Tag Archives: Research

Runkeeper & Research: The Keeping Pace Study

Yesterday evening I laced up my running shoes, connected my bluetooth headphones, turned on my Spotify playlist, and most importantly, hit “Go Running” on my Runkeeper app. About an hour later and I had run 6 miles at a decent pace of around 8 minutes per mile. And I knew this thanks to Runkeeper.

Founded in 2008, Runkeeper is designed to assist individuals who want to track their activities with GPS precision, whether that is walking, running, hiking, or cycling. If you’re moving outdoors, Runkeeper and similar apps, such as Strava or MapMyRun, use your smartphone’s GPS to pinpoint exactly where you are and how fast you’re moving. With all that data, you can train for your next marathon, discover new routes, and now, thanks to efforts by New York University researchers, take part in advancing public health research.

“We know from the existing literature that spatial characteristics like walkable neighborhoods and green spaces encourage exercise, but a lot of the details are still unknown.”


Last week, Dr. Rumi Chunara and her colleagues launched the Keeping Pace study. Over the next few months they hope to enroll participants who are willing to share their geo-located exercise data from Runkeeper. Because Runkeeper keeps a log of not only what you did, but where you did it, researchers hope to use the large amount of aggregated data to better understand physical activity patterns in communities around the United States.

“Typically, this type of research takes a long time and includes long, ardorous surveys or giving out GPS devices to participants,” said Dr. Chunara. “But with this type of data from apps people already use, we will be able to understand how the environment and exercise are related over more rapid and longer time periods.” With this data being contributed, the research team hope to understand differences in exercise choice between commuting and recreational activities, variation in activities among neighborhoods, and where people spend their time while being active.

Participants who enroll in Keeping Pace will be asked to complete a short demographic survey and then connect their Runkeeper account so researchers can access the type of activities they do and the GPS-based map associated with the activity. The Runkeeper data connection is being handled by a unique research platform, Open Humans.

A few years ago, Dr. Rumi Chunara was at a meeting hosted by the US Department of Health and Human Services. She was there to present and speak with colleagues about the growing importance of citizen science and crowdsourced data. There she met Jason Bobe, Executive Director of PersonalGenomes.org. They got to talking about some of their common insterests in open data, research, and new models for research participation. Later, when Dr. Chunara was designing GoViral, a project to examine how to leverage crowdsourced flu symptom information and diagnostics to predict illness risk, she ran into some issues with hosting and handling the amount of data participants were contributing. “It was obvious we needed some sort of platform to handle data,” said Dr. Chunara. She got back in touch with Jason, who helped her think about the issues and how to solve them.

This year, when it came to build out the infrastructure for the Keeping Pace study, Dr. Chunara decided to get back in touch with Jason and his colleagues, who were now developing OpenHumans.org. As we’ve written before, Open Humans represents a new way of thinking regarding researcher studies, participants, and the data being transferred between the two. The two teams, Open Humans and Dr. Chunara’s lab at NYU, worked together to develop an easy method for individuals to simultaneously allow researchers access to their Runkeeper data, and also maintain control over where and how that data flowed. Specifically, each individual who chooses to participate in the Keeping Pace study will be asked to create an Open Humans account, connect their Runkeeper account, and then authorize the Keeping Pace study to access their data. It sounds like a lot of work, but thanks to the designers and the use of Runkeeper’s API, it takes no more than five minutes to complete.

But why go through that trouble at all? Why not just have participants export their Runkeeper data and send it to the researchers? Why didn’t Dr Chunara and her colleagues build that data connection themselves?

Thanks to the proliferation of sensors, wearables, and smartphones, the ability to generate data about our lives is rapidly expanding. Pair that data with new efforts like the Precision Medicine Initiative and it’s easy to see the potential for researchers to understand our lives and our health in new and interesting ways. But what about the people who create that data? People, like myself, who strap on their phones when they go out for a run or log onto a website to report their flu symptoms. What do they have a right to in regards to their data? This is the question many researchers and scientific institutions are grappling with. But some have already taken a stand.


“What data are collected and how varies across research studies, but the question remains, ‘Who owns it?’ If someone is spending time generating then they should have control over it.”


Dr. Chunara and her colleagues chose to work with Open Humans because they shared the same perspective — participants should be in control of their data. “Open Humans has created an infrastructure that makes it easy to share and learn while respecting the participant and their data. That’s a noble motive, and it’s important,” said Dr. Chunara. Today, Keeping Pace is the first study to use Open Humans for data access and management for a research study. If successful, researchers may not only learn about exercise and the environment, but also about how studies that place an emphasis on participants’ data access and control may engage the public in new ways.

Keeping Pace is currently enrolling participants. If you’re a Runkeeper user and want to contribute your data to research, please visit the study website to learn more.

Keeping Pace was funded as part of the Agile Projects grants by the Health Data Exploration Network. If you’re a researcher, company, or individual interested in personal health data, sign up to become a network member. Membership is free.

Quantified Self Labs is dedicated to sharing stories and insights about the role of data access for personal and public health. We invite you to share your data access stories and this article. Then follow along on quantifiedself.com and @quantifiedself.


Posted in QS Access | Tagged , , , | Leave a comment

Open Research, Open Data, Open Humans

“Open Humans aims to break down data silos in human health and research. We believe data has a huge potential to live and grow beyond the boundaries a single study or program. Our online portal allows members to aggregate data from the research they participate in. By connecting individuals willing to share existing research data about themselves with researchers who are interested in using that data, data can be re-used and built upon.” — OpenHumans.org

On March 24, 2015 the Open Humans Network officially opened their virtual doors and began allowing individuals to sign up and engage in a new model of participatory research. We spoke with Co-founder & Principal Investigator of the Public Data Sharing study, Madeleine Ball, Ph.D. about Open Humans, what it means for research, and what we can look foward to from this exciting initiative. The following is an edited transcript of that conversation.

It’s been a lot of work up to this point.

We’re grateful to have the funding support of two organizations to help get this off the ground, the Knight Foundation and the Robert Wood Johnson Foundation. It’s been a lot of work to get to this point, from hiring Beau Gunderson as our Senior Software Gardener to launching with our first three studies. We’re excited to be partnering with the Harvard Personal Genomes Project, the American Gut study, and the GoViral study. These are the seed studies, what we’ll build off of in the coming months and years. Today, we’re excited to start letting participants in these projects, and all individuals interested in participating in research, know about Open Humans.

This is an open invitation to join us.

We’re also working to make it easier for research partners to join the Open Humans Network. We’ve already started receiving interest from researchers that want to integrate with Open Humans or start working with our already growing public data sets. We’ve set parameters regarding how you have to behave as a study as well as how researchers looking to work with our members should engage with us. (You can find out more about that here.)

For members who sign up with us we’ve developed methods for them to control access to their data. Whether that is data from personal health devices and apps like Runkeeper (adding this to our next project), genetic data, or other data sources derived from participating studies, each individual member will have the ability to establish a peer-to-peer interaction. Members can allow access to some data, but not others. They may choose to release some or all of their data publicly, or the may choose to only share with one study. In the end it’s up to them and their individuals goals.


What excites me about Open Humans is the potential we have to transform future research studies — from how they treat data to how they think about data sharing. We’re building our system so that participants are central to the data process. A good example of this when researchers use our member’s data they must also agree to return any new data that results from their research back to the original participant. This decentralization of data is a key component of our design. No single person, researchers, or study has all the data.

We’ve also built in the ability for researchers to contact our members who contribute data. The idea that researchers must come up with all the right questions before starting a study is a recipe for failure. Researchers are not psychic, that can’t forsee what interesting questions might come up in the future. By opening up the ability for these connections to take place in the design of Open Humans, we’re creating the ability to continue asking questions of specific individuals, or groups of people, far in to the future.

We’re founded on the principle of transparency. You as a researcher, or participant member can see what we’re all about. You can even see our Open Human member profiles (Madeleine BallJason BobeBeau Gunderson). We worked with Marcia Hoffman, special counsel to the Electronic Frontier Foundation, to develop our Terms of Use and Data Use Policies so that they’re readable and easily understood. We want people to read them, we want them to ask us questions. We want people to be engaged and involved.

I think this work is creating a new form of data sharing that will unlock a world of new exciting possibilities. Our hope is that when participants start getting data back from studies, and have the ability to use it and share it how they wish, that participation in research will be more rewarding. This model helps participants become a respected member of the evolving research conversations happening all over world. We know a lot of people don’t participate in research, even researchers who rely on participants don’t participate in studies. Hopefully this work will help move the needle.

It’s wonderful to see the long scroll of members.

As of this writing the Open Humans Network has over 200 individuals who have created member profiles. If you’re interested in participating in open research you can learn more and sign up here. If you are a researcher or personal data company interested in integrating with Open Humans you can get in touch with the team here.

We invite you to share your data access stories, and this article with the #qsaccess hashtag and follow along here, on Medium, and @quantifiedself.

Posted in QS Access, QS Access Conversations | Tagged , , , , , , | 2 Comments

What We Are Reading

Ernesto is in sunny Austin for SXSW, so I’m filling in to gather this week’s articles and links for your reading pleasure.


Apple ResearchKit concerns, potential, analysis by MobiHealthNews. ResearchKit was a big surprise coming out of Apple’s Special Event this week. It was quite difficult to select just one representative article about the ensuing conversation, so this round-up serves nicely.

#WhatIfResearchKit: What if Research Kit actually, truly, worked… by Christopher Snider. Okay, I failed to keep to one article on ResearchKit. This post chronicles a series of Twitter conversations on the question: if ResearchKit does work, what are the possibilities?

The Electric Mood-Control Acid Test by Kevin Bullis. Thync is a sort of evolved version of a transcranial direct current stimulation (TDCS) device. A technology with a lot of potential and controversy, this article explores why the brain-enhancing effects of the TDCS only work for some people. By the way, if you are a fan of Philip K. Dick, Thync may remind you of the mood organ that was in Do Androids Dream of Electric Sheep?

Automated Learning by Nichole Dobo. Some school classrooms are experimenting with ”Blended learning”, a method of combining classroom teachers and computer-assisted lessons. A detail that stuck with me is the description of three large displays that show where each student is supposed to go that day, based on the results of the previous day’s lesson.

The Mouse Trap: Can One Lab Animal Cure Every Disease? by Daniel Engber. An in-depth how science’s predominant use of lab mice could be limiting our knowledge of disease. Of relevance to self-trackers because many models of optimal health are in part based on mouse studies.



Analyzing a Year of My Sleep Tracking Data by Bob Troia. This is a superb exploration of Bob’s sleep data from 2014 as collected by his Basis watch.


Notes on 416 Days of Treadmill Desk Usage by Neal Stephenson. The author of Snow Crash and The Cryptonomicon is a long time user of a treadmill desk, but when he began having pain in his left leg, he had to reevaluate how he used his favored tool.



Qualities of #QuantifiedSelf by Christina Lidwin. A fascinating analysis of the #quantifiedself hashtag.

Access Links
First medical apps built with Apple’s ResearchKit won’t share data for commercial gain by Fred O’Connor
Talking Next-Gen Diabetes Tools with Dexcom Leaders by Mike Hoskins

From the Forum
Mood Tracking Methods?
Howto track laptop uptime
CCD or CCR conversion tools?
What gets measured, gets managed – Quantified Self in the workplace
Best ECG/EKG Tool for Exercise
Best iOS app to track water/coffee/alcohol intake?

This Week on QuantifiedSelf.com
QS15 Sponsor Highlight: RescueTime
Quantified Self and Apple’s ResearchKit
Better by Default: An Access Conversation with John Wilbanks
QS15 Conference Preview: Jamie Williams on Tracking My Days
Quantified Self Styles

Lastly, I’ll leave you with a lovely little comic with a message that many self-trackers can relate to.

The Secret by Grant Snider


Posted in What We're Reading | Tagged , , , , , , , , , , | Leave a comment

Better by Default: An Access Conversation with John Wilbanks

Photo by Joi Ito

Photo by Joi Ito

John Wilbanks is the Chief Commons Officer for Sage Bionetworks. You may not recognize John’s name or the name of his organization, but after today, you may want to commit it to memory. On March 9, 2014 Tim Cook introduced us to the idea of ResearchKit and then turned it over to Jeff Williams who described the new initiative and the five applications that were launched in collaboration with leading academic labs, advocacy groups, and research organizations. Sage Bionetworks is responsible for two of those apps, and John is responsible for leading design and development of many of the novel research tools and methods used by Sage. The following is an edited transcript of a conversation we had with John after the announcement.

Parkinson mPower App

This is Big

Think about what we’re doing in the context of it being open source. ResearchKit is going to be great for the QS community. They are perfectly suited to take advantage of ResearchKit. Let’s look at the five apps that were released today. They cover complex, but ubiquitious diseases. Plus the diseases are a great fit for the technology and sensor capabilities. But what is going to be amazing is when ResearchKit is released as open source. Anyone will have the ability to reskin these existing five apps to make novel research tools. Our Parkinson’s app might become a Huntington’s app. Our breast cancer app measures cognition really well, maybe it gets reskinned into a focused cognition tool. The app that focuses on diabetes could become a diet research app.

Large research institutions are going to jump all over this, but I’m more excited about the idea that small groups of people who have a bit of technical skill can work together to reskin the apps and turn them into longitudinal studies of 20 or 50 people. There may be five of those small groups, maybe 100. We can then work to stich those groups together and learn even more.

So the rare and chronic disease folks and the tech and QS community are involved. Now we come to everyone else. Everyone who isn’t suffering or presenting with a medical issue. We have to figure out ways to get them to participate.

We have a long tradition of creating incentives to participate. There is no reason we can’t create novel incentive methods to bring people into research. Maybe you receive a federal income tax credit. Maybe you’re allowed to enroll in long term care insurance at a reduced rate. There are a lot of ways to bring people in and that’s not touching the innate altruism of people and their curiosity. We’re already seeing this with the Precision Medicine Initiative. People want to take part, they want to engage.

Once we start getting that engagement, and we begin to see a diverse ecology of applications built on top of ResearchKit, then we’ll start to see success. It may take a bit of time. It won’t happen with this iteration. Maybe not even the second, but when we get enough devices, apps, participants, and improved interoperability between them all we’ll start to see the power of network effects.

You have to remember, there are no “killer apps” without network effects. Email wasn’t a big deal in the late 80′s because you couldn’t reach anyone outside your system. But then the web came, we connected the dots, the nodes, and then there it was, the power of the newtork.

I see QS and our current state of devices, apps, and tools being very similar to all those nerds typing away in the 80′s. They were okay with what they had because they could work with it. Then the net came and you have more control and more interefaces. That leads to the killer apps. That’s why we’re building this, for the third or fourth wave.

Share the Journey app

Everything we’re doing, the whole stack working together, is new. Consenting participants using well-designed and open-source Participant-Centered Consent toolkits. Giving participants direct access to their data. Securely hosting automatically de-identified data in the cloud using our Bridge Server. None of these have been done before at the same time.

We’re a non-profit, so we can be this icebreaker. We can take these risks and experiment and iterate and learn. No one asks us how we’re going to make money. We have a different bottom line — what is best for everyone involved? That outlook gives us the freedom to do this work.

I’m honored to have the opportunity to create a product that pushed my beliefs: participants-centric principles. If you’re a participant, then you decide. You’re in control. You can do cool stuff with the data. You can stop answering questions whenever you want. You can delete the app. We live in a world in which the politics of technology are dicteated by code and they often don’t share those ideas. To get this opportunity is amazing.

These moments don’t happen often. This is IBM going open source in the late 90s. It’s that big. I wasn’t sure I was going to be a piece of something that big. We’re trying to change culture.

In 1998 Lawrence Lessig proposed the pathetic dot theory in his book, Code and Other Laws of CyberspaceHe theorized that four forces control what we do: Law, Architecture, Social Norms, and the Market. He went futher and differentiated west coast law, which was quickly becoming dominated by software code and east coast law, what we normally think of when we think of laws.

I think up until now we’ve failed to properly take up that lever, to use software code, west coast law, to express something better. People often forget that building software means expressing an opinion. We created our apps, our tools, our systems to reflect our opinion, that participants should be at the center of research. And then we’re giving it away. That’s our position, and it’s better by default.

But, we’ve just started. The fun now is that we get to test it. I’ve always said that we can’t screw it up any worse than it already is. This isn’t the end. It’s not finished. We’re going to keep changing and learning.

John can be found online at del-fi.org and @wilbanks.

We invite you to share your data access stories, and this article with the #qsaccess hashtag and follow along on quantifiedself.com@quantifiedself and our Access Matters Medium publication.

Posted in QS Access, QS Access Conversations | Tagged , , , , , , , , , , | 1 Comment

Quantified Self and Apple’s ResearchKit


Apple’s announcement of ResearchKit is strong evidence that Quantified Self practices are emerging as a major influence on medical research and other forms of knowledge making.

Apple talked about how their new effort focused on opening up health research is designed to combat five main current issues:

  • Limited Participation
  • Small sample sizes limit our understanding of diseases
  • Reliance on subjective data
  • Infrequent data provide only snapshots through time
  • One-way communication from researcher to participant (and only at the end of the study, if at all)

Furthermore, the design of ResearchKit allows the  participant to decide how data is shared. Apple will not see the data. Participants are allowed to be involved in the data collection in real-time, using the data they’re collecting to understand and inform their own health improvement plans.

In light of today’s announcement we wanted to highlight some of our favorite and most powerful examples of taking the research process into one’s own hands, making their own knowledge through thoughtful data collection and reflection. We invite you watch what’s possible now, and imagine with us what could be accomplished tomorrow.

Mark Drangsholt: Deciphering My Brain Fog

Lindsay Meyer on Tracking Hearing Loss

Thomas Christiansen on Learning from 60,000 Observations

Nan Shellabarger: 26 Years of Weight Tracking

Rob Rothfarb on Tracking My Blood

Last year we gather a fantastic group of researchers, toolmakers, and science leadership at the 2014 Quantified Self Public Health Symposium to discuss how personal data can impact personal and public health. That meeting culminated in a great report that touches on many of the aspects discussed today regarding ResearchKit. We invite you to download, read, and share that report. For a more nuanced look into how ResearchKit may impact the research community, we’re highlighting four great talks from the the meeting.

Susannah Fox shares research from the Pew Internet and Life Project and describes the challenges ahead for promoting self-tracking.

Margaret McKenna explores the issues, challenges, and ideas large scale self-tracking applications have in mind when they consider working with the research community.

Jason Bobe talks about the lessons learned from involving research participants in the data ownership and discovery process.

Doug Kanter describes what he’s learned from tracking and visualizing his diabetes data.

If you’re interested in how ResearchKit will be affecting self-tracking, personal data, and access to information, research and knowledge making, then stay tuned to our Access Channel here on QuantifiedSelf.com and on Medium.

We are sure to have many great talks and sessions that focus on ResearchKit at our QS15 Conference and Actrivate Exposition. We invite you to join us.

We invite you to share your data access stories, and this article with the #qsaccess hashtag and follow along on quantifiedself.com and @quantifiedself.

Posted in QS Access | Tagged , , , , , , | 1 Comment

What is in My Gut?


This past fall we learned about a unique study, conducted at Stanford University, designed to contribute to the understanding of the human microbiome. This study also has a component not common to academic research — data is being returned to the participants. Intrigued, I contacted the principle investigator, Les Dethlefson, to learn more.

Ernesto: Tell me about the Dynamics of Human Microbiota study.

les_dethlefsen_at_the_benchLes Dethlefsen: Since I joined the Relman Lab at Stanford, I’ve been looking at the human gut microbiota, focusing on what affects it and how it changes over time. In our study, we are looking at three different perturbations, deliberate changes to the gut ecology, to see how the microbiota population is affected.

We are very interested in the patterns that emerge. In people who have very stable gut microbiota, does their microbiota remain that way when they undergo diet shifts, a colon cleanout, or an antibiotic? Or maybe people who have a stable gut microbiota most of the time are the ones who are most affected by something unusual such as taking antibiotics. We just don’t know enough to understand these patterns right now. So, we’re really looking for basic ecological information.

Ernesto: If you look at the popular press, it seems the microbiome is the new golden child of biological life sciences. We’re even seeing companies in Silicon Valley get involved with this kind of work.

Les: It is broader than that. It really is a worldwide interest on the parts of both the scientific community and the public. And unfortunately, we are probably going to see some overhype, just as we did with the Human Genome Project. But I do believe this is a very important area. I think there will be a lot of payoffs and health impacts from this research, although it’s not going to be everything.

The shift that, I think, would be good for us to make intellectually is to get rid of the “us vs. them” thinking, because we are symbiotic organisms.

We have evolved with a native gut microbiota, and native microbiota is pretty much everywhere. We have evolved together, so it’s fallacious — an artifact of our past ignorance — that we don’t think of our microbes as part of our physiology.

Ernesto: It seems like exploring the deep sea, an unknown world that we’re just starting to peek into.

Les: It’s along those lines. You’re not wrong about that. But unlike, let’s say, the deep waters surrounding an undersea hydrothermal vent, we already know a lot about human physiology. There are a lot of molecular details and genetic pathways that we already have worked out. The context is somewhat understood.

And now, we have a reasonable start on the initial research: What microbes are present, and where? What’s the range of what we think is the normal distribution? We certainly don’t know enough, because we only know about people in the developed world. However, this may not represent all of human diversity or a very natural state of the gut microbiota.

Candida albicans

Candida albicans

Ernesto: Let’s get back to your study. You are asking participants to send microbiome data in the form of fecal matter and urine to your lab. What are you doing with those samples?

Les: We ask participants to provide both stool and urine samples. With the stool sample, we apply four different methodologies to turn it into data. One is the very common 16S ribosomal RNA (16S rRNA) gene sequencing approach. It’s relatively standard and inexpensive. It acts like an ID card for microbial taxa — telling us approximately what strains are present and in what relative abundance. We have a lot of data like that already for comparison.

sequencerThe second approach we will be applying is metagenomic sequencing, wherein we will be sequencing a random selection of all the genomes of the microbial types that are present. We can’t take this to completion, even with the dropping cost of sequencing, especially because there are some very, very rare microbes that we barely even have the chance to see at all. But we can get a pretty good swathe of genetic sequence data from all the microbes.

The third approach is even more ambitious. It’s called metatranscriptomics. Genes can be carried by any critter, you and I included, but not expressed. Knowing which genes are turned on, and to what extent they’re turned on is a better measure of the biological activity that is actually happening. The metagenome is a measure of potential activities, what the bugs can do. The metatranscriptome shows what the microbes are actually doing. Metatranscriptomics is even more challenging than metagenomics partly because of the nature of messenger RNA (mRNA). It’s a highly unstable molecule. There are technical challenges, but we’re ambitious enough to try to collect information on gene expression.

The fourth approach is not based on gene sequences, but on chemical composition. Metabolomics is the name given to a number of these approaches that are not directed to a specific chemical. These are techniques that try to measure a broad swathe of chemicals present in the environment and their relative abundance. This is a technology that we, in the Relman Lab, know very little about. We’re collaborating with the Nicholson Lab in Imperial College in London, and they will be doing the metabolomic analyses on the stool samples. That may be even closer to where the rubber meets the road — knowing not just the gene expression but also the resulting chemical changes that are happening in the environment.

MicrobeHeatMapMetabolomics takes us to the other type of sample we’re collecting: the urine samples. We aren’t doing this because we have an interest in the urinary microbiome itself, but because, as the Nicholson Lab suggested, the urine provides a more complete, integrated picture of the co-metabolism between the human host and most of the gut microbiota. So while metabolomics for the stool samples would primarily measure the gut microbial activity and what they contribute to the host’s physiology, the urine provides a more integrated picture about how the host metabolism works in concert with the gut microbiota.

Ernesto: If a participant is going to be contributing all of that data, will they have access to it?

Les: As someone with similar interests, I certainly knew that a huge motivation for people to join the study would be the access to their own data. We offer monetary compensation, but for the amount of time that will be spent in contributing samples, it is probably trivial. We knew we would attract the curious, scientifically inclined, and practising scientists. Of course, they would want to see their data.

The Institutional Review Board (IRB) was quite open to us sharing information with the participants about their own microbiota. It probably helped that there’s publicity  about ways people can get this information. There is the American Gut project, offering an assessment of your microbiota for a donation, and uBiome, a private company offering the same kind of service.

I, or another staff member of the study, are going to share this microbiota data with each participant in a conference call. So in effect, I’m going to be a microbiota counselor. It’s nowhere near as high-stakes as sharing genome information. We don’t know enough to say, for example, that this microbiome is definitively healthy, or that it’s unhealthy, or what the exact risks of diseases are due to this particular composition. So we will be putting this information in context, and we will be available as interpreters of the scientific literature. We may be able to say that there is a statistical association between a particular microbial group that someone may have in their gut and some health-related outcome.


Microbiome sequence data

Ernesto: Will participants be getting a copy of their data as well?

Les: Yes, we will provide that. I have an open source mentality. Added to that is the fact that there are many practicing scientists signing up for the study and saying they want data, not just a PDF summary. I am happy to provide the data in as raw a format as people want. They can get the raw sequence information, a low-level summary (which is the result of the first pass of data processing), or the final summary. I have permission and full intention to share all the data derived from a person’s samples with that person.

Ernesto: Do you think we will see this happening more in the future?

Les: I think we will probably see more of it in the future. We’re moving in the direction of access to information. The open source movement has reached the health and medical realm from its origins in tech and computing. I think the participatory nature of access to data and scientific information is a good thing. It has started, and I don’t see any way of reversing the trend. I would hope that it becomes the norm that there is some appropriate level of sharing, that research participants have access to their data if they wish, and in a way that lets them interpret that data appropriately.

I believe that people have a right to that level of knowledge about their bodies, and if we, scientists, are generating that knowledge, there’s no reason not to share it with the individuals.

The Dynamics of Human Microbiota study is currenlty recruiting participants. If you’re interested in learning more about the ecosystem within read more about the study and check to see if you’re eligible to participate here.

We invite you to share your data access stories, and this article with the #qsaccess hashtag and follow along on quantifiedself.com and @quantifiedself.



Posted in QS Access, QS Access Conversations | Tagged , , , , , , , , , , | Leave a comment

QS Access: Precision, Patients, and Participation

This morning President Barack Obama announced a new Precision Medicine Initiative, a key $215 million piece of the proposed 2016 budget. Much has been written since last week’s State of the Union, when this initiative was first mentioned by President Obama. In brief, the initiative is an investment in new programs and funding initiatives at major government bodies that influence the current and future health of all Americans, including the National Institutes of Health (NIH), the Food and Drug Administration (FDA), and the Office of the National Coordinator for Health Information Technology (ONC). These programs will focus on developing “a new model of patient-powered research that promises to accelerate biomedical discoveries and provide clinicians with new tools, knowledge, and therapies to select which treatments will work best for which patients.”

There is a lot of information being circulated about this new initiative, and we’ve collected some links below, but we’d like to highlight something directly related to our interests in self-tracking data, personal data access, and new models of participatory research. In this morning’s announcement President Obama mentioned a long-term goal of creating a participatory research cohort comprised of 1 million volunteers who will be called upon to share personal medical record data, genetic samples, biological samples, and diet and lifestyle information. This is truly an ambitious goal and we are happy to see the President take care to mention the importance of including patients and the individuals who collect this data in the decision making and research process. For example, here is the description of this specific program from the NIH Precision Medicine Infographic


Here at QS Labs, we’re dedicated to helping create and grow a culture that enables everyone to generate personal meaning from their personal data. Sharing, participation, and exploring new models of discovery are a core themes we’re exploring as part of our QS Access work. We’ll be following this initiative as it moves from today’s announcement to tomorrow’s reality. Be sure to stay tuned to our QS Access Channel for more updates as we learn more.

Learn more about the Precision Medicine Initiative
NIH mini site describing the initiative
White House Blog: The Precision Medicine Initiative: Data-Driven Treatments as Unique as Your Own Body
FACT SHEET: President Obama’s Precision Medicine Initiative
A New Initiative on Precision Medicine by Francis Collins and Harold Varmus (New England Journal of Medicine).

Posted in QS Access | Tagged , , , , , , , , | Leave a comment

QS Access: Self-Report & Quantified Self in Health Research

As part of our new Access channel we’re going to highlight interesting stories, ideas, and research related to self-tracking data and data access issues and the role they take in personal and public health. We recently found this expert report, published in the International Journal of Obesity, that tackles issues with the data researchers rely on for understanding diet and physical activity behaviors, and ultimately concludes that the data is fundamentally flawed.

Researchers has known for a long time that relying on individuals to understand, recall, and accurately report what they eat and how much they exercise isn’t the best way to understand the realities of everyday life. Unfortunately for many years, this was the only way to track this information – interviews, surveys, and research measures. Only recently have tools, devices, and methods matured to a point where objective information can be captured and analyzed.

The authors of this article make the case that obesity and weight management fundamentally relies on getting these numbers right, and unfortunately most research hasn’t. Reading the background on self-report data and the call to action the authors make for developing and using more objective measures we can’t help but wonder about the role of commercial personal self-tracking tools. How can we, as a community of users, toolmakers, and researchers work together to open up access pathways so that the millions of people tacking pictures of their meals and uploading their step data can have a positive impact on personal and public health? This is an open question, one that we’re excited to be working on.

If you’re interested in these type of questions, or working on projects related to data access we invite you to get in touch and keep following along here with us.






Posted in QS Access | Tagged , , , , , , , | Leave a comment

The State of Wearables

In our work supporting users and makers of Quantified Self tools we pay close attention to how others talk about trends and markets. In the past year, the most-used catch all term for devices that help us track ourselves has been “wearables.” Now, it’s clear that wearables covers only a fraction of QS practices. Many of the ways people are using numbers, computing, and technology to learn about themselves do not involve wearing anything special. However, the term is useful to us in following relevant research. Below you’ll find links to last year’s best reporting on the wearables market, gathered into a single post for easy reference.

Pew Research Center (January 2013)

The most important work in this space remains the Tracking for Health report from the Pew Research Center, which found that 69% of adults track their health or the health of others, and that 21% of those who track use technology.
Link: QS Analysis of the Pew Research Center Tracking for Health

Forrester, January 2013
A report about the market for fitness wearables “like the Nike+ Fuelband and Jawbone UP” predicts that 8 million US online will be purchasing such devices.
Link: Fitness Wearables — Many Products, Few Customers

Nike, August 2013
Announces in a press release for their “Just Do it” campaign that they have over “18 million global” members of their Nike+ ecosystem.
Link: Nike Redefines “Just Do It” With New Campaign

CCS Insight, October 2013
Surveyed over 700 adults in both the UK and US. They found smart watch adoption was low with only 1.3% of adults (both countries) currently owning and using one and 1.5% no longer using (had owned). For “Wearable Fitness Trackers” they found 2.3% currently owned and used one and 1.2% no longer use it.
Link: User Survey: Wearables UK and US

Endeavor Partners, January 2014 (Part 1)
A survey of “thousands of Americans” completed in late 2013 found that 10% own an activity tracker. Activity trackers were most popular with younger adults (25–34 years) when compared to other age groups. They found that 50% of individuals who have owned an activity tracker no longer use it and one third stopped using it within six months.
Link: Inside Wearables

IDC, March 2014
“This IDC study presents the five-year forecast for the worldwide wearable computing devices market by product category. The worldwide wearable computing devices market (commonly referred to as “wearables”) will reach a total of 19.2 million units in 2014”
Link: Worldwide Wearable Computing Device 2014–2018 Forecast and Analysis

Nielsen, March 2014
A survey conducted in late 2013 of 3,956 adults found that 15% currently “use wearable tech—such as smart watches and fitness bands—in their daily lives.” Device ownership leaned heavily toward “fitness bands” with 61% of wearable technology users reporting ownership. This was followed by smart watches (45%), and mobile health devices (17%).
Link: Are Consumers Really Interested in Wearing Tech on their Sleeves?

Rock Health, June 2014
“While the activity tracker segment has about 1-2% U.S. penetration, wearables overall are expected to grow significantly”
Link: The Future of Biosensing Wearables

Endeavor Partners, July 2014 (Part 2)
As of June 2014, they found that the percentage of adult consumers that still wear and use their activity tracker has improved with 88% still wearing it after three months, 77% after 3–6 months, 66% after 6–13 months, and 65% after a year. They also found that majority of respondents (1,024 of 1,700 surveyed) reported obtaining their divide within the last six months
Inside Wearables – Part 2

PWC, October 2014
“21% of American adults already own a wearable device” They also found in their survey of 1,000 adults that 2% no longer use it, 2% wear it a few times per month, 7% wear it a few times a week, and 10% use it everyday.
Links: The Wearable FutureHealth Wearables: Early Days

Acquity Group, November 2014
A survey of 2,000 US consumers found that 13% plan to purchase as wearable fitness device with in the next year, and 33% within the next five years. Additionally, smart clothing is on slower trajectory with 3% planning to purchase in the next year and 14% in the next five years.
Link: The Internet of Things: The Future of Consumer Adoption

Gartner, November 2014
Gartner forecasts that worldwide shipments for “wearable electronic devices for fitness” will reach 68 million units in 2015, a slight decrease from the forecasts from 2014 and 2013 (70.2 and 73 million units, respectively). Additionally, according to Angela McIntyre, Gartner has found that “20 million online adults in the U.S. own and use a fitness wristband or other activity monitor and that 5.7% of online adults in the U.S. own and use a fitness wristband.”
Link: Forecast: Wearable Electronic Devices for Fitness, Worldwide, 2014

Berg Insight, December 2014
This is a market research report that states “fitness and activity trackers is the largest product category” and shipments are forecasted to reach 42 million units in 2019. Smart watches are predicted to reach 90 million units.
Link: Connected Wearables

Accenture, January 2015
Using a survey of 24,000 individuals across 24 countries Accenture found that 8% currently own a “Fitness Wearable”. Furthermore, they found that 12% plan to purchase in the next year, 17% in the next 1–3 years, and 11% in the next 2–5 years.
Link: Engaging the Digital Consumer in the New Connected World

Global Web Index, January 2015
In their Q3 2014 Device Summary report, GWI labeled wearable devices as “highly niche” after finding that 7% of US online adults own a “smart wristband” (Nike Fuelband, Jawbone Up, Adidas miCoach) and 9% own a smart watch.
Link: GWI Device Summary – Q3 2014

Rocket Fuel, January 2015
A survey of 1,262 US adult consumers conducted in December of 2014 found that 31% currently use a QS tool to track their health and fitness. This includes apps, devices, and websites. More specifically, 16% use a wearable device and 29% use a website or app not associated with a wearable device to track health and fitness.
Link: “Quantified Self” Digital Tools

Posted in Lab Notes | Tagged , , , , , | 1 Comment

What We Are Reading

We hope you enjoy this week’s list!

Big Data in the 1800s in surgical science: A social history of early large data set development in urologic surgery in Paris and Glasgow by Dennis J Mazur. An amazing and profoundly interesting research paper tracing the use of “large numbers” in medical science. Who knew that is all began with bladder stones!

Civil Rights, Big Data, and our Algorithmic Future by Aaron Rieke, David Robinson and Harlan Yu. A very thorough and thoughtful report on the role of data in civil and social rights issues. The report focuses on four areas: Financial Inclusion, Jobs, Criminal Justice, and Government Data Collection and Use.

Caution in the Age of the Quantified Self by J. Travis Smith. If you’ve been following the story of self-tracking, data privacy, and data sharing this article won’t be all that surprising. Still, I can’t help but read with fascination the reiteration of tracking fears, primarily a fear of higher insurance premiums.

Patient Access And Control: The Future Of Chronic Disease Management? by Dr. Kaveh Safavi. This article is focused on providing and improving access and control of medical records for patients, but it’s only a small mental leap to take the arguments here and apply them all our personal data. (Editors note: If you haven’t already, we invite you to take some time and read our report: Access Matters.)

Perspectives of Patients with Type 1 or Insulin-Treated Type 2 Diabetes on Self-Monitoring of Blood Glucose: A Qualitative Study by Johanna Hortensius, Marijke Kars, and Willem Wierenga, et al. Whether or not you have experience with diabetes you should spend some time reading about first hand experiences with self-monitoring. Enlightening and powerful insights within.

Building a Sleep Tracker for Your Dog Using Tessel and Twilio by Ricky Robinett. Okay, maybe not strictly a show&tell here, but this was too fun not to share. Please, if you try this report back to us!

Digging Into my Diet and Fitness Data with JMP by Shannon Conners, PhD. Shannon is a software development manager at JMP, a statical software company. In this post she describes her struggle with her weight and her experience with using a BodyMedia Fit to track her activity and diet for four years. Make sure to take some time to check out her amazing poster linked below!

The following two visualizations are part of Shannon Conners’ excellent poster detailing her analysis of data derived from almost four years of tracking (December 2010 through July 2014). The poster is just excellent and these two visualizations do not do it justice. Take some time to explore it in detail!



Tracking Energy use at home by reddit user mackstann.


“The colors on the calendar represent the weather, and the circles represent how much power was used that day. The three upper charts are real-time power usage charts, over three different time spans. I use a Raspberry Pi and an infrared sensor that is taped onto my electric meter. The code is on github but it’s not quite up to date (I work on it in bits and pieces as time permits I have kids).”

From the Forum
Help With Livestrong Data Export
Need Help Deciding Which Device
New to Fitness Tracking

Posted in What We're Reading | Tagged , , , , , , , , | Leave a comment