Why does dairy fat, and perhaps other similar fats like tallow and coconut, seem to prevent diabetes?
A broken omega 6:3 ratio becomes more likely with higher PUFA intakes. There is something about having a low PUFA intake that preserves the balance, even at relatively low omega 3 intakes.
We can see this in the recent fatty liver study comparing olive oil with canola oil and soy/safflower oil (control). For 6 months 20g of oil per day was used to cook food; this is not much (and it seems likely to me that many participants would have used more than they were directed to, if only to increase the palatability of their meals). There was no change in fatty liver and insulin resistance scores in those using soy/safflower oil, which is presumably what all subjects cooked with before.
The pre- and post-intervention difference in liver span was significant only in the olive (1.14 ± 2 cm; P 0.05) and canola (0.66 ± 0.33 cm; P 0.05) oil groups. In the olive and canola oil groups, post-intervention grading of fatty liver was reduced significantly (grade I, from 73.3% to 23.3% and from 60.5% to 20%, respectively [P 0.01]; grade II, from 20% to 10% and from 33.4% to 3.3%, respectively [P 0.01]; and grade III, from 6.7% to none and from 6.1% to none, respectively). In contrast, in the control oil group no significant change was observed.
So canola oil and olive oil were about equally good for reversing steatosis; this might be an expected effect of supplying fats with an omega 6:3 ratio of 2:1 for six months. But when it came to glucose and insulin, there was a marked difference:
In a comparison of olive and canola oil, a significant decrease in fasting insulin level, HOMA-IR, HOMA-βCF, and DI (P 0.001) was observed in the olive oil group.
In fact, fasting insulin and blood glucose were normalised in the olive oil group, but not in the canola oil group. With regard to these measures of glycemic control, a 50% lower intake of linoleic acid (with substitution of MUFA from oleic acid) produced more benefit than a 20-fold increase in alpha linolenic acid.
Here we have a paper that compares the effect of LA restriction (from 8%E to 4%E) with the effect of DHA in immune-deficient mice bearing human breast cancer cells;
Tumor prostaglandin E2 concentrations were reduced by feeding the lower LA level; further dose-dependent decreases occurred in the DHA dietary groups and were accompanied by reduced levels of 12- and 15-hydroxyeicosatetraenoic acids.
According to Raheja et al. (1993) "prostaglandin E2 is a potent inhibitor of first-phase insulin release, whereas an arachidonic acid lipoxygenase product, possibly 12-hydroxyeicosatetraenoic acids (12-HETE) sustains increased second-phase insulin release". A pattern also known as insulin resistance, or if sufficiently elevated, NIDMM or type 2 diabetes. These elevated prostaglandins are also seen in type 1 diabetics.
And, what do you know, ghee reduces PGE2 in Wistar rats:
Ghee, the anhydrous milk fat, is one of the most important sources of dietary fat in India. Male Wistar rats were fed diets containing 2.5, 5.0 and 10 wt% ghee for a period of 8 weeks. The diets were made isocaloric with groundnut oil. The results showed that serum thromboxane levels decreased by 27-35%, and 6-keto-prostaglandin F1alpha by 23-37% when ghee was incorporated at level of 10% in the diet. Prostaglandin E2 levels in serum and secretion of leukotrienes B4, C4 and D4 by peritoneal macrophages activated with calcium ionophore decreased when increased amounts of ghee from 2.5 to 10% were included in the diet. Arachidonic acid levels in macrophage phospholipids decreased when incremental amounts of ghee were fed to rats. These studies indicate that ghee in the diet not only lowers the prostaglandin levels in serum but also decreases the secretion of leukotrienes by macrophages.
(I haven't seen fulltext for that, but control, groundnut oil, is around 30% LA, and 10 wt% will be more than 10%E).
With regard to ALA, this epidemiological paper on prostate cancer, while perhaps irrelevant, has an interesting line:ALA intake was unrelated to the risk of total prostate cancer. In contrast, the multivariate relative risks (RRs) of advanced prostate cancer from comparisons of extreme quintiles of ALA from nonanimal sources and ALA from meat and dairy sources were 2.02 (95% CI: 1.35, 3.03) and 1.53 (0.88, 2.66), respectively. The multivariate RR of advanced prostate cancer from a comparison of extreme quintiles of the ratio of LA to ALA was 0.62 (0.45, 0.86).
Do you have any idea how much dairy fat it takes to get into a high quintile for ALA? Anyway, just another epidemiological paper where animal fats come out safer than their vegetable equivalents. One of the ones you don't hear about.
As I mentioned previously here, in New Zealand per capita weekly butter consumption at the beginning of the Second World War was 415 grams. It is now 112 grams, which is half of the reduced 1940s wartime ration. Not much Type 2 diabetes in New Zealand prior to the Second World War. Not much consumption of heart-healthy oils either, but plenty of consumption of sugar and white flour.
The second hit: In children and young individuals, a high intake of n-6 PUFA is correlated with fasting hyperinsulinaemia, and dietary supplementation with n-3 PUFA leads to an improved lipid profile but not insulin sensitivity. In adults, high-carbohydrate meal consumption was reported to cause hyperinsulinaemia, postprandial hyperglycaemia and hypertriacylglycerolaemia. (Misra, A. 2009).
Take a child, and raise them on this high-LA, vegetable oil diet (because saturated fat and high cholesterol, don't you know, cause heart disease in toddlers). By the time they reach adulthood, they'e primed for the second hit:
That's grains, by the way, not sugar, not HFCS.
Dairy fat intake is associated with glucose tolerance, hepatic and systemic insulin sensitivity, and liver fat but not β-cell function in humans.
A broken omega 6:3 ratio becomes more likely with higher PUFA intakes. There is something about having a low PUFA intake that preserves the balance, even at relatively low omega 3 intakes.
We can see this in the recent fatty liver study comparing olive oil with canola oil and soy/safflower oil (control). For 6 months 20g of oil per day was used to cook food; this is not much (and it seems likely to me that many participants would have used more than they were directed to, if only to increase the palatability of their meals). There was no change in fatty liver and insulin resistance scores in those using soy/safflower oil, which is presumably what all subjects cooked with before.
The pre- and post-intervention difference in liver span was significant only in the olive (1.14 ± 2 cm; P 0.05) and canola (0.66 ± 0.33 cm; P 0.05) oil groups. In the olive and canola oil groups, post-intervention grading of fatty liver was reduced significantly (grade I, from 73.3% to 23.3% and from 60.5% to 20%, respectively [P 0.01]; grade II, from 20% to 10% and from 33.4% to 3.3%, respectively [P 0.01]; and grade III, from 6.7% to none and from 6.1% to none, respectively). In contrast, in the control oil group no significant change was observed.
So canola oil and olive oil were about equally good for reversing steatosis; this might be an expected effect of supplying fats with an omega 6:3 ratio of 2:1 for six months. But when it came to glucose and insulin, there was a marked difference:
In a comparison of olive and canola oil, a significant decrease in fasting insulin level, HOMA-IR, HOMA-βCF, and DI (P 0.001) was observed in the olive oil group.
In fact, fasting insulin and blood glucose were normalised in the olive oil group, but not in the canola oil group. With regard to these measures of glycemic control, a 50% lower intake of linoleic acid (with substitution of MUFA from oleic acid) produced more benefit than a 20-fold increase in alpha linolenic acid.
Here we have a paper that compares the effect of LA restriction (from 8%E to 4%E) with the effect of DHA in immune-deficient mice bearing human breast cancer cells;
Tumor prostaglandin E2 concentrations were reduced by feeding the lower LA level; further dose-dependent decreases occurred in the DHA dietary groups and were accompanied by reduced levels of 12- and 15-hydroxyeicosatetraenoic acids.
According to Raheja et al. (1993) "prostaglandin E2 is a potent inhibitor of first-phase insulin release, whereas an arachidonic acid lipoxygenase product, possibly 12-hydroxyeicosatetraenoic acids (12-HETE) sustains increased second-phase insulin release". A pattern also known as insulin resistance, or if sufficiently elevated, NIDMM or type 2 diabetes. These elevated prostaglandins are also seen in type 1 diabetics.
And, what do you know, ghee reduces PGE2 in Wistar rats:
Ghee, the anhydrous milk fat, is one of the most important sources of dietary fat in India. Male Wistar rats were fed diets containing 2.5, 5.0 and 10 wt% ghee for a period of 8 weeks. The diets were made isocaloric with groundnut oil. The results showed that serum thromboxane levels decreased by 27-35%, and 6-keto-prostaglandin F1alpha by 23-37% when ghee was incorporated at level of 10% in the diet. Prostaglandin E2 levels in serum and secretion of leukotrienes B4, C4 and D4 by peritoneal macrophages activated with calcium ionophore decreased when increased amounts of ghee from 2.5 to 10% were included in the diet. Arachidonic acid levels in macrophage phospholipids decreased when incremental amounts of ghee were fed to rats. These studies indicate that ghee in the diet not only lowers the prostaglandin levels in serum but also decreases the secretion of leukotrienes by macrophages.
(I haven't seen fulltext for that, but control, groundnut oil, is around 30% LA, and 10 wt% will be more than 10%E).
With regard to ALA, this epidemiological paper on prostate cancer, while perhaps irrelevant, has an interesting line:ALA intake was unrelated to the risk of total prostate cancer. In contrast, the multivariate relative risks (RRs) of advanced prostate cancer from comparisons of extreme quintiles of ALA from nonanimal sources and ALA from meat and dairy sources were 2.02 (95% CI: 1.35, 3.03) and 1.53 (0.88, 2.66), respectively. The multivariate RR of advanced prostate cancer from a comparison of extreme quintiles of the ratio of LA to ALA was 0.62 (0.45, 0.86).
Do you have any idea how much dairy fat it takes to get into a high quintile for ALA? Anyway, just another epidemiological paper where animal fats come out safer than their vegetable equivalents. One of the ones you don't hear about.
As I mentioned previously here, in New Zealand per capita weekly butter consumption at the beginning of the Second World War was 415 grams. It is now 112 grams, which is half of the reduced 1940s wartime ration. Not much Type 2 diabetes in New Zealand prior to the Second World War. Not much consumption of heart-healthy oils either, but plenty of consumption of sugar and white flour.
The second hit: In children and young individuals, a high intake of n-6 PUFA is correlated with fasting hyperinsulinaemia, and dietary supplementation with n-3 PUFA leads to an improved lipid profile but not insulin sensitivity. In adults, high-carbohydrate meal consumption was reported to cause hyperinsulinaemia, postprandial hyperglycaemia and hypertriacylglycerolaemia. (Misra, A. 2009).
Take a child, and raise them on this high-LA, vegetable oil diet (because saturated fat and high cholesterol, don't you know, cause heart disease in toddlers). By the time they reach adulthood, they'e primed for the second hit:
Refined grain consumption and the metabolic syndrome in urban Asian Indians (Chennai Urban Rural Epidemiology Study 57).
Compared with participants in the bottom quartile, participants who were in the highest quartile of refined grain intake were significantly more likely to have the metabolic syndrome (odds ratio, 7.83; 95% confidence interval, 4.72-12.99). Higher intake of refined grains was associated with insulin resistance and the metabolic syndrome in this population of Asian Indians who habitually consume high-carbohydrate diets.That's grains, by the way, not sugar, not HFCS.