23and me gives the following statistics for APO E4. Like you said, although about 20% of people have 1 E4, 2 is much less common.
APOE variant frequencies in Europeans
Genotype Frequency
- ε2/ε2 0.4%
- ε2/ε3 13.7%
- ε3/ε3 63.9%
- ε2/ε4 1.3%
- ε3/ε4 19.0%
- ε4/ε4 1.7%
Data from Lahoz et al (2001).
They also say that
while there is evidence that the APOE ε4 variant may associate with higher LDL cholesterol levels, consistent associations have not been demonstrated for coronary heart disease or response to statins.
The connection to LDL is complicated. The biggest effect is seen in people with 2 ApoE4s.
This study suggests that it only makes a difference if you drink alcohol, for example.
This paper, shows a table with the different values of lipid panels associated with the different Apo E combinations. They show, for example that a typical someone with 2 E4's had LDL of 157, a 3 and a 4 with 143, a 4 and a 2 (which was the lowest!) with 101.
Here is a paper that acknowledges some of the complexity of the problem:
Interestingly, even though both apoE phenotype and diet modify LDL size, studies showing the interaction between these 2 factors are scarce or discrepant. Thus, a higher saturated fat intake was associated with smaller LDL particles in apoE2 subjects, and larger LDL particles in apoE4 subjects (31). However, another study showed that when subjects changed from a high- to a low-fat diet, there was a shift from large, buoyant, cholesterol-rich particles, to smaller, denser LDL particles, with progressively greater reductions in levels of larger LDL from apoE3/2 to apoE3/3 to apoE3/4 (32).
These two points strongly suggests that people with Apo E4 should not lower their fat / saturated fat intake!
The rest of that study is even more complex, in part because they compare only two conditions (which in my opinion aren't very comparable -- moderately high CHO with moderately high MUFA), and it's really unclear what conclusions to draw. The resulting discussion section is interesting to read, if only to show how little is understood. The authors do conclude something related to your suggestion: that there are many factors which make optimal diet individual.
My opinion is that LDL doesn't really matter very much. It has been shown that a diet with plenty of saturated fat often increases LDL, but it also increases the particle size. So it really isn't clear that high LDL in this scenario is dangerous. In fact, if the diet is low in carbs, the larger particle size, combined with the lowered triglycerides, and raised HDL, suggests to me that the perceived risk of higher LDL is more than compensated for. The problem in answering your question is that most studies that compare the cholesterol outcomes of people based on their Apo E variant and fat intake do not have a low carbohydrate intake in mind, and don't test the conditions that would make that part of the picture come into focus.
