These patients exhibited a plasma cytokine profile that was exactly the same as that of Alzheimer's disease (AD) transgenic mice that were given caffeinated coffee and didn't progress to dementia. It's therefore very likely that it's caffeine from coffee, and not from other sources, that affords the cognitive protection, said study senior author Gary W. Arendash, PhD, research scientist, Bay Pines Veterans Affairs Hospital, St. Petersburg, Florida.

The research also suggests that certain cytokine patterns could signal for impending conversion to dementia among those with MCI, said Dr. Arendash. The study is published in the June issue of the Journal of Alzheimer's Disease.

Lower Caffeine Levels

The new case-control study included 2 cohorts of 124 participants in a Florida Alzheimer's Disease Research Center study of persons aged 65 years and older. All participants had undergone a battery of baseline neurologic assessments and cognitive tests and were categorized as normal, MCI, or dementia. As well, researchers had access to fasting blood samples taken at baseline.

Over the next 2 to 4 years, researchers annually reassessed the cognition of the participants. They separated the participants into 5 groups: (1) initially normal and remained normal, (2) initially normal but converted to MCI, (3) initially MCI and remained so, (4) initially MCI but converted to dementia, and (5) initially dementia and remained so.

Analysis of plasma caffeine levels from the initial visit showed significantly lower caffeine levels in participants with MCI relative to the normal group (P < .03). Caffeine levels were also lower in participants with dementia than in those with normal cognition, but this association did not reach statistical significance (P < .07).

There was a 26% lower plasma level of caffeine in normal persons who converted to MCI over the course of the study compared with those who remained normal, but this was not significant because of considerable variability in caffeine levels among individuals in both of these subgroups.

However, 11 patients with MCI who progressed to dementia had plasma caffeine levels that were 51% below levels at study initiation vs those with MCI who remained MCI (P < .02).

None of the MCI participants who converted to dementia had initial caffeine levels above 1200 ng/mL while half of those with stable MCI had higher levels. Baseline plasma caffeine levels greater than 1200 ng/mL in MCI patients were associated with a 100% chance of avoiding progression to dementia during the 2- to 4-year follow-up.

Patients with MCI in both the Miami (n = 81) and the Tampa (n = 43) study cohorts independently showed the same relationship between blood caffeine levels and later risk for dementia progression.

Critical Level

That 1200-ng/mL level appears to be an important threshold, said Dr. Arendash. The amount of coffee needed to reach this critical level appears to be 3 to 5 cups of daily, with a target of 5 cups or 500 mg of caffeine. Those previous AD mouse studies showed that 1 to 2 cups, or between 100 and 200 mg of caffeine (which is what typical Americans drink daily), was not enough to ward off dementia, he said. It's not known whether it's necessary to spread those 5 cups throughout the day, he added.

It's important to remember, though, that half of the patients with stable MCI in the study who had caffeine levels below 1200 ng/mL also didn't progress to dementia. Clearly, other factors play a role. Such factors probably include the level of cognitive and physical activity, the presence of hypertension, and antioxidant intake, especially from fruits and vegetables, said Dr. Arendash.

The study also found that 3 cytokines - granulocyte colony-stimulating factors (G-CSF), interleukin-10, and interleukin-6 - were lower in the plasma of patients with MCI who were destined for AD conversion than in both the nonconverting MCI participants and the participants with dementia. None of the 8 other plasma cytokines that were measured showed any such profile when the same 2 MCI subgroups were compared.

"When that initial blood sample was taken, MCI patients that went on to convert to AD had low levels of all those cytokines," said Dr. Arendash. "That could be diagnostic; it could be a very important plasma indicator of impending AD."

The studies of AD transgenic mice, which produce the same abnormal human protein as the human brain, amyloid-beta, demonstrated that long-term oral administration of caffeinated coffee prevents cognitive impairment.

The cytokine profile of the participants in this current study was the same as that in these AD mice. "Their profiles matched identically to the mice given coffee but not other sources of caffeine;" said Dr. Arendash. "That's why we strongly believe that most, if not all, of those MCI patients who did not convert were on habitual coffee intake."

The mouse research allowed investigators to identify disease-modifying mechanisms for caffeine. The studies showed that caffeine alone suppresses brain levels of enzymes required for amyloid-beta production via targeting of specific signal transduction mechanisms. This research also suggested that something in coffee increases plasma levels of those 3 key cytokines: G-CSF, interleukin-10, and interleukin-6. G-CSF, in particular, has beneficial cognitive actions in AD mice that involve synaptogenesis and neurogenesis.

Aside from caffeine, coffee is rich in antioxidants and anti-inflammatory compounds that may also contribute to reduced risk for AD.

This study was a retrospective analysis, so a definitive relationship will have to be derived through a clinical trial in which participants consume caffeinated coffee, other caffeinated products, or decaffeinated coffee, over a period several years, said Dr. Arendash. He suggested that residents of China, where coffee consumption is very rare, would make an ideal control population for such research.

Accumulating Evidence

Reached for a comment, Karen Ritchie, PhD, Faculty of Medicine, Imperial College, London, United Kingdom and Directeur de Recherche, Institut National de la Santé et de la Recherche Médicale, Montpellier, France, said the study, which involved a direct measure of caffeine in plasma rather than just reports of caffeine consumption, adds to accumulating evidence of a beneficial effect of caffeine.

However, she told Medscape Medical News, the study's "weak point" is that, unlike the epidemiologic studies, such as the ones she and her colleagues have carried out, alternative explanations of this observation were not taken into account.

"For example, persons drinking less coffee may also have more hypertension, more depression, more heart disease, less social activity than those with higher levels, and these factors are in themselves related to onset of dementia."

Still, one of Dr. Ritchie's own previous studies, published in Neurology, concluded that the psychostimulant properties of caffeine appear to reduce cognitive decline in women without dementia, especially at higher ages.

The question of whether some people are protected against dementia because they drink coffee or because they do or have something else that non–coffee drinkers don't, remains unanswered, she said.