Hazard Ratios for Diabetes Complications

KK123

Well-Known Member
Messages
3,967
Type of diabetes
Type 1
Treatment type
Insulin

Oldvatr

Expert
Messages
8,470
Type of diabetes
Type 2
Treatment type
Tablets (oral)
I didn't realise that the Hazard Ratios were so bad even in the 'pre-diabetic' region of the HbA1C levels. Note that the HbA1C's are shown measured in % (i.e. old style).

https://twitter.com/DikemanDave/status/1274583863570755585

http://www.lesterhightower.com/diabetes/HazardRatios_202004_ADA_paper.html
Hazard ratio (HR) or Relative Risk (RR) is a way of inflating artificially the results of a statistics based study to make things seem better or worse. So if the actual number of events in a database associated with one variable was say 0.5% , then this can be magnified to give an HR of about 40. This was prevalent in drug trial reports published before 2006, when the statistical methodology was finally banned, and all studies since are supposed to follow new methodologies. We still see HR calculated, but it is supposed to be more representive. Unfortunately, because the raw data itself is now kept hidden or commercially confidential, there is no way any independent analysis can be done, and even peer reviews can be carried out without recourse to the original database. There are lies, damned lies, and statistics.
 

Mr_Pot

Well-Known Member
Messages
4,573
Type of diabetes
Type 2
Treatment type
Diet only
I didn't realise that the Hazard Ratios were so bad even in the 'pre-diabetic' region of the HbA1C levels. Note that the HbA1C's are shown measured in % (i.e. old style).

https://twitter.com/DikemanDave/status/1274583863570755585

http://www.lesterhightower.com/diabetes/HazardRatios_202004_ADA_paper.html
I haven't studied this in detail but I noticed that the hazard ratio is with respect to someone with an HbA1c of 4.6% which is 26.8. I imagine someone with that level ie no where near being diabetic has almost zero chance of having diabetic retinopathy for example. If the hazard ratio is 12.73 for someone with an HbA1c of 48 then that may be still a very low risk.
 

ianf0ster

Moderator
Staff Member
Messages
2,423
Type of diabetes
Treatment type
Diet only
Dislikes
exercise, phone calls
I agree with you about that way that the drug companies use both Relative Risk and Absolute Risk.
However I had been led to believe that once I got my HbA1C down into the 'pre-diabetic' range and kept my BG spikes down to no higher than 7.8 mmol. I would have no greater risk than a similar person without Diabetes.
 

Oldvatr

Expert
Messages
8,470
Type of diabetes
Type 2
Treatment type
Tablets (oral)
I agree with you about that way that the drug companies use both Relative Risk and Absolute Risk.
However I had been led to believe that once I got my HbA1C down into the 'pre-diabetic' range and kept my BG spikes down to no higher than 7.8 mmol. I would have no greater risk than a similar person without Diabetes.
We all seek the Holy Grail of Reversal or Cure, but sadly it seems to remain elusive to most of us. Although I control my T2D well, I know that a bad meal choice or a sneaky snack can totally disrupt my equilibrium. Even when i did get reversal for a short while, where I seemed to be able to eat normal food at will, i relapsed and no longer benefit. I am close to that sweet spot again, but cannot allow myself to stray too far from the LC pathway. At least I am not progressing down the slippery slope.