This week we want to give you some insight on how GrassrootsHealth collects data through D*action, why we collect the data we do, and how it is used to help move research into practice.

Why the long questionnaire?

GrassrootsHealth could provide a vitamin D test without asking you to fill out a questionnaire. It would be a public service, especially back in 2009 when we started and it was hard to get a doctor to prescribe a vitamin D test. But we founded GrassrootsHealth for more than testing. We founded GrassrootsHealth based on science, by first gathering together researchers to form a panel, then having those researchers come to a consensus that optimal vitamin D blood levels should be in the range of 40-60 ng/ml.  

By asking questions along with each test, we collect information that can be used to inform public health officials and conduct vitamin D research. By asking about health, supplementation, and exposure to sun or UV light - we can connect each participant's blood level with their lifestyle and health status.  

By asking about demographic and lifestyle information, health history and current conditions we track how vitamin D blood levels are associated with disease. We want to determine if an association between vitamin D blood level and a particular health condition exists, and if it does, how strong the association is or how it may change based on vitamin D status.  

What happens to the data that is recorded through the questionnaire?

All data is held in a secure database. We do not share this data with any other company or entity.  

Our customer service staff uses the contact information to send notifications about test results and when it is time for the next test.  

One staff member on our research team reviews questionnaire responses and test results to try and spot things out of the ordinary that may need clarification or correction. If there is a newly reported disease, they may contact that participant to request medical records.

All other research team members see only the questionnaire and test result data - they do not have access to the personal identification information.  They conduct data analysis using the de-identified data and put together summary data reports and graphs. We have published newsletter topics around this data such as how vitamin D relates to sex, age, or latitude (Your Data, Your Answers). GrassrootsHealth has also published many peer-reviewed papers - see sidebar.

Top 3 Areas of Interest

We could ask many more questions than we do on our questionnaire. It is great to have lots of data to compare with vitamin D levels, but we understand that there is a limit to how much time people are generally willing to spend filling out a questionnaire.

Most of the known research on vitamin D has been done in populations with low blood concentrations of 25(OH)D, with very low supplementation (< 2,000 IU/day), or with bulk supplementation, large doses given once a month. Our data set is unique in that many of our participants take over 4,000 IU/day and our average serum level is 44 ng/ml.

Area of Interest #1 - Dose Response

Is it really true that one person can supplement with 1000 IU/day and have a vitamin D level of 50 ng/ml, while another takes 10,000 IU/day and only has 40 ng/ml? What can we surmise by reported supplementation and sun exposure vs. the measured vitamin D level?

Above is one of the first graphs we published. It is a dose-response curve using data from 3667 participants that showed an average dose response, as well as the wide variation at each level of intake. The curve also showed that the response is not linear - it is steeper at lower levels and then flattens out at higher levels.

Recently, we used our dose-response data - your data - to respond to the IOM regarding the current RDA. We helped publish a paper that used a statistical analysis to show how much supplementation it would take to get 97.5% of the population above 20 ng/ml.

Area of Interest # 2 - Disease analysis

We have researchers on our panel of scientists who specialize in specific diseases or conditions, e.g. diabetes, cancer, or pregnancy. Our data can help them research relationships with vitamin D using diagnosis date, vitamin D level before diagnosis, and vitamin D level after diagnosis. Our data also includes other factors that might be important, such as gender, age, smoking status, exercise, latitude, weight, BMI, and occupation.

Overall we can explore which health conditions are associated with vitamin D status. We can also drill down by disease and find what serum level range is ideal for the prevention of that disease. From there we can compute how much disease would be prevented with serum levels in the range of 40-60 ng/ml.

Area of Interest #3 - Fluctuations of D-levels

Our database has also shown that many of our participants D-levels change throughout the year. By looking at the other data included in the questionnaire, we can try and correlate the vitamin D level change to other factors - sun exposure, UV light, exercise, etc.

When information from the questionnaire cannot explain observed differences, we may contact the participant to try to gather more information. For example, we know that a recent illness or injury can result in lower than usual levels, sometimes we see that it is higher, sometimes much higher. With our additional questioning, we have learned that some people will take an extra large dose for a few days if they feel a cold coming on, others have reported taking extra after an injury such as a sprained ankle. These extra large doses can result in a temporary spike in D levels.

We have also learned from several anecdotes that foods and supplements that may not contain any vitamin D can have an impact on how we absorb the D we get from other foods/supplements.

Why doesn't my doctor ask for this data when I get a test through my doctor (a lab)?

Your doctor already has a lot of the information - they have your current health history; they do generally ask if you have been smoking or drinking alcohol; they do take your height, weight, and blood pressure measurements.

The questions that differ are specifically related to vitamin D research - to gauge skin exposure we ask how much time is spent outdoors, clothing worn when outdoors, sunscreen usage, and use of indoor UV equipment; for vitamin D intake we also ask about brand and type of vitamin D supplement used in addition to the total amount.  

What good has my data done for vitamin D research?

Most vitamin D research has been done with low supplement levels and low serum levels. The GrassrootsHealth cohort is providing information on what happens at higher levels.

Please see the sidebar for some of the papers and newsletters that have been published from the D*action data. One of the first key findings was our dose-response curve mentioned earlier. Another paper challenged the belief that high intake of vitamin D, and higher serum levels, were associated with increased rates of kidney stones.

We have a paper under review right now that shows a lower incidence of diabetes with higher vitamin D levels. We are in the process of gathering data so that we can look at the relationship of vitamin D levels with post-menopausal breast cancer.