that is, the US sources for the Canadian guidelines.7 The ARND diagnosis, within the Canadian guidelines, requires “evidence of impairment in three or more of the following CNS domains: hard and soft neurologic signs; brain structure; cognition; communication; academic achievement; memory; executive functioning and abstract reasoning; attention deficit/hyperactivity; adaptive behavior, social skills, and social communication.”1, p S12 However, abnormalities in 3 of the listed domains would also
www.TheCJP.ca
The Canadian Journal of Psychiatry, Vol 60, No 12, December 2015 W 587
Perspective
be commonly found in many children with various mental health disorders. PAE may not be uncommon in children with such problem clusters. However, the fraction for whom PAE is primarily etiologic is unknown, and to assume it is typically the leading etiology is highly problematic.8 Complicating the picture is the inclusion of an etiologic variable, in this case PAE, in the diagnostic criteria for FASD. This is at odds with a key direction taken in psychiatry as reflected in contemporary versions of the DSM, that is, to avoid etiologic assumptions within diagnostic criteria.9 That the DSM-5 has included neurobehavioural disorders associated with prenatal alcohol exposure under the section “Conditions for Further Study”10 is perhaps a new exception. Unfortunately, as currently written, this DSM5 provisional criteria may contribute to further diagnostic overlap with mental disorders not caused by PAE. This is in part due to the low threshold set for diagnostic criteria for this provisional disorder. A leading epidemiologist has noted that the inclusion of proposed etiological variables in diagnostic criteria may undermine the validity of epidemiologic evaluations of such defined disorders.11 Further complicating considerations of diagnostic overlap and (or) potential comorbidity is a direction taken by some in the FASD community to label mental disorder manifestations in children with PAE or FASD as being secondary disabilities.12 Although this may be valid in some situations, it is concerning if there is a general assumption that manifestations of mental disorders in the context of PAE or FASD are typically secondary phenomena. Relying primarily on the temporal relation is inadequate to make this linkage. Various mental disorders have different times of manifestation, regardless of PAE, and always manifest postpartum. It is not clear how a clinician could determine that the emergence of a mental disorder is secondary to PAE or FASD and not a function of genetic or other environmental factors that may underlie timing of mental disorder emergence. Leaving the problem of diagnostic criteria overlap and etiologic assumptions aside, referral bias (also known as Berksonian or selection bias) may also lead to misattribution. Referral bias may occur when a factor of interest in the clinic population (for example, psychiatric symptoms) also influences whether a patient is seen in the clinic.13 This may lead to spurious correlations and generalizations. Apparent associations identified from studies using clinic Abbreviations ADHD
attention-deficit hyperactivity disorder
ARND
alcohol-related neurodevelopmental disorder
CNS
central nervous system
DSM
Diagnostic and Statistical Manual of Mental Disorders
FAS
fetal alcohol syndrome
FASD
fetal alcohol spectrum disorders
PAE
prenatal alcohol exposure
588 W La Revue canadienne de psychiatrie, vol 60, no 12, décembre 2015
Clinical Implications •
Clinicians and researchers should attend to diagnostic criteria overlap and etiologic assumptions, referral bias, and potential influence of confounders when considering associations between different disorders, symptom clusters, and purported risk factors.
•
Some mental health problems in children diagnosed with FASD may be inaccurately attributed to in utero alcohol exposure and (or) FASD.
•
In some cases, children with mental health disorders may be mismanaged secondary to misattributing most or all their difficulties to FASD.
Limitations •
This paper is not based on a systematic literature review.
•
The number of high-quality source documents to evaluate the relation between FASD and child mental health disorders is sparse.
samples that have not been verified in nonreferred samples warrant considerable skepticism. Although it is possible that a child may be sent to an FASD clinic based solely on unique aspects of facial dysmorphology, factors such as disruptive behaviours are much more likely to drive referrals. Consequently, specialized FASD assessment clinics are likely receiving referrals of children and youth with a disproportionally high occurrence of disruptive behaviours. Clinicians then experience high rates of cooccurrence, and the researchers who rely on studies of clinically referred samples may report strong but potentially spurious associations. Unfortunately, most FAS or FASD studies appear to rely on clinic samples. Fortunately, there are a few population-based samples that include an examination of psychiatric morbidity among children assessed for PAE and (or) FAS or FASD and that may help identify whether referral bias underlies some of the high comorbidity noted in clinical populations. ADHD may be the most studied psychiatric comorbidity in FASD. One review found an overall ADHD prevalence of 48% in children with FASD14; however, most source studies in that review used clinic-referred samples. In contrast, a nonreferred, school-sampled study found a combined-type ADHD prevalence of 8.7% among children with FASD based on teacher ratings, compared with 5.3% among children without FASD, a nonsignificant difference.15 The marked differences in ADHD prevalence found in children with FASD in this population-based study, compared with the review of clinic-based studies, may lead to the hypothesis that a substantial portion of ADHD found in children with FASD in referred clinic samples have ADHD as a function of a wider range of possible etiologies.16 Unfortunately, prevalence of other psychiatric disorders have not been reported from population-based studies of FASD; however, referral bias is also likely for other conditions that are apt to precipitate referrals (for example, aggression, conduct disorder, and developmental delays). www.LaRCP.ca
Misattributions and Potential Consequences: The Case of Child Mental Health Problems and Fetal Alcohol Spectrum Disorders
Whether such distortions extend to internalizing problems and disorders is unknown. Confounding variables may also contribute to an overestimation of the strength of relations between the FAS, FASD, and PAE cluster and mental health problems. Children referred to FASD clinics are often found to have had many risk exposures in addition to PAE, such as being victims of physical or sexual abuse and exposure to parental substance use after birth.17 It is unclear how clinicians can attribute mental health presentations to the PAE, compared with one or more of the other risk exposures. In some assessment and advocacy contexts, mental health symptoms and disorders may be framed as primarily or entirely explained by PAE. Such assumptions do not seem to hold up as suggested by findings from numerous studies aimed at examining the extent to which PAE independently explains adverse child outcomes relative to other known risk factors for child mental health problems. For example, in a FASD study within a school-based sample, measures of PAE explained only a modest portion of cognitive and behavioural outcomes and these association were no longer significant once other risk factors (for example, lower maternal education) were included in the multivariate model.18 Another study, based on a sample derived from a maternity clinic, found PAE was related to child behavioural problems at follow-up but explained only 0.6% to 1.7% of outcome variance; in contrast, a measure of maternal psychopathology explained 13.0% to 29.1% of the variance.19 In a population-based study examining the impact of lowto-moderate PAE, the association with ADHD symptoms became statistically insignificant once the multivariate model factored in measures of social adversity.20 Finally, another large population-based study attempted to consider genetic and other environmental contributions beyond PAE by including siblings and cousins with variable PAE.21 They found that the significant differences in attentional and impulsivity problems between children with and without PAE for unrelated families was not found within related family.21 However, this study did find some evidence for a significant relation between PAE and conduct problems after controlling for family membership.21 Although measurement issues and epidemiologic-based findings may, at times, seem disconnected from clinical practice, in this case, failure to appreciate these underlying patterns may have distorting effects on services and policies. For example, at a service level, if a comorbid psychiatric disorder is inaccurately attributed to FASD, this may lead to an inappropriate diversion to a newly designed FASD specialty intervention that lacks evidence of effectiveness for the comorbid psychiatric problem rather than being directed to an evidence-based intervention specific to the identified psychiatric comorbidity. For example, use of evidence-based behavioural interventions or medications for ADHD may be discouraged if the presenting attentional problem is attributed to FASD. A second example may include the setting of eligibility for certain services and supports based on the presence of an FASD diagnosis www.TheCJP.ca
rather than a given identified need that may not be specific to FASD (for example, a social skills deficit). This may reinforce unsubstantiated assumptions made about children with FASD (for example, they all have social skill deficits) and compromise access to a service for a child who may have a need for additional services (for example, social skills training) but who had, say, prenatal polysubstance exposure rather than PAE. A further concern is a potential for harm. It has been proposed that flawed causal models that may be employed within some FASD-focused clinics may violate the ethical principle of nonmaleficence.8 More specifically, assumptions made about causes may inadvertently revive past psychiatric frameworks that emphasized maternal behaviour as etiologic in child psychiatric disorders.8 Finally, the failure to consider an epidemiologic perspective may also distort prevention efforts which may, for example, overemphasize a single risk exposure (for example, elimination of drinking during pregnancy) in efforts to improve child mental health in the population. This may result in relative neglect of other important risk factors contributing to child mental health problems, such as poverty.22 None of the preceding argument proposes that PAE is not an important risk exposure that can adversely impact child development or that FAS is not an important medical disorder. However, as legitimately concerned as advocates, clinicians, and researchers may be about PAE, learnings from epidemiology have the potential to improve the evidence base of our understanding of child mental health and, by extension, improving intervention services and health policies. To ignore such evidence raises the risk for misattribution, lost opportunities, and potential harm. It is also critical to retain attention on the relative impacts of multiple etiological factors contributing to mental health and to be wary of approaches emphasizing single risk factor models.
Acknowledgements
Thanks to Dr Peter Braunberger, Dr Kathy Georgiades, and Dr Michael Boyle for critical feedback on earlier drafts of this paper, as well as recommendations by the anonymous reviewers that helped improve the final paper. In addition, thanks to several people who have previously raised the issue of potential misattribution in the field of FASD, such as Dr Elizabeth M Armstrong, Dr Ben Gibbard, and Dr Barry Zuckerman. No funding was provided for this research and the author has no conflicts of interest to declare.
References
1. Chudley AE, Conry J, Cook JL, et al. Fetal alcohol spectrum disorder: Canadian guidelines for diagnosis. CMAJ. 2005;172(5 Suppl):S1–S21. 2. Stratton KR, Howe CJ, Battaglia FC. Fetal alcohol syndrome: diagnosis, epidemiology, prevention, and treatment. Washington (DC): National Academy Press; 1996. The Canadian Journal of Psychiatry, Vol 60, No 12, December 2015 W 589
Perspective 3. Astley SJ, Clarren SK. Diagnosing the full spectrum of fetal alcohol-exposed individuals: introducing the 4-digit diagnostic code. Alcohol Alcohol. 2000;35(4):400–410. 4. Mattson SN, Riley EP, Gramling L, et al. Neuropsychological comparison of alcohol-exposed children with or without physical features of fetal alcohol syndrome. Neuropsychology. 1998;12(1):146–153. 5. Nash K, Rovet J, Greenbaum R, et al. Identifying the behavioural phenotype in fetal alcohol spectrum disorder: sensitivity, specificity and screening potential. Arch Womens Ment Health. 2006;9(4):181–186. 6. Canadian Association of Paediatric Health Centres. National screening tool kit for children and youth identified and potentially affected by FASD [Internet]. Ottawa (ON): Canadian Association of Paediatric Health Centres, Public Health Agency of Canada; 2012 [cited 2015 May 14]. Available from: http://ken.caphc.org/xwiki/bin/ view/FASDScreeningToolkit/National+Screening+Tool+Kit+for+ Children+and+Youth+Identified+and+Potentially+Affected+ by+FASD. 7. Hoyme HE, May PA, Kalberg WO, et al. A practical clinical approach to diagnosis of fetal alcohol spectrum disorders: clarification of the 1996 institute of medicine criteria. Pediatrics. 2005;115(1):39–47. 8. Price KJ, Miskelly KJ. Why ask why? logical fallacies in the diagnosis of fetal alcohol spectrum disorder. Ethics Behav. 2015:25(5):418–426. 9. Rogler LH. Making sense of historical changes in the diagnostic and statistical manual of mental disorders: five propositions. J Health Soc Behav. 1997;38(1):9–20. 10. American Psychiatric Association (APA). Diagnostic and statistical manual of mental disorders. 5th ed. Washington (DC): APA; 2013. 11. Rothman KJ. Epidemiology an introduction. New York (NY): Oxford University Press; 2002.
590 W La Revue canadienne de psychiatrie, vol 60, no 12, décembre 2015
12. Streissguth A. Fetal alcohol syndrome: a guide for families and communities. Baltimore (MD): Paul H Brookes Publishing Co, Inc; 1997. 13. Berkson J. Limitations of the application of fourfold table analysis to hospital data. Biometrics. 1946;2(3):47–53. 14. Burd L, Carlson C, Kerbeshian J. Fetal alcohol spectrum disorders and mental illness. Int J Disabil Hum Dev. 2007;6(4):383–396. 15. Aragon AS, Coriale G, Fiorentino D, et al. Neuropsychological characteristics of Italian children with fetal alcohol spectrum disorders. Alcohol Clin Exp Res. 2008;32(11):1909–1919. 16. Waldman ID, Gizer IR. The genetics of attention deficit hyperactivity disorder. Clin Psychol Rev. 2006;26(4):396–432. 17. Streissguth AP, Bookstein FL, Barr HM, et al. Risk factors for adverse life outcomes in fetal alcohol syndrome and fetal alcohol effects. J Dev Behav Pediatr. 2004;25(4):228–238. 18. May PA, Tabachnick BG, Gossage JP, et al. Maternal factors predicting cognitive and behavioral characteristics of children with fetal alcohol spectrum disorders. J Dev Behav Pediatr. 2013;34(5):314–325. 19. Sood B, Delaney-Black V, Covington C, et al. Prenatal alcohol exposure and childhood behavior at age 6 to 7 years: I. dose-response effect. Pediatrics. 2001;108(2):E34. 20. Rodriguez A, Olsen J, Kotimaa AJ, et al. Is prenatal alcohol exposure related to inattention and hyperactivity symptoms in children? Disentangling the effects of social adversity. J Child Psychol Psychiatry. 2009;50(9):1073–1083. 21. D’Onofrio BM, Van Hulle CA, Waldman ID, et al. Causal inferences regarding prenatal alcohol exposure and childhood externalizing problems. Arch Gen Psychiatry. 2007;64(11):1296–1304. 22. Costello EJ, Compton SN, Keeler G, et al. Relationships between poverty and psychopathology: a natural experiment. JAMA. 2003;290(15):2023–2029.
www.LaRCP.ca
