Homocysteine and Cognitive Function in a Population-based Study of Older Adults

Jyme H. Schafer, MD, MPH; Thomas A. Glass, PhD; Karen I. Bolla, PhD; Margaret Mintz, MS; Anne E. Jedlicka, MS; Brian S. Schwartz, MD, MS


J Am Geriatr Soc. 2005;53(3):381-388. 

In This Article


In crude and adjusted analysis, elevated homocysteine levels were associated with poorer scores in all tested cognitive domains in this large population-based study of randomly selected adults aged 50 to 70. Homocysteine was an independent predictor of worse test scores in all subgroups examined, including men and women and African Americans and whites, and across a range of ages, blood lead levels, and LDL-C levels. Interesting evidence was also found that ApoE haplotype may modify the influence of homocysteine on neurobehavioral function, with a more pronounced effect in subjects with the ApoE-ε4ε4 haplotype. Furthermore, the average magnitude of the associations was large; an increase of homocysteine from 7.6 µmol/L to 11.3 µmol/L, the interquartile range, was equivalent to an increase of 4.2 years of age at the first study visit. On average, those with higher homocysteine levels had almost twice the odds of having the lowest test scores. Although in this cross-sectional analysis causal inferences cannot be made, it is guardedly believed that, because the observed associations were consistent and large, the results have important public health implications.

The Baltimore Memory Study has several strengths, including a large, community-dwelling, general-population sample, with large numbers of African Americans and whites and a comprehensive neurobehavioral test battery spanning several cognitive domains. Previous studies had suggested an inverse relationship between homocysteine levels and test scores in various cognitive domains in late life. In the Normative Aging Study, an association was observed, but only with spatial copying.[12] In one study, two birth cohorts (Aberdeen 1921 Birth Cohort and Aberdeen 1936 Birth Cohort) from the Scottish Mental Surveys of 1932 and 1947 were tested using the MMSE (a screening test for global cognitive function), National Adult Reading Test, Raven's Progressive Matrices (nonverbal intelligence), Auditory Verbal Learning Test, and the Block Design (spatial copying) and the Digit Symbol subtests of the revised Wechsler Adult Intelligence Scale.[6] In the older cohort (aged 78 at testing), an association between homocysteine and cognitive function was observed in four of the six tests. In participants with a mean age of 74, it was found that homocysteine levels explained 11% of the variance in the Cambridge Cognitive Examination score, independent of intelligence quotient.[4] The Rotterdam Scan Study evaluated 1,077 subjects aged 60 to 90, with the MMSE, Stroop (three parts), Letter-Digit Substitution Task, Verbal Fluency Test, Paper and Pencil Memory Scanning Task (three parts), and 15-word verbal learning test.[10] Comparing the upper quintile of homocysteine levels with the lower four quintiles, there was a mean difference in psychomotor speed of -0.26 (95% CI=-0.37 to -0.14), memory of -0.13 (95% CI=-0.27-0.01), and global cognitive index of -0.20 (95% CI=-0.30 to -0.11). In a study of 1,270 National Health and Nutrition Examination Survey III participants who were aged 60 and older, hyperhomocysteinemia (>13.7 µmol/L) was related to poorer performance on a paragraph and word delayed-recall test.[9] Although not evaluated in many prior studies, in at least one study, ApoE haplotype was found to be unrelated to the homocysteine and cognitive function relation.[5]

Determinants of homocysteine levels include physiological factors (age, sex, renal function, postmenopausal state); lifestyle factors (smoking, alcohol); disease states (renal failure, thyroid disease, diabetes mellitus); and pharmacological, nutritional, and genetic factors.[38] Homocysteine levels are modifiable, and such modification could prevent cognitive dysfunction and decline. The exact mechanism of homocysteine's effect on the vascular system is unknown but has been postulated to include vascular damage through impaired vascular endothelial and smooth muscle cell function, possibly involving inhibition of nitric oxide synthesis, increased oxidative stress, endoplasmic reticulum stress, receptor-induced cell death, and activation of pro-inflammatory factors.[17,19,39,40] Although it is thus possible that homocysteine may be associated with poorer cognitive function because of an influence on cerebral microvasculature, it is also important to note that homocysteine has been found to directly damage neurons.[18,20,40]

There is considerable biological variation in homocysteine levels, as well as significant measurement error.[38] A recent summary of expert opinions about homocysteine concluded that, in cross-sectional studies, a single measurement of homocysteine in each individual will underestimate the true strength of association between homocysteine and disease by about 15%.[38] Therefore, the associations reported herein, already consistent, strong and unlikely to be due to chance, are likely to be an underestimate of the true associations. Similarly, the variability between fasting and nonfasting samples and the fact that a nonfasting result was used is also likely to result in an underestimation of the true associations.

If homocysteine is indeed an independent, modifiable risk factor for cognitive dysfunction, what is the evidence that interventions that lower homocysteine levels also decrease the risk of cardiovascular or central nervous system outcomes? Unfortunately, studies on risk reduction by vitamin therapy to lower homocysteine levels have been performed only in subjects who already have significant and clinically apparent vascular disease. In patients with known cardiovascular disease, studies have shown a slowing in the progression of coronary and peripheral atherosclerosis and, in a separate, more recent study, an increased risk of in-stent restenosis.[41,42,43] In another recent study, moderate reduction of homocysteine after nondisabling stroke had no effect on vascular outcomes through the 2-year follow up period.[44] Published randomized trials assessing the ability of vitamin therapy to improve cognitive function by lowering homocysteine levels have had significant limitations, including having study populations that were already cognitively impaired or demented, using global screening tests such as the MMSE that lack sensitivity, and having short follow-up durations. The longest duration of study was 12 weeks.[38,45]

Ultimately, an improved understanding of the risk factors for cognitive dysfunction is needed to prevent and possibly reverse the commonly observed age-related decline in cognitive function. Homocysteine may be a biomarker for vascular disease induced cognitive dysfunction. The existing evidence supports further investigation of whether interventions can be directed at this risk factor.

CLICK HERE for subscription information about this journal.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.