It's time to replace the Friedewald Method for calculating LDL

[librarising]

Yesterday I managed to get my LDL down from 3.6 to 3.1 in next to no time. Since LDL, the so-called 'bad' cholesterol, is one of the figures your doctor may use to frighten you into statins, you may want to see if you can get yours down as easily.
LDL can be directly measured, but since this is expensive it is usually estimated. It is done (mostly) using the Friedewald Method where LDL

in UK (mmol/L) = TC - HDL - (trigs/2.19)
in US (mg/dl) = TC - HDL - (trigs/5)

It is known that at large levels of trigs the formula becomes unreliable. It has also been found that at low levels of trigs it is also unreliable.
This 2012 study looked at various methods for more accurately estimating LDL on 10,664 fasted individuals. The table below shows how the proposed formula (LDL = [TC - HDL] x 0.75) outperforms all other methods in 11 out of 12 sub-sections. Only where trigs exceed 2.11 mmol/L does the Teerakanchana formula fractionally beat the proposed new formula into second place. However the new formula clearly outperforms the Friedewald.
Did you spot the defining feature of this new formula? It ignores trigs. Since trigs rise after eating/drinking, their inclusion would affect any LDL calculation. Compare my results from 5/17 and 10/17

TC 5.2 / 5.0
HDL 1.1 / 1.25
Trigs 1.1 / 1.7 (the latter non-fasting as test unexpected)
LDL 3.6 / NOT GIVEN

The new formula brings my 5/17 LDL from 3.6 to 3.1.
If I assume fasting trigs of 1.1 the 10/17 test would give me an LDL of 2.98. Under the new formula it would be 2.8.

Check your own figures. You may be in for a nice surprise.
Geoff


http://journals.sagepub.com/doi/full/10.1258/acb.2012.011259

http://journals.sagepub.com/na101/h...ges/large/10.1258_acb.2012.011259-table1.jpeg

 

[Bluetit1802]

Tried it. First I checked the LDL using the Friedewald formula and it tallies exactly with the figure on my lab report. Then I used to new formula and it dropped from 3.6 to 3.0.

You say that low levels of trigs can skew the calculation using the existing Friedewald method. Do you know how low?

 

[librarising]

Tried it. First I checked the LDL using the Friedewald formula and it tallies exactly with the figure on my lab report. Then I used to new formula and it dropped from 3.6 to 3.0.

You say that low levels of trigs can skew the calculation using the existing Friedewald method. Do you know how low?

Dr Michael Eades :
"I’ve always thought the same held true for triglycerides under 100 mg/dl, which would apply to almost everyone who sticks to a low-carb diet for any length of time. Triglyceride levels of 40-90 mg/dl are not uncommon, and are, in fact, typical. When Friedewald did his work, the triglyceride levels were mainly up in the 150 – 250 mg/dl range, and in this range his equations match pretty well to directly measured LDL levels, but all bets are off with triglycerides above 400 mg/dl and, I suspect, triglyceride levels below 100 mg/dl. MD and I did find this ourselves in a few patients that we did direct LDL measurements on in our practice."
https://proteinpower.com/drmike/200...al&utm_source=twitter.com&utm_campaign=buffer
Geoff

 

[librarising]

Tried it. First I checked the LDL using the Friedewald formula and it tallies exactly with the figure on my lab report. Then I used to new formula and it dropped from 3.6 to 3.0.

You say that low levels of trigs can skew the calculation using the existing Friedewald method. Do you know how low?

100 mg/dl equates to 1.1 mmol/L

 

[Bluetit1802]

100 mg/dl equates to 1.1 mmol/L

Thanks, just did the conversion and got it to 1.129mmol/l so we agree. There is a cholesterol converter here for anyone interested http://www.onlineconversion.com/cholesterol.htm )

As mine have never been above that figure, even at diagnosis, and have been well under 1 for the last 4 years, it may be my LDL has been miscalculated completely forever. I bet I can't convince my GP of this.

 

[ringi]

Provided triglyceride are low and HDL is high, I don't care what LDL is. (Likewise for total)

 

[Art Of Flowers]

My calculated LDL went down from 3.1 down to 2.775 using the new calc.

 

[librarising]

Provided triglyceride are low and HDL is high, I don't care what LDL is. (Likewise for total)

That's great for the informed. Not so great for the misinformed or the ill-informed.
Like you, I don't care about a stand-alone TC/LDL figure.I care about as-true-as-possible figures.It's the libra in me
Geoff

 

[Bluetit1802]

Am I being thick, but if this new formula became the standard, the way they measure the total would have to change??
The total is needed in the new formula so needs to be correct.

Currently the total is HDL+LDL+ 46% of trigs (in UK measurements) so if the LDL drops, the 46% of trigs would have to be changed? Can a scientist explain this to me?

 

[ringi]

It is easy to measure the total. The formula is used in reverse at present to find LDL as its the hardest to measure.

 

[librarising]

Am I being thick, but if this new formula became the standard, the way they measure the total would have to change??
The total is needed in the new formula so needs to be correct.

Currently the total is HDL+LDL+ 46% of trigs (in UK measurements) so if the LDL drops, the 46% of trigs would have to be changed? Can a scientist explain this to me?

TC, HDL, and trigs are all measured. LDL under whatever formula will always remain an estimate, so you can't invoke an unknown and change a total.
Trigs are used in calculating LDL because they're a surrogate for VLDL, which isn't so easily measured.Neither are trigs a fixed percentage of TC. The 46% was felt by Friedewald to be a reasonable fit for estimating LDL. The proposed formula would reject 46%, and in fact rejects trigs from its formula.
As Dr Eades said, when Friedewald was introduced, trigs levels were generally higher, and could more accurately reflect LDL levels. That is no longer always true.

If you play around with your figures, you should be able to work out what that new % would need to be to achieve the revised LDL level. But that's useless information.
Geoff

 

[Bluetit1802]

Thank you both.

 

[pete254]

Yesterday I managed to get my LDL down from 3.6 to 3.1 in next to no time. Since LDL, the so-called 'bad' cholesterol, is one of the figures your doctor may use to frighten you into statins, you may want to see if you can get yours down as easily.
LDL can be directly measured, but since this is expensive it is usually estimated. It is done (mostly) using the Friedewald Method where LDL

in UK (mmol/L) = TC - HDL - (trigs/2.19)
in US (mg/dl) = TC - HDL - (trigs/5)

It is known that at large levels of trigs the formula becomes unreliable. It has also been found that at low levels of trigs it is also unreliable.
This 2012 study looked at various methods for more accurately estimating LDL on 10,664 fasted individuals. The table below shows how the proposed formula (LDL = [TC - HDL] x 0.75) outperforms all other methods in 11 out of 12 sub-sections. Only where trigs exceed 2.11 mmol/L does the Teerakanchana formula fractionally beat the proposed new formula into second place. However the new formula clearly outperforms the Friedewald.
Did you spot the defining feature of this new formula? It ignores trigs. Since trigs rise after eating/drinking, their inclusion would affect any LDL calculation. Compare my results from 5/17 and 10/17

TC 5.2 / 5.0
HDL 1.1 / 1.25
Trigs 1.1 / 1.7 (the latter non-fasting as test unexpected)
LDL 3.6 / NOT GIVEN

The new formula brings my 5/17 LDL from 3.6 to 3.1.
If I assume fasting trigs of 1.1 the 10/17 test would give me an LDL of 2.98. Under the new formula it would be 2.8.

Check your own figures. You may be in for a nice surprise.
Geoff


http://journals.sagepub.com/doi/full/10.1258/acb.2012.011259

http://journals.sagepub.com/na101/h...ges/large/10.1258_acb.2012.011259-table1.jpeg

Very interesting Libra.
I have just calculated my LDL over the last 4 years (5 tests) using both formulae - (thank god for spreadsheets!).
The results are quite interesting.
The first 4 LDL levels were exactly the same using the new formula. (trigs were between 1 and 1.3 in these tests).
The last test my trigs were 0.8 and my LDL using the new calculation LDL dropped from 1.7 to 1.6.
This seems to confirm that using the new formula when trigs are below 1 mmol/L will result in a lower calculated LDL.
Obviously the same is happening at higher trig levels from your results.
However within a certain range (I'm guessing between 1 and 2 mmol/L) both calculations will give the same result.
Pete

 

[sally and james]

It's worth reading Zoe Harcombe on this subject, http://www.zoeharcombe.com/the-knowledge/we-have-got-cholesterol-completely-wrong/ See, in particular, paragraph (iv), where the key point for this discussion is that there are four variables (Total, LDL, HDL and trigs), but only two are measured. As Zoe says, any mathematician will tell you that one equation with two unknows and no supporting equation, isn't a fat lot of use.
Sally