- Reaction score
- 907
here a great article from a respected poster of hairlosshelp about the expectation of propecia for Norwood 1 and 2 , and why sometimes it fails for them:
The "86% maintained or improved vs 14% worsened" figure we get from the Rossi 10-year finasteride study is not relative to baseline. I'd assumed it was for the longest time, as per similar studies. I was wrong.
It's relative to the 5-year data point, which makes sense, because the real purpose of the study was to see if finasteride "stops working" in the long-term.
If you really read the study until you finally really understand it (took me a few times, but I think I finally got it), you'll find it's even more remarkable than we thought.
Let's slap it around for a while, and see what shakes out:
If you look at Table 8 in Rossi, however, those numbers are relative to baseline, and you can easily calculate the 5 year and 10 year result relative to baseline from them. You can then use those number to make an apples-to-apples comparison with the 5 year Kaufmann study we all know and love ('cuz they use the same method).
Furthermore, because Rossi breaks down the figures by Androgenetic Alopecia group, you can use a little bit of logical reasoning (surprisingly harder than it sounds) to figure out the ten year results for the different Norwood levels.
I chose NW1/2s because they are harder to maintain than other Norwood levels:
Here are 10 year results relative to baseline (5 year result in parenthesis):
93% Maintained or Regrew (93%)
50% Regrew (38%)
43% Maintained (54%)
07% Worsened (7%)
A couple of things jump out at you. One is that the two five-year studies have very similar results, which indicates reliability and validity. Then you see that the ten year result is actually better than the five-year result.
A bit of a disappointment is the NW1/2's result. A full 25% of them were below baseline. However, if you dig a little deeper, this is not a result of being a NW1/2. It's a result of higher androgen levels at younger ages, as most NW1/2s tend to be younger.
Also, there seems to be a bottleneck at baseline at year 5. In the first 5 years 7% of men dropped below baseline, in the next 5 0% did. This supports a "leveling-off" model of finasteride long-term results as extrapolated from the Kaufmann study hair count results:
Note that I'm not saying men don't lose hair after 5 years on finasteride. What I am saying is that this study strongly suggests that after 5 years -- and after hair loss has stabilized in even the very worst responders -- the men losing hair will be outnumbered by the men still gaining hair. As a result, the entire group appears to be doing better.
Finally, a few interesting facts about the 16 guys in the NW1/2 group:
-2 guys continued to improve between years 5 and 10.
-Based on the 9 guys improved at the 5 year mark -- the remaining 7 that didn't see continued improvement maintained.
-That would mean 3 guys maintained a score of 0 (baseline) between years 5 and 10
-Which leaves the 4 of 16 guys worsened between years 5 and 10
-It follows that the overall ten year numbers from baseline (not from year five) for NW2 or less in this study are:
--9 men improved (56.25%) over baseline, with 2 of those men (12.5%) continuing to improve after year 5.
--3 men unchanged throughout the study (18.75%)
--The remaining 4 men got worse (25%), though it is not known when they started to get worse, and either 0 or 1 of those 4 continued to worsen between years 5 and 10.
-In this group we can calculate the overall improvement rate. It was 56.25% at 5 years, and 56.25% at ten years, so it was maintained in this group through year 10. 18.75% were at baseline for a total 75% maintenance rate.
Here the full link to the topic http://www.hairlosshelp.com/forums/messageview.cfm?catid=10&threadid=106061catid=10&threadid=106061
The "86% maintained or improved vs 14% worsened" figure we get from the Rossi 10-year finasteride study is not relative to baseline. I'd assumed it was for the longest time, as per similar studies. I was wrong.
It's relative to the 5-year data point, which makes sense, because the real purpose of the study was to see if finasteride "stops working" in the long-term.
If you really read the study until you finally really understand it (took me a few times, but I think I finally got it), you'll find it's even more remarkable than we thought.
Let's slap it around for a while, and see what shakes out:
If you look at Table 8 in Rossi, however, those numbers are relative to baseline, and you can easily calculate the 5 year and 10 year result relative to baseline from them. You can then use those number to make an apples-to-apples comparison with the 5 year Kaufmann study we all know and love ('cuz they use the same method).
Furthermore, because Rossi breaks down the figures by Androgenetic Alopecia group, you can use a little bit of logical reasoning (surprisingly harder than it sounds) to figure out the ten year results for the different Norwood levels.
I chose NW1/2s because they are harder to maintain than other Norwood levels:
Here are 10 year results relative to baseline (5 year result in parenthesis):
93% Maintained or Regrew (93%)
50% Regrew (38%)
43% Maintained (54%)
07% Worsened (7%)
A couple of things jump out at you. One is that the two five-year studies have very similar results, which indicates reliability and validity. Then you see that the ten year result is actually better than the five-year result.
A bit of a disappointment is the NW1/2's result. A full 25% of them were below baseline. However, if you dig a little deeper, this is not a result of being a NW1/2. It's a result of higher androgen levels at younger ages, as most NW1/2s tend to be younger.
Also, there seems to be a bottleneck at baseline at year 5. In the first 5 years 7% of men dropped below baseline, in the next 5 0% did. This supports a "leveling-off" model of finasteride long-term results as extrapolated from the Kaufmann study hair count results:
Note that I'm not saying men don't lose hair after 5 years on finasteride. What I am saying is that this study strongly suggests that after 5 years -- and after hair loss has stabilized in even the very worst responders -- the men losing hair will be outnumbered by the men still gaining hair. As a result, the entire group appears to be doing better.
Finally, a few interesting facts about the 16 guys in the NW1/2 group:
-2 guys continued to improve between years 5 and 10.
-Based on the 9 guys improved at the 5 year mark -- the remaining 7 that didn't see continued improvement maintained.
-That would mean 3 guys maintained a score of 0 (baseline) between years 5 and 10
-Which leaves the 4 of 16 guys worsened between years 5 and 10
-It follows that the overall ten year numbers from baseline (not from year five) for NW2 or less in this study are:
--9 men improved (56.25%) over baseline, with 2 of those men (12.5%) continuing to improve after year 5.
--3 men unchanged throughout the study (18.75%)
--The remaining 4 men got worse (25%), though it is not known when they started to get worse, and either 0 or 1 of those 4 continued to worsen between years 5 and 10.
-In this group we can calculate the overall improvement rate. It was 56.25% at 5 years, and 56.25% at ten years, so it was maintained in this group through year 10. 18.75% were at baseline for a total 75% maintenance rate.
Here the full link to the topic http://www.hairlosshelp.com/forums/messageview.cfm?catid=10&threadid=106061catid=10&threadid=106061