Monday, November 16, 2009

Glycaemic load and breast cancer

This struck me as fascinating when Dr Briffa posted it some time ago. I had this feeling that being skinny while eating as many sweets as you like might just be possible because you were running your metabolism in overdrive on glucose. That might just run the metabolism of a cancer cell in overdrive on glucose too. Being young and skinny does not appear to protect against breast cancer if you are sufficiently unlucky. No warnings necessarily given in terms of external markers of glucose dysregulation...

Of course the writing has been observable on this same epidemiological wall for some time.

Peter

34 comments:

Stan (Heretic) said...

The following study also showed massive correlation between carbohydrates and breast cancer. Based on their data, for every additional 1% of carbohydrate calories consumed above 52% (% calories of the total), the relative risk of breast cancer was increasing by about 12%!

-----------

"Carbohydrates and the Risk of Breast Cancer among Mexican Women",

Isabelle Romieu, Eduardo Lazcano-Ponce, Luisa Maria Sanchez-Zamorano, Walter Willett,
and Mauricio Hernandez-Avila,

Cancer Epidemiol Biomarkers Prev 2004 13: 1283-1289.

Gyan said...

Your epidemological reference says large PUFA intakes are protective against breat cancer.
Should that make us rethink on PUFA?

Peter said...

Hi Gyan,

Yes, and there is also the EPIC Norfolk epidemiological study which shows HbA1c to be inversely associated with PUFA intake. In Norfolk it's vegetable oil rather than fish oil for PUFAs. An equally interesting conundrum in an epidemiology paper... Epic didn't give raw data, it may have been corrected for calories, which makes it pretty meaning less unless you know what the calorie intake was too.

Peter

Peter said...

Ah, this one was calorie corrected. 5% of your 1000kcal slimming diet is not quite the same as 5% of 3000kcal... certainly in terms of grams per day total.

Peter

donny said...

The highest glycemic load group were the least likely to smoke, least likely to drink, ate the least protein, and got the most fiber from healthful fruits and grains.

It'd be nice if they'd given more detailed information on the differences in intake between highest and lowest quintiles of glycemic load in the skinny women.


http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2066187/
-----------------------------------
The liver produces plasma sex hormone–binding globulin (SHBG), which transports sex steroids and regulates their access to tissues. In overweight children and adults, low plasma SHBG levels are a biomarker of the metabolic syndrome and its associated pathologies. Here, we showed in transgenic mice and HepG2 hepatoblastoma cells that monosaccharides (glucose and fructose) reduce human SHBG production by hepatocytes.
---------------------------------

Since breast cancer is a sex hormone related cancer, I figured this might be a good time to pull out this study.

I guess in premenopausal women, if there's carb-derived palmitic acid in the liver, it probably means they ate some carbs.
In post-menopausal women, a larger fraction of the palmitic acid in the liver would probably come from disregulation of metabolism--
chronic metabolic syndrome-related elevation of glucose, maybe even the release of free fatty acids from fatty liver or visceral fat.

Steve Parker, M.D. said...

A high-glycemic index may increase the risk of breast cancer, but not by much (8% comparing the highest with lowest quantiles).

-Steve

Reference: Barclay, Alan W.; Petocz, Peter; McMillan-Price, Joanna; Flood, Victoria M.; Prvan, Tania; Mitchell, Paul; and Brand-Miller, Jennie C. Glycemic index, glycemic load, and chronic disease risk - a meta-analysis of observational studies [of mostly women]. American Journal of Clinical Nutrition, 87 (2008): 627-637.

mark said...

Peter, I have a thought about your inability to gain weight (muscle or fat) from eating crazy amounts of fat calories. So when you break down stored fat, you release a glycerol molecule from triglyceride.

Would it follow that if you were to gain fat mass, you'd need sufficient glycerol to bind the fats to transport them and then store them? Just enough carbs to stay out of ketosis some of the time is to provide some for energy utilization and keep the body's stock of glucose normal without need for gluconeogenesis. Shouldn't leave enough excess for you to store fat.

I don't suppose the liver would do gluconeogenesis, then convert that to glycerol just so you could store fat calories lol. So maybe ketosis and half ketosis work to keep your fat mass down through lack of glycerol.

Mark.

Peter said...

Donny, that's an interesting link, saved.

Steve, I always struggle with meta analysis because it is a mamoth task to go through and check what you think about the data going in to the analysis before you can believe the result coming out. But if the data are good I would agree that any effect which requires this number of people before you can pick it up must be trivial.

However I really struggle with this correct for calories issue. The role of carbohydrate as causal in heart disease and the protective effect of saturated fat suggested by these folks

http://www.ncbi.nlm.nih.gov/pubmed/15531663

especially here:

http://www.ajcn.org/cgi/content/full/80/5/1175/T1

These groups were not "corrected for calorie intake", they just all had about the same intake of calories. I have found (and unfortunately binned) essentially identical tables from observational studies showing the opposite, but these were "corrected for calories". People don't eat %s of calories, they eat grams of sugar. You lose this when the raw data get converted to percentages. How many studies in the meta analysis corrected for calories?

So I struggle with this approach and, given the likely quality of the data, picking up any trend seems quite impressive to me. If there is any causation of course!

Mark,

Don't forget that I don't calculate anything but I do weigh myself daily and I suspect I back off on calories if the scale cracks 65kg. I've only got Gunther's estimates of what I eat calorie wise since I stopped counting somewhere back in 2005!

Peter

lightcan said...

Peter,

can I ask you a question?
Does an LDL of 8.63 mean FH? Can it be always there (I mean the genetic difference, with moderately elevated TC and LDL) and only reacting big time to lowcarb high satfat diet? Maybe 1 in 100 subjects, just guessing, because there was a study regarding low carb diets in 2004 and two subjects dropped out, one because of huge increase in LDL and the other because of smaller increase.
And if it's FH I just have to eat like the Dutch in 1807 and I'll be fine? What other choice is there? It's not as if I will start eating seed oils or more carbs.
The doctor couldn't find a cause for it and puts it down to satfat.

Peter said...

Hi Lightcan,

Technically no, there are other marker such as tendon xanthomas and CVD in the clinical (non genetic) definition. And your value is close to that of Barry Groves, Jimmy Moore and myself. If we do all have heterozygous FH, yes, all that we can do is live like a dutchman in the 1800s. No margarine and just a little sugar...

Doesn't sound so bad to me!

Peter

lightcan said...

Hi Peter,

well the news are not good. My repeat tests show the LDL has gone up again in the last two months from 8.6 to 11.2 with HDL and trigs the same and the doctor says as you do that it is unlikely as in 2007 TC was 6.4 and in 2008 it was 7.1, and now 13.7. I must be beating all the records. The other thing is that the lp(a) is 101, which is bad right?

blogblog said...

I eat a high fat meat-based diet with no fruit or vegetables and <40g carbohydrate (mainly from cream). I only weigh myself every 4-6 weeks on a very accurate professional digital scale. My weight always stays between 70-71kg. However I am definitely getting leaner. This is despite no calories counting and no conscious attempt at weight control.

I have had several cats over the years. I noticed that their adult weight never varied by more than 100g when fed a quality zero carbohydrate diet. This was despite them eating as much food as they wanted. As the cats got older and less active they just ate less.

Dr. B G said...

Great post! Reminds me Peter when you discussed the BRCA gene is just a carb related gene. Funny eh?

The root cause of colon cancer is the same as breast cancer (is the same as prostate, is the same as other 'inflammatory' conditions like vascular disease):
--high glycemic index
--total carbs (of course inflammatory gluten) Dude, look at Fig 1
--FRUCTOSE

At the highest quintile of glycemic index, Giovannucchi showed that the multivar adjusted RR was 2.85 in the Women's Health Study for risk colon cancer.

Journal of the National Cancer Institute, Vol. 96, No. 3, February 4, 2004

I think I'll stay low carb, low fruit and HIGH FAT for now! Would prudently avoid all things that raise insulin and/or insulin resistance (e.g. Crestor and synthetic hormone contraceptives). *haaa* FYI The Kostapanos post still coming along... Too many distractions!

-G

Bris said...

Steve:
"A high-glycemic index may increase the risk of breast cancer, but not by much (8% comparing the highest with lowest quantiles)."

I expect that the carbohydrate intake between the highest and lowest quantiles is similar. It seems as though at least 90% of the population consume a diet that is 40-60% carbohydrate.

lightcan said...

Peter,
dr. B.G. suggested the apo E research done by KRAUSS. That's funny. Me and the apo E mouse.

Peter said...

Hi Lightcan,

You do hold the record at the moment (just) for LDL among people who have given me their numbers. You aren't quite up to the highest numbers cited by Kwasniewski as occasional findings, but you are very close.

The question is; what are you going to do about any of the numbers? Lp(a) of 100 is not what you would want as a youngster, though you might be glad of it in old age. But however worrying you might find it, what can you do? You can increase Lp(a) it by increasing either carbs or MUFA, but the only foods that I know of that decrease it are saturated fat and protein... The DELTA changes have been replicated at least once.

Because most of us do not have access to Lp(a) testing it's hard to say much more than that. Dr BG has a number of posts and is a niacin fan, but anyone eating high fat without the lab available simply wouldn't know they had elevated Lp(a).

I would be quite interested in your HbA1c... And I can't remember HDL and trig values.

BTW it's always worth considering whether you have an endocrine problem which might raise TC, but that's not needed for extreme responders to high fat diets.

Oh, and extra thought. CAC score to look at CV disease rather than surrogates. And if you do have any history of glucose dysregulation you may end up in the same situation as myself, needing to rescan to get a trend if you are not zero...

Peter

lightcan said...

Yes, Peter, I know, I read dr. B.G and she was kind to respond to my question, see above, indeed, not much to do about lp(a), besides eating more sat fat and less carbs which is anyway advisable so I don't oxidise my LDL even more. As there is no NMR in Dublin, that I know of, and they didn't want to do a CAC scan for a public patient, I don't know my real risk. I tried to look into the hormonal issues but I was sent back into my corner.
HBa1c stable at 5.2, HDL grew to 2.5, which is close to the top of the range.
A point I didn't talk about. I had severe periodontitis and consequently I had severe bone loss. There is a huge link to cardiovascular disease and it just makes me think that I must have plaque. So I will follow Track the plaque as much as I can, which means high fat and low carb. I can't reach their LDL goal though!
I was just hoping I can have some of the Christmas junk food. It seems not.

lightcan said...

Oh, I forgot Trig consistent for the last 14 months around 0.6-7.

Nigel Kinbrum BSc(Hons)Eng said...

lightcan, do you know your Vit K2 intake? K2 gets calcium out of your arteries & into your bones.

I've blogged about K2 as has Stephan Guyenet.

Cheers, Nige.

Kurt G. Harris MD said...

@lightcan

I'm glad to see you have gone to some reliable sources with these record setting lab values of yours. I have been pondering your dilemma before responding more at PaNu. I agree with Peter that FH is unlikely given your initial values. You need to make sure that clinical hypothyroidism has been excluded.

As far as Lp(a) this is not near the "risk factor" many believe it to be. My reading of the literature on lp (a) makes me differ a bit from Dr. Davis on this. Hecht has shown that it does not correlate at all with CAC scores. I can confirm it does not in my case. Despite my Lp (a) of 85, I have zero family history of early coronary disease, and my CAC score is 0 and my CIMT shows values more typical of a 35 year old than my age which is 48.

I also have a direct LDL of 194 and LDL-p of only 1100, with small dense and IDL both ZERO. So my direct LDL overestimates my true amount by over 80%. This is why it would be so helpful to have NMR done in your case. Dr. BG might say, and I might agree, that LDL-p is irrelevant if you have zero small dense particles like me. You can get it done in the states through a mail order lab for under $100. There has to be some private place you can get this done in the UK, I would think. Otherwise, go visit Peter in Glasgow, have him draw your blood and freeze it, and express mail it to me and I can run it for you at cost!

A Peter said, the main thing is what are you going to do about it? The only thing that lowers lp(a) is high sat fat diet ( or truckloads of niacin). The best way to limit the amount of ox LDL in your possibly high number of LDL particles is to avoid PUFAS and have big fluffy particles loaded with antioxidant cholesterol -eat a high sat fat diet.

As far as CAC, that is fairly insensitive to atherosclerosis in men under 40 and women under 50, If you are young CIMT is much less likely to give false reassurance. If it's positive, what are you going to do about it anyway?

It is my developing opinion that a high sat fat diet with low O-6 PUFAs, no wheat and limited fructose is already the maximum anyone can do to avoid atherosclerosis and once endocrine abnormalities are excluded, anxiety provoking lab tests should be left to self-experimenting bloggers

lightcan said...

I apologise for pestering everybody about it. (yes, I am a bit neurotic, high cortisol) It's important that I go to my next appointment with the resolution that I'm not going to change my diet, that I consider to be the best for me. I am very grateful to all, and feel very fortunate that I could ask somebody knowledgeable and open-minded.

Dr. Harris, that was really funny, the trials and tribulations of my blood samples.

I have been reading on Apo E, I think I might have the e4 allele.

Nigel, how are you?
I don't know what my K2 intake is. Cheese not so much, Emmental and goats, liver often enough, eggs every day, butter yes, and all meats. I was thinking about supplementing for my gums. I will have to get if from iherb.

A big hug for all

Nigel Kinbrum BSc(Hons)Eng said...
This comment has been removed by the author.
Nigel Kinbrum BSc(Hons)Eng said...

2nd attempt with correct source of K2
@lightcan

I'm doing much better thanks (details in Blog). I've just ordered 90 caps of Vitamin Research Products 15mg K2 from Seek Natural and I take 15mg/week as a maintenance dose.

I'm putting mum on 15mg/day (to try to raise her brain sulfatide level which hopefully will lower her brain ceramide level).

I increased my lumbar bone density from -2 to normal from 2003 to 2006 using 600mg/day Ca, 300mg/day Mg, 400iu/day D3 & 15mg/day K2 which, according to my Endocrinologist, was impossible.

Cheers, Nige.

Dr. B G said...

Hi Lightcan!

I should've asked your A1c but Peter beat me to it!

Hypothyroidism can raise glucoses and post-prandial insulin. Vitamin K2, D3 and n-3 fish oil can help tremendously with the insulin, thyroid and bone issues. One study in rats showed placebo lead to easy traumatic tooth removal, whereas those supplemented with n-3 had less tooth removal.

Influence of dietary n-3 polyunsaturated fatty acid on experimental tooth movement in rats.
Iwami-Morimoto Y, Yamaguchi K, Tanne K.
Angle Orthod. 1999 Aug;69(4):365-71.

Omega-3 fatty acid effect on alveolar bone loss in rats.
Kesavalu L, Vasudevan B, Raghu B, Browning E, Dawson D, Novak JM, Correll MC, Steffen MJ, Bhattacharya A, Fernandes G, Ebersole JL.
J Dent Res. 2006 Jul;85(7):648-52.



Actually I had periodontal disease with my 2nd child -- totally reversed on n-3 and vitamin D alone within 6-9mos. I got the OK from the dental hygienist to never come back for the extra cleaning/de-scaling/brutality that I was paying out of pocket for at $100 per pop!!

(now I spend it on all this fermented CLO, vitamin D etc!!)

The AA:EPA test (n-6, n-3 respectively) from Canada advocated by Sears is the best way to know if the appropriate balance has been achieved; he is REALLY good. He evaluates DGLA, GLA (n-6, a good one from borage, primrose and hemp(no, POT/THC doesn't count *haa ha*)) and other markers of inflammation.

A1c of > 4.0 % is kinda high in my opionion. DR. Bernstein the diabetes expert advises < 4.0 - 4.5 %. Again -- correctable with insulin and insulin sensitivity and thyroid improvements.

Correction! The official TrackYourPlaque program is LOW sat fat (8-10% per Davis at this time). I am an 'adjunct' to the program and we differ on this point. Just FYI -- though IMHO the only members who achieve regression on EBCT by my observation are the ones consuming a mod to high sat (doesn't matter what supplements they are on unless ultra high n-3, and doesn't matter how hypothyroid they are). I concur completely w/your thoughts and Kurt and Peter.

High sat fat, low carb (no carb sometimes or often) in order to get the IDL and sdLDL to 'ZERO' honey!! It works obviously.

-G

Peter said...

Hi Lightcan,

I'd concur with the comments here. I'd also suggest that HbA1c is still a blunt instrument and for those with long lifespan RBC the value will be higher at a given blood glucose than those with short lifespan RBCs, so 5.2% is OK provided you are not getting post prandial glucose levels in to vascular damage levels, general gleaning from what research focuses on this seems to suggest anything below 6.5mmol/l is safe. Back when I had the time to browse Dr B's forum I can't remember people getting HbA1c below 4% and anything below 5% was the bees knees, so you are not far off here. I've been looking for PUFA as a RBC lifespan shortening factor, as oxidation (glycoxidation) of the surface lipoproteins seems to be the signal for their destruction, but no hard data so far.

Where you ever vegetarian? I just ask on account of the gum disease, which suggests you have been over feeding your mouth bacteria, they like glucose and hydrogen... ie sugar, starch and fibre.

Peter

Dr. B G said...

Peter,

On Dr.Bernstein's forum, these were all diabetics (type 1 and 2), right? (and probably hypothyroid added) What about non-diabetics?

Several of my T2D achieve low 5s < 5.5% on low carb semiPaleo and some (or none) metformin. I have one after correcting Cushing's his BP reversed and the a1c decreased to low 5s (mabye once even < 5% but I to verify) off of meds though he has terrible PVD and a foot ulcer that won't go away (turns out he is TOTALLY hypothyroid though the TSH is like perfect 1.0-1.4, due to a meningioma that screwed probably the pituitary/ hypothalamus axis). BillyE, one of Dr.T's (nephropal) biggest supports (and freq heartscan commenter) has achieved I believe A1c 5.1% now off of ALL diabetic meds including insulin on high fat, low carb Paleo. Apparently the kidney condition is much reversed as well.

I'm not sure about the long-lived RBC stuff (?related to hypothyroid) but perhaps it is the opposite as Thallesemia -- where hgba1c is greatly UNDERestimated? I have a couple of these (?undiagnosed borderline thallesemia) and I have to track their fructosamine (which provides a 2-week BG average) and 1-hr postprandial glucoses. Interestingly, some don't have complications at all as one might expect. Others, horrible.

-G

lightcan said...

Peter,

I was some kind of lacto-ovo-vegetarian for a short while, can't remember for how long, in the 90s; I remember eating dried soy mince in my spag bol and margarine on bread. Even doing fruit juice fasts for a day. Whole grains and magic raw vegetables. Madness. I think I was eating chicken though. When I saw the dentist in Budapest in 1997 she kept asking if I was smoking.

Gyan said...

Surely it cant be just carbs that are problem. High-carb eating tropical countries have low BC rates.
Could it be High Fructose and/or high PUFA?
And GI Index is just proxy for sucrose?

Peter said...

Gyan, I certainly suspect that high carb intake is a surrogate for high fructose. Once the liver is "cooked" in situ then starches drive insulin and hyperglycaemia, but are probably not causal. Those people who can eat starches without either hyperglycaemia or inappropriate hyperinsulinaemia probably don't have metabolic syndrome.

The PUFA, apart from liver issues, have all sorts of tissue growth factor functions tied up with them and are probably acting in addition to excess insulin...

Peter

Gyan said...

Peter,
I eat a high-fat (relative to other Indians) mostly vegetarian diet (about 50% fat). Fat is ghee, coconut and some rapeseed oil.
I have prominent abdominal fat. But my fasting glucose and HbA1c are normal (4.8%). I did lost 3-4 kg last year by going to higher-fat diet and deleting PUFA oil.

How do I find out if I have inappropriate hyperinsulinaemia?
If I get fasting insulin measured, what is the appropriate value re: my diet?

Peter said...

Hi Gyan,

This is almost impossible and there is no database to compare it to anyway, what would be needed would be repeated insulin over the time course of a typical meal to compare to some sort of standard. I suppose the closest would be a GTT with insulin measurement, but this is not what is happening after a real meal. I don't know that has ever been looked at in detail.

You might consider surrogate markers like HDL:trig ration but even there you don't really know what is going on. That's one positive aspect to LC eating in general, even if insulin response is high per unit carb, total exposure will still be quite low compared to the exposure after a high carb diet. The absolute degree of hyperinsulinaemia must be quite small if total carbs is small. My favouring LC in general comes from our inability to be certain we can handle higher carb levels. An excellent HbA1c is excellent unless it is "bought" at the cost of a pancreas with muscles of steel...

Peter

Gyan said...

Well in my case, not many muscles, I am sedentary, only exercise I get from playing with my 4 yr old girl.
But maybe I am overdriving my pancreas?
I did gain 18 kgs in one year a few years back, then I switched to higher-fat and cutting PUFA, and the weight gain stopped.
Now I am trying to cut wheat partially. It is difficult culturally as it is unheard of in India.
Though people with colitis, gallstone problems etc are not uncommon but nobody imagines that cutting on wheat could be a solution. Even when I have told people, they say it is impossible for them. Without wheat, it is impossible to imagine a meal.

Peter said...

Well Gyan, that's how it is... Oddly enough we eat out at Indian restaurants a fair bit and just skip the nan (and rice) while choosing ghee based meat dishes... It's the mindset which is difficult!

Peter

Gyan said...

Ghee based meat dishes-- rarely found in India itself. The meat is mostly cooked with vegetable oils. I think thats why in India, the meat-eaters do not tend to have the best health. The centuarians esp tend to come from Jains, (no study but anecdotally) who follow lacto-vegetarian diet that is heavily ghee based and interestingly, they avoid too many vegetables-- their dietry laws forbid roots so avoid potatoes, carrots etc.