“I was surprised to see significant biomarker movement in a study with so few people and with treatment for such a short amount of time,” said Anne Fagan at Washington University in St. Louis, Missouri. She was not involved in the study. “I think the results lend credibility to pursuing this line of research further.”

In animal models of AD, high-saturated fat diets are known to accelerate amyloid deposition. In people, epidemiological studies have suggested that a diet low in saturated fats and simple sugars, and high in fruits, vegetables, polyunsaturated fats, and lean proteins may help delay or prevent AD (see, e.g., ARF related news story), but there was little evidence linking diet to changes in AD pathology.

To begin to make this connection, first author Jennifer Bayer-Carter looked at the effect of an intensive diet intervention on CSF biomarkers. Bayer-Carter and colleagues provided all food for four weeks to the participants, who consisted of 20 healthy elderly volunteers and about 30 elderly people with amnestic mild cognitive impairment (aMCI). The volunteers were randomly assigned to eat either a diet that was high in saturated fats and simple sugars, or one that was low in both. The researchers designed the meals to have the same number of calories as the participants normally ate, so weight gain or loss was not a factor. Because the researchers provided all food, they could tightly control the diets, which Craft speculates may be one of the reasons they were able to get significant results over such a short time period. The scientists kept the study short to limit any negative health consequences to participants eating high-saturated fat diets. The researchers collected CSF at the beginning and end of the experiment. Participants fasted for 12 hours before donating CSF, which Craft notes is becoming the standard way to do it, as CSF Aβ levels change throughout the day in response to metabolism.

After four weeks, the researchers found numerous statistically significant changes between groups. Several measures of health, such as cholesterol levels and insulin sensitivity, improved in both groups on the healthy diet and worsened in both groups on the unhealthy diet. People with aMCI seemed to be more sensitive to the effects of diet, however, in many cases improving twice as much as the healthy controls on the good diet, and worsening twice as much on the bad diet. This effect did not correlate with any difference in baseline diet or weight between controls and people with aMCI, Craft said; instead, it seemed to be a disease-related vulnerability.

When the researchers looked at AD biomarkers, they saw significant changes in levels of CSF Aβ42. Low levels of CSF Aβ42 are strongly linked to AD, with the explanation being that as Aβ42 clumps into plaques, less of it is available to enter the CSF. In this study, however, the authors found that cognitively normal volunteers had lower levels of CSF Aβ42 on the healthy diet, and higher levels on the unhealthy diet. People with aMCI, on the other hand, showed a dramatic increase in CSF Aβ42 on the healthy diet, and little change on the unhealthy diet. Although puzzling at first, these seemingly contradictory findings would make sense if concentrations of Aβ42 in brain reach a “tipping point” where they begin to form plaques, the authors suggest. Under this hypothesis, cognitively normal people have not yet reached this tipping point, so their brain Aβ42 remains soluble and moves easily into the CSF. In that case, CSF Aβ42 levels would directly reflect brain levels, and dropping levels could actually indicate better brain health, for example, if the brain is producing less Aβ. Cognitively impaired people would be past the tipping point, however, with brain Aβ42 already tied up in plaques. The healthy diet may begin to break up plaques, resulting in a rise in CSF Aβ42 as the newly soluble peptide is cleared from the brain. This model would go against the conventional wisdom that low CSF Aβ is bad news, but scientists contacted for this article found the concept intriguing and worthy of further study.

“This idea of a tipping point is a very interesting hypothesis, and one that is supported by the animal literature,” Fagan said. “A lot of us view low CSF Aβ as bad and high CSF Aβ as good, but that may not be the case, depending on where in the disease trajectory you are.” Animal studies show that CSF Aβ rises until deposits begin to form in the brain, after which the levels drop off. People with Down's syndrome, who accumulate Aβ throughout life due to an extra copy of the gene for APP, show a similar pattern of initially high CSF Aβ followed by low.

Other AD biomarkers changed on the two diets as well. Lipids known as F2-isoprostanes, which mark neuronal injury, decreased in the CSF after four weeks of healthy eating, but increased in tandem with Aβ42 in cognitively normal volunteers eating a poor diet. Levels of the cholesterol chaperone ApoE increased on the healthy diet and decreased on the unhealthy diet. A variant of ApoE is the primary genetic risk factor for sporadic AD. Although scientists do not yet fully understand what ApoE does, the data suggest another mechanism by which diet may affect AD risk, Craft speculated.

Importantly, the two diets also had a functional effect. At the end of the study, people on the healthy diet had improved in a test of delayed visual memory, while the performance of cognitively normal volunteers who ate a high-saturated fat, high-sugar diet worsened. Tests of executive function, motor speed, and other forms of memory showed no change. To find any effect on memory in such a short time period is striking, said Mark Mattson at the National Institute on Aging, Bethesda, Maryland. If the results hold true in future studies, it would suggest that “Your diet can affect your learning and memory, and perhaps your vulnerability to cognitive impairment,” Mattson said. Mattson pointed out that lowering total caloric intake slows down disease and improves memory in AD transgenic mice (see also ARF related news stories on the benefits of caloric reduction in animals [ARF related news story] and humans [ARF news story]). Caloric restriction pumps up levels of brain-derived neurotrophic factor (BDNF), which protects neurons, Mattson said. It would be interesting to compare CSF BDNF levels in people on the high-saturated fat, high-sugar diet with those on the healthy diet, Mattson suggested. Mattson also recommended neuroimaging studies, such as functional MRI and MR spectroscopy, to more directly measure what might be happening in the brains of participants.

Fagan agreed that it would be nice to see more direct measures of brain pathology, for example, amyloid imaging, in people on different diets. “The biggest unresolved issue is, What does the change in CSF Aβ42 really reflect?” Fagan said. Amyloid imaging and longer-term studies might help answer this question, she suggested.

Craft agrees. She is particularly interested in what happens to CSF Aβ during mid-life, the hypothetical tipping point when Aβ starts to deposit in some people’s brains. To answer this, Craft is examining CSF Aβ from longitudinal studies on middle-aged adults, and notes that the Dominantly Inherited Alzheimer Network might also be a rich source for such data. In addition, Craft is repeating the diet intervention study in middle-aged volunteers to see if they respond differently from the elderly participants, whose average age was close to 70.

In follow-up studies in both animals and humans, Craft and colleagues will look more closely at the biology that underlies the dietary effects. One possibility is that diet alters ApoE’s efficiency in chaperoning Aβ, Craft suggested. A high-saturated fat diet induces insulin resistance, and that may interfere with the brain’s ability to clear β amyloid as well. Although the mechanisms are not yet clear, “Our study shows in a proof-of-concept manner that a controlled diet intervention can modulate levels of critical markers of AD pathology,” Craft said.

Fagan suggested that it would be interesting to look at the biomarker changes in individual participants in the diet study and see if they correlated with people’s baseline biomarker levels. “Ultimately, people are going to want individualized medicine,” Fagan said. “People will want to know, if my diet changes, what’s the change in my AD risk? We are clearly not there yet.”—Madolyn Bowman Rogers.

Reference:
Bayer-Carter JL, Green PS, Montine TJ, Vanfossen B, Baker LD, Watson GS, Bonner LM, Callaghan M, Leverenz JB, Walter BK, Tsai E, Plymate SR, Postupna N, Wilkinson CW, Zhang J, Lampe J, Kahn SE, Craft S. Diet intervention and cerebrospinal fluid biomarkers in amnestic mild cognitive impairment. Arch Neurol. 2011 Jun;68(6):743-52. Abstract