I wouldn't worry about the Mensink study (reference 12). The same authors updated their study in 2003 analysing 60 controlled trials, and give a more favourable view of saturated fat. They looked at the ratio of total to HDL cholesterol, which they say is a more specific (better) marker of coronary artery disease. They found that unsaturated fats lowered the ratio (good thing) and saturated fats did not change it (neither good nor bad).

Then they looked at individual saturated fatty acids: lauric acid increased HDL and lowered the ratio (good); myristic and palmitic acid had no effect (neither good nor bad, neutral); stearic acid slightly lowered the ratio ( slightly good).

This does NOT show that saturated fat worsens blood lipid profiles, but to be more or less neutral. They end by saying that the findings should be confirmed by prospective observational studies. Just such a study by Walter Willett and his Harvard team does indeed confirm these results.

If the alzheimers article is up to date (why does no one ever date their articles?), I wonder why they used the old study? Fishy.