Science is not solitary. The image of a lone researcher working late into the night in the lab doesn’t come close to capturing how scientific progress is made. Dedicated scientists, money, and time are all important, but amid public uncertainty over scientific funding, one more component valuable to the research process often flies under the radar. Data sharing is where researchers at one institution provide their results upon request to those working on the same disease worldwide, and it has significantly accelerated progress towards cures for many serious diseases. Ailments that were considered death sentences 25 years ago, such as childhood cancer and AIDS, have been rendered treatable more rapidly than otherwise possible thanks to data sharing. Now, data sharing is creating a positive impact on Alzheimer’s disease research.
Alzheimer’s is a debilitating disease, but its tragedy spreads beyond just the patient. Dr. Eliezer Masliah, Director of the National Institute on Aging’s (NIA) Division of Neuroscience, said, “[Alzheimer’s] has a devastating effect … on the whole family.”
Currently, as many as 5.5 million older Americans live with Alzheimer’s. By 2050, the NIA expects that number to reach nearly 14 million. In the opinion of Dr. Masliah, we need treatment for Alzheimer’s, and we need it fast. Data sharing helped advance our understanding of the disease to where it is today and only promises to take us further. He noted, “[The global research community] has learned tremendously about the evolution of Alzheimer’s disease … and that was only possible through sharing data.”
To the extent that it can, the federal government is working to promote data sharing at research institutions nationwide. The NIA funds thirty-two separate Alzheimer’s Disease Centers across the country, located at major medical establishments and universities. Since their funding ultimately comes from taxpayers, these centers have an obligation—often contractual—to open their data to peers. As the NIA continues to fund Alzheimer’s research, it is now requiring that research be shared.
So how exactly does data sharing help researchers?
The simple answer: divide and conquer. While quality data are the building blocks of scientific progress, they require interpretation once collected. A single dataset can be repurposed and analyzed in a myriad of ways, but limited skillsets, time, and money hold researchers back from fully exploring what their data mean. Thus, opening research to the scientific community lets others do that work for them. Researchers can draw new conclusions from old studies or recognize broad trends across many datasets. As an added benefit, data sharing also validates the original research. Dr. David Bennett, Director of Rush University’s Alzheimer’s Disease Center (RADC) said, “I find it particularly flattering when the other twenty-eight [Alzheimer’s research centers] want to use our data … I view that as a sign of the quality of the data and material we’re collecting.”
The RADC has had an open data policy for decades, and now promotes its own online hub, which organizes their extensive collection of complex data and biospecimens. Its mission is to assist non-RADC researchers with finding data to support their own projects.
Dr. Sterling Johnson, Principal Investigator for the Wisconsin Registry for Alzheimer’s Prevention (WRAP) study also views data sharing as essential for Alzheimer’s research progress. “We recognize,” he said, “that the effort and commitment of our participant-volunteers is best honored by using the data they have provided to find discoveries about AD. The fastest way to new knowledge about Alzheimer’s disease is to make our rich dataset available to talented researchers.”
Outside researchers can request WRAP’s data either directly through the study’s website or through gaain.org, an online repository sponsored by the Alzheimer’s Association. The scientific executive committee reviews and generally grants each request. Then, WRAP’s data team works with the researchers to give them the correct data and make sure they understand how it’s organized. The team takes care to ensure their findings are accessible to skilled researchers across the globe.
The most prominent project to embrace an open-data policy is the expansive Alzheimer’s Disease Neuroimaging Initiative (ADNI). The public and private-sector institutions that collaborated to found the study nearly two decades ago had the visionary idea to share the data amassed, not after results were published, but as it was being collected. Dr. Masliah noted, “Nowadays, we all talk about [data sharing], but in those days, nobody was talking about it. Now we find ourselves almost two decades after ADNI was funded, and the impact has been extraordinary.”
But for all its positive impacts, sharing data does not come without complications.
For businesses, research data is both costly and valuable, so opening the data it produces to other researchers is not often seen as a wise business strategy. Pharmaceutical companies have earned a bad reputation for keeping their data private, but in recent years they have actually begun to create means which by interested parties can access their patient-level clinical trials data.[1] By cataloging thousands of past trials, these companies are offering a trove of previously-untapped, quality data of which researchers are just now beginning to take advantage.
Nevertheless, this newfound openness is the exception, and corporations aren’t the only institutions with secretive attitudes towards their data. Even top universities—producing excellent data—aren’t always eager to share their research. At many universities, conducting “original” research is a requirement for promotion. The academic boards don’t consider new analysis of preexisting data independent enough to merit personal distinction. This forces aspiring professors to keep their studies very close to the chest.
So as Alzheimer’s research advances, how can more research centers share? Dr. Bennett sees two paths forward. First, the National Institutes of Health can and already are coercing more research institutions to open their data. Second, the reward system for scientists could be altered. At many universities, if a researcher shares her data and another group uses it to publish their own findings, she gets zero credit. “We need to change the reward structure to bring it into the 21st century,” said Dr. Bennett.
Interestingly enough, the physics field has long embraced a team-oriented view of progress. Smashing atoms and building rockets require thousands of people from varied backgrounds. Young researchers work with their mentors, rather than under them. A single physics paper can have over 400 coauthors. Everyone works together, and everyone thrives together.
Likewise, Alzheimer’s disease, Dr. Masliah said, “is not something one lab or one person working in a silo can resolve. It takes a huge effort by the whole community.”
Chase Gardner
Intern
University of Chicago
[1] Navar, Ann Marie, Michael J Pencina, Jennifer A Rymer. “Use of Open Access Platforms for Clinical Trial Data.” JAMA. 2016; 315 (12): 1283-1284. doi:10.1001/jama.2016.2374