Association between homocysteine, depression, and cognitive function in community-dwelling older women from Australia.

Abstract

To the Editor: Depression in later life has been associated with cerebrovascular disease1,2 and, more recently, high circulating levels of homocysteine (Hcy). This study recruited 262 community-dwelling women aged 70 and older (mean age±standard deviation=74.7±4.4) with the aim of investigating the association between total plasma Hcy and depression (as assessed using the Beck Depression Inventory (BDI)), anxiety (as measured using the Beck Anxiety Inventory (BAI)), and cognitive function (as measured using the Cambridge Mental Disorders of the Elderly Examination–Cambridge Cognitive Capacity Scale (CAMDEX-CAMCOG). General health was assessed with a single question: “In general, would you say that your health is good to excellent, fair, or poor?” There was a moderate and significant direct association between BDI scores and total plasma Hcy (Spearman rho=0.20, P=.001). Because of high skewness in the distribution of the data, BDI scores were divided into quartiles (0–3, 4–7, 8–10, and ≥11) (Figure 1). Total plasma Hcy levels increased progressively with increasing BDI quartiles (F=4.79, df=3, P=.003). The graded association between total plasma Hcy and BDI quartiles remained significant (F=3.92, df=3, P=.009) even after the analysis was simultaneously adjusted for the effects of age, poor self-perceived health, and folate and B12 levels. Furthermore, participants who were clinically depressed (BDI>13, n=44) had significantly higher total plasma Hcy levels than nondepressed subjects (mean±SD=12.4±4.5 vs 10.8±3.6, t=2.49, P=.013). Again, this association was independent of age, poor self-perceived health, and folate and B12 levels (F=9.10, df=1, P=.003). Total plasma homocysteine according to Beck Depression Inventory (BDI) quartile scores. There was no clear association between BAI scores and Hcy levels (rho=0.10, P=.11). There was a weak negative association between total plasma Hcy and CAMDEX-CAMCOG score (rho=−0.12, P=.05) and a significant difference in the distribution of plasma Hcy according to CAMDEX-CAMCOG quartiles (F=3.00, df=3, P=.031). However, the graded association between total plasma Hcy and CAMDEX-CAMCOG quartiles disappeared once the analysis was adjusted for age and folate and B12 levels (F=0.35, df=3, P=.789). These results confirm the existence of a graded association between increasing depression scores and total plasma Hcy in community-dwelling women aged 70 and older. In addition, they show that women with clinically significant depressive symptoms have higher total plasma Hcy than nondepressed subjects. These findings are compatible with those of another study3 and the Women's Health and Aging study4 and add to currently available evidence by demonstrating that the association between depression and Hcy remains, even after adjustment for other explanatory variables such as age, self-perceived health, and plasma levels of folate and B12. The mechanisms underlying the association between plasma Hcy and depression are presently unknown, but at least two possibilities should be considered. First, there is compelling evidence that high plasma Hcy increases the risk of cerebrovascular disease,5 which in turn, is associated with depression in later life.1,2 Second, plasma Hcy seems to be a marker of dysfunction in the metabolic pathway of methionine. Methionine is the immediate precursor of S-adenosylmethionine (SAM), the methyl donor of numerous methylation reactions in the brain, many of which are directly involved in the synthesis and metabolism of monoamines.6 As a consequence, SAM seems to regulate several steps in the synthesis of dopamine, norepinephrine, and serotonin, which are neurotransmitters postulated to play an important role in the pathogenesis of depression and anxiety. High plasma Hcy has also been associated with increased risk of Alzheimer's disease in some case-control7 and cohort studies,8 and there is preliminary evidence from cross-sectional neuroimaging investigations that subjects with high plasma Hcy have more cortical atrophy (including the hippocampus) than controls.9 Our results indicate that cognitive function is inversely associated with plasma Hcy, but age and plasma levels of folate and B12 at least partly explain such an association. The Rotterdam study, which observed no association between high Hcy levels and subsequent cognitive decline in a community-representative sample of 702 subjects aged 55 and older, reported similar findings.10 In conclusion, the results of the present investigation show that higher depression scores are associated with increasing levels of total plasma Hcy. This study also found that plasma Hcy is not associated with cognitive performance, as assessed using the CAMDEX-CAMCOG. These findings may have important implications for the design of effective measures to reduce the personal, social, and financial burden associated with depression in later life. The authors are indebted to Dr. Steve Fenner, Dr. Helen McGowan, Dr. Brett Coulson, Dr. Ria Kotynia, Lee Barclay, Felicity Roche, and Kelly Cassidy for their assistance with data collection. The study was supported by grants from the Australian Rotary Health Research Fund, Australasian Menopause Society, and the National Health and Medical Research Council of Australia.