Toward predicting vitamin D levels without a blood test.

---

Many models have failed to predict vitamin D levels based on about 10 variables.

The problems appear to be due to:

1. Not enough variables in the model
2. Not enough subjects
3. Not powerful enough software/model
4. Large variance between vitamin D testers ( 5% to 10%)
5. Not record other health influences - emotional stress as a child, antibiotics as a child

VitaminDWiki is considering solving all of the above problems

• Might be able to predect vitamin D test results to be in one of the following ranges
  0-20 15-25 20-30 25-35 30-50 40-60 50-70 60-80 70-100 80–100+
  Goal of model: less than 5% chance of being in an adjacent category
  Goal of model: 10X less chance (0.5%) of being off by 2 categories
      Example: predicted 20-30 ng but was actually 30-50 ng
• Vast majortiy of the data samples will be in the range of 10-30 ng
• Age range: 18 ==> 100 – majority in the range of 30-80
• Data for training and testing the model might be crowd-sourced via the web
  People could fill in both the variables AND their vitamin D test results
  Rarely can we expect that all of the variables will be filled in
  Especially expect to lack the model of the tester

Variables/questions

Will need to quantize the details to train a model with the data
Might want to provide the user an indication of the importance of each question.

Age

Body

Weight, BMI, Muscular (compensates for BMI in-accuracy)
Skin color
Red hair (less likely to be out in the sun)

Have a disease/health problem which is strongly associated with low vitamin D

Trauma, Surgery
plasma exchange – reduces vitami D level by 50% in a single day
Inflammation – including depression -which consumes vitamin D
Have infection
Gut, Kidney, liver poor
Low stomach acid ==> less Magnesium
Fatty liver disease
Sore bones
No gall bladder
Poor sleep

Female

On the pill
Currently pregnant or lactating
Pregnancy during past 12 months
Twins during past 12 months
Excessive clothing for reliigious reasons

What is consumed

Vitamin D supplements : e.g. 400 IU, 4,000 IU (more than 3 months) D2 vs D3
Take Vitamin D3 supplements with evening meal or monsaturated fats
Magnesium supplements > 300 mg of Mg increases vitamin D levels about 30%, Mag in hard water
Probiotics
Low cholesterol diet
HFCS, pop
Omega-3
Vitamin A
Low cholesterol diet
Does not drink fortified milk
Smoking

Drugs increase/decrease response

statins

Environment/access to sun in the middle of the day

Urban/rural
Season
Latitude
Recent suncation
Outdoor worker
Work second shift
Work long hours
Altitude (30% more UV at the altitude of Colorado)
Shut-in
Air pollution – reduces UVB and likelihood that person will go outdoors
Use air conditioing (home, work, car)
Tanning bed use
Getting sunshine thru window > 3 hours a day for – UVA decreases vitamin D levels

Note: Vitamin D is not a constant Canadian vitamin D levels fell by 6% in 2 years (2009-2011)

 Download a Grassroots Health 2013 questionaire

 Download a Grassroots Health 2019 questionaire for Henry Lahore

Want some visibility as to the confidence of each prediction.

---

VitaminDWiki - Predict Vitamin D category contains

It is very difficult to predict the response to supplementation of Vitamin D, or additional sun/UV
There are a huge number of factors involved.
 
This page also has studies predicting deficiency without Vitamin D tests

Examples of 82 studies that Predict Vitamin D levels

• Predicted Vitamin D levels for health young women had 95% accuracy using neural network (paywall) – July 2024
• Vitamin D deficiency predicted with 91% accuracy ( AI, age, paywall) - April 2024
• Predictors of low vitamin D: race, age, and BMI (confirmed now by Machine Learning) – Feb 2024
• Low Vitamin D screening with just 5 questions (for less than 12 ng) – June 2022
• Top 10 signs of Vitamin D Deficiency (9 minute Video) - Oct 2021
• Estimate Vitamin D levels based on questionnaires (12 studies) – July 2020
• Is a senior Vitamin D insufficient - a 2 minute questionnaire is 85 percent accurate – Nov 2019
• Simple Vitamin D deficiency scoring system – Feb 2016
• Toward predicting vitamin D levels without a blood test. by VitaminDWiki
• Excellent prediction of very low vitamin D in elderly from just 16 questions (analyzed by ML) – June 2017
• Quick, free, self test of vitamin D deficiency 90% chance <20 ng