Search This Blog

Sunday, August 21, 2016

ASD Validity

From the Review Journal of Autism and Developmental Disorders
via Springer Link

By Lynn WaterhouseEric London, and Christopher Gillberg
August 10, 2016

"The paper concludes that the evidence does not provide support for the neurobiological or construct validity of the ASD diagnosis, and therefore the ASD diagnosis should be disbanded in research."

Abstract

ASD research is at an important crossroads. The ASD diagnosis is important for assigning a child to early behavioral intervention and explaining a child’s condition. But ASD research has not provided a diagnosis-specific medical treatment, or a consistent early predictor, or a unified life course.

If the ASD diagnosis also lacks biological and construct validity, a shift away from studying ASD-defined samples would be warranted.

Consequently, this paper reviews recent findings for the neurobiological validity of ASD, the construct validity of ASD diagnostic criteria, and the construct validity of ASD spectrum features. The findings reviewed indicate that the ASD diagnosis lacks biological and construct validity. The paper concludes with proposals for research going forward.

Download Article PDF

The goal of the DSM-3 nosology (American Psychiatric Association 1980) was to create reliable and standard categorical psychiatric diagnoses (Robins and Guze 1970). However, in the past 30 years, clinical, genetics, and neuroscience findings have revealed that the DSM diagnoses are not biologically valid.


The National Institutes of Mental Health (NIMH) responded by proposing the Research Domain Criteria (RDoC) framework for a brain-based transdiagnostic psychiatric symptom nosology (Cuthbert and Insel 2013; Insel et al. 2010; Lilienfeld and Treadway 2016). Peterson (2015) and Weinberger et al. (2015) argued that the RDoC could not replace the DSM-5 psychiatric nosology (American Psychiatric Association 2013) or the parallel International Classification of Diseases (ICD) psychiatric nosology (World Health Organization2012).

But, “For the foreseeable future, RDoC is not envisioned as a system of psychiatric classification in its own right. Instead, in the near term, RDoC and DSM-ICD are expected to coexist. Nevertheless, RDoC is intended to provide scaffolding for a large-scale research program that will ultimately yield an alternative to DSM-ICD” (Lilienfeld and Treadway 2016, p. 445).

RDoC advocates accept that DSM-5/ICD psychiatric categories remain necessary in clinical practice, but argue that researchers should shift to RDoC study designs immediately. They assert that studying psychiatric categories lacking biological validity blocks the discovery of brain bases for psychopathology, and thus cannot lead to effective medical treatments for specific psychiatric symptoms (Cuthbert and Insel 2013; Insel et al. 2010; Lilienfeld and Treadway 2016; Yee et al. 2015).

Against this RDoC imperative for biological validity, Weinberger et al. (2015) countered that current DSM-5 psychiatric behavioral diagnoses were valid when they yielded effective medical treatment, clear prognosis, and a life course specific to a diagnosis.

Autism spectrum disorder (ASD) research has been productive (Dawson 2016; de la Torre-Ubieta et al. 2016; Szatmari et al. 2016), but no ASD research findings have met the validity criteria of Weinberger et al. (2015).

DSM-5 ASD research has found no specific effective pharmacotherapy or other medical treatment (Na Young and Findling 2015). The only broadly successful ASD treatment has been early behavioral intervention programs (Kasari 2015; Schreibman et al. 2015; Smith and Iadarola 2015; Volkmar et al. 2014), but these treatments are not unique to ASD (Losinski et al. 2014), and the long-term effectiveness of these programs is not yet known (Fernell et al. 2011, 2014).

Researchers who studied infant siblings of children with ASD concluded no single early behavior could predict ASD diagnoses (Zwaigenbaum et al.2015), and ASD has been found with widely varied trajectories of development (Fountain et al.2012; Lord et al. 2015) and varied life outcomes (Fein et al. 2013; Helles et al. 2015; Steinhausen et al. 2016).

ASD research is at a crucial decision point. The ASD diagnosis remains necessary in the clinic to assign a child to early behavioral intervention and to explain a child’s condition. But researchers must decide whether to continue studying ASD-defined samples or not.

Given that ASD lacks any diagnosis-specific medical treatment, any consistent early predictor, or any specific life course, if the ASD diagnosis also lacks biological and construct validity, a shift away from studying ASD-defined samples would be warranted.

Consequently, this paper reviews recent findings for ASD biological and construct validity. This paper is not a meta-analysis; instead it brings together competing and unresolved findings. The first section examines evidence for the neurobiological validity of ASD. The second section outlines evidence for the construct validity of ASD diagnostic criteria. The third section explores evidence for the construct validity of ASD spectrum features beyond ASD diagnostic symptoms.

The paper concludes that the evidence reviewed does not provide support for the neurobiological or construct validity of the ASD diagnosis, and therefore the ASD diagnosis should be disbanded in research.

1 comment: