Good open science improves the lives of America’s youth
In the United States, adolescents spend more than one hour a day on social media during school hours. Measures of brain connectivity in childhood can predict future depression symptoms. One in four 9–10-year-olds report having sipped alcohol. These and hundreds of other findings relevant for youth mental and physical health were made possible by the landmark Adolescent Brain Cognitive DevelopmentSM (ABCD) Study, the largest long-term investigation of brain development and child health in the United States.
For the past ten years, over 10,000 youth participants have contributed their brain imaging, cognitive test results, questionnaire responses about physical health, mental health, and everyday activities, hormone samples, genetic data, and much more to help scientists better understand the factors underlying risk and resilience for mental health and substance use disorders. A recent NYT article reports that scientists involved in conducting the ABCD Study® did not keep these data safe, leading to unethical and unauthorized use by a group of individuals seeking to argue for the innate intellectual superiority of White people. As scientists conducting the study, we want to clarify misleading statements about ABCD in the NYT article and expand on some important open science principles and ethical guardrails that were not fully covered in the story.
ABCD was designed using an open science model, which maximizes the scientific benefits of the data collected by making them available to qualified researchers. It is designed with careful attention to data governance. All data are “de-identified”; this means that personally-identifying information (such as names and date of birth) is never available to data users. Access to the de-identified data is restricted to qualified researchers who must submit proposals, agree to strict data use certifications, and undergo training in responsible data handling and ethical use. However, technical safeguards to data access cannot protect against analytic and interpretive misuse.
We were profoundly disturbed and concerned that data intended to enhance the scientific understanding of healthy development in youth were accessed by individuals promoting a stigmatizing agenda. The validity of their claims has been challenged by many scientists who point out that racial groups often experience different social and environmental conditions that likely influence mental functions. The science is complex and nuanced, but there is no ethical defense for conducting research intended to amplify bias against certain racial groups. This unethical use of scientific data disproportionately harms individuals from marginalized communities.
We did not publicize the fact that some individuals were using ABCD Study data to conduct faulty, harmful studies because 1) no confidential information was breached, 2) we did not want to promote unethical science that could cause harm to study participants, and 3) most of the findings were posted on blogs, and not published in peer-reviewed academic journals. We immediately brought the issue to the attention of multiple journal editors and published a series of guidelines for responsible use of data. What we did not do is give a platform to the unethical and distressing claims made by the individuals who had wrongfully gained access to multiple datasets, including ABCD. We are actively communicating with families who participate in the ABCD Study to provide clear information regarding what transpired, receive feedback, and address these issues.
The case of data misuse reported by the NYT focused on an individual who gave deliberately misleading information when applying to gain access to de-identified ABCD Study data. After accessing the data, he also re-shared it with an unauthorized person – further violating the terms of the agreement. After an investigation confirmed his violations, the first individual was barred from access to data. The NYT article reported that 16 papers purporting to find biological differences in intelligence between races were published. Many readers assumed that these papers used ABCD data. We have learned that none of them did. The author of these 16 papers has published nothing using ABCD data; however, the associate with whom he shared the data has subsequently published results in online journals. This misuse of ABCD data is highly unethical and disturbing.
We are proud that the open science model adopted by the ABCD Study has led to over 1800 peer-reviewed research articles in respected scientific journals, providing valuable insights into mental health, physical health, and brain development during adolescence. Unethical activity and misuse of the study data are not acceptable, and our National Institutes of Health sponsors have continued to take actions to help prevent such activity in the future by adding new safeguards to ensure that only qualified scientists are granted data access. They also require anyone accessing the de-identified data to complete a responsible data use training module. This type of training helps scientists understand the ethical implications of their research and the historical inequities that should be carefully considered when interpreting their results. Additional guardrails protecting access to data and emphasizing strict guidelines for data management have recently been put in place.
While technical safeguards like de-identifying data and strictly overseeing data access are necessary in research, they are not enough. Scientists, journal editors, reviewers, and the public all play a role in how research is interpreted and shared, particularly as scientific findings circulate through social media. Without critical review of research output, data can be misused in ways that reinforce biases and ignore social context, causing harm to individuals and more globally by increasing health disparities for the most impacted individuals. Mitigating this harm requires a shared commitment to uphold standards of accountability and ethical responsibility.
As the ABCD Study enters its second decade, we are poised to make increasingly impactful contributions to science and society. Data from ABCD can be used to answer some of the most pressing questions in public health. How does cannabis use affect brain development? How does social media use during adolescence affect adult health and functional outcomes? What social and environmental factors influence mental health? What contributes to a happy and healthy life? Because of data generously provided by the participants of the ABCD Study, we will soon begin to have answers to these and many other questions relevant to today’s youth. While doing this work, we will continue to uphold the highest standards around strict data access policies and ethical use of data. We will not lose sight of our primary mission to conduct science that improves the lives of others.