|Home | About | Journals | Submit | Contact Us | Français|
Depression associated with low plasma Amyloid-β peptide 42 (Aβ42) leading to a high ratio of Aβ40/Aβ42, a biomarker of Alzheimer’s disease (AD), may represent a unique depression subtype. The relationship between low plasma Aβ42 in depression and the major risk factor of AD, Apolipoprotein E4 (ApoE4), is unknown. With the goal of clarifying this relationship, we analyzed 1060 homebound elders with ApoE characterization and depression status in a cross-sectional study. Plasma Aβ40 and Aβ42 were measured, and cognition were evaluated. In the absence of the ApoE4 allele, depressed subjects had lower plasma Aβ42 [median (Q1, Q3): 17.1 (11.6, 27.8) vs. 20.2 (12.9, 32.9) pg/ml, P = 0.006], a higher Aβ40/Aβ42 ratio [median (Q1, Q3): 7.1 (4.6, 11.3) vs. 6.9 (3.4, 9.7), P = 0.03], and lower cognitive function (mean ± SD of Mini-Mental State Examination: 24.5 ± 3.1 vs. 25.5 ± 3.3, P < 0.0001) than those without depression. In contrast, these relationships were not observed in the presence of ApoE4. Instead, regardless the depression status ApoE4 carriers had lower plasma Aβ42 and a higher Aβ40/Aβ42 ratio than non-ApoE4 carriers. Using multivariate logistic regression, it was found that depression was not associated with ApoE4 allele, but with the interaction between plasma Aβ42 and ApoE4 (OR = 3.94, 95% CI = 1.50, 10.33, P = 0.005), denoting low plasma Aβ42 in the absence of ApoE4. Both ApoE4 carriers and non-ApoE4 carriers with depression had lower Aβ42 and a higher Aβ40/Aβ42 ratio in plasma compared to non-ApoE4 carriers without depression in the homebound elderly. Since a combination of low plasma Aβ42 and high plasma Aβ40 has been shown to increase the risk of AD in two large cohort studies, amyloid-associated depression shown in this study may suggest a risk factor of AD in the absence of ApoE4.
Depression is a clinical syndrome with multiple pathologies and etiologies, and prodromal depression of Alzheimer’s disease (AD) is one subtype of depression in the elderly 1-3. Clinically, it is often seen that some depressed, elderly patients are diagnosed and treated for depression, and then longitudinally these patients develop cognitive decline and AD. Therefore, finding specific biomarkers to differentiate prodromal depression of AD from other depression subtypes is critical, especially when therapies targeting at the early stage of AD are currently under development.
Our recent study has shown that depressed elders without cardiovascular disease (CVD) have a lower concentration of plasma Aβ42, a biomarker of AD, compared to depressed elders with CVD and those without depression in a homebound population 4. In addition, we found that depression with a high plasma Aβ40/Aβ42 ratio, determined by both low Aβ42 and high Aβ40, is associated with poor memory in the same population 5. Because two large cohort studies have shown that a high Aβ40/Aβ42 ratio in plasma increases the risk of developing AD 6, 7, it is hypothesized that depression characterized by low Aβ42 and a high Aβ40/Aβ42 ratio in plasma may represent prodromal depression of AD that is distinct from vascular depression 8. We have termed it “amyloid-associated depression”.
The Apolipoprotein E4 (ApoE4) allele is the major genetic risk factor of late-onset AD 9. Some studies have shown that late life depression is not associated with ApoE4 10, 11. However, among the elderly aged 80 and older who have a much higher AD prevalence than those with a younger age, a relationship between ApoE4 and depression is observed, suggesting the existence of a prodromal depression of AD 12, 13. ApoE4 is associated with higher amyloid plaque burden composed mainly by Aβ42 in AD brains than those carrying ApoE2 and E3 14, probably through reduced clearance or accelerated aggregation of the peptide. Although the meaning of the relationship between depression and low concentration of Aβ42 in plasma is unclear, plasma Aβ42 declines significantly in the preclinical stage of AD 15-18. These studies suggest that reduced plasma Aβ42 either plays a role in AD pathogenesis or is a biomarker predicting the onset of AD. Since both ApoE4 and depression could be a prodromal stage of AD, we hypothesize that they may share some AD biomarkers, such as Aβ in plasma. To investigate this possibility, we explored the relationship between plasma Aβ42 and depression in those with and without the ApoE4 allele in a homebound elderly population.
We studied a group of 1060 subjects, all of whom had been characterized for ApoE and depression status from an ongoing, population-based study, the Nutrition, Aging and Memory in the Elderly (NAME) study. Subjects included homebound elderly clients who were enrolled in one of four homecare agencies in the Boston area between 2003 and 2006. Elders aged 60 and older were invited to participate in the study. Anyone receiving homecare services was registered with one of these agencies if he/she lives in the city of Boston, had an annual income < $18,890 and needed homecare service. All homebound elders aged 60 and older at each of the four agencies were invited to participate in the study. Eligibility for enrollment required that the participants spoke English, were physically able to participate in the study home visits, and had sufficient vision and hearing to read and hear the content of the neuropsychological tests. Those with Mini-Mental State Examination (MMSE) ≤ 10 or verbal IQ < 75 were not eligible to continue in the study. Of all eligible subjects, 66% enrolled in the study, and gave informed consent. Each subject participated in three home visits conducted by a research assistant, who drew fasting blood and collected data on depression and medical conditions 19.
A 244 bp fragment of the apoE gene including the two polymorphic sites was amplified by PCR using a robotic Thermal Cycler (ABI 877, Perkin-Elmer/Applied Biosystems), using oligonucleotide primers F4 (5′-ACAGAATTCGCCCCGGCCTGGTACAC-3′) and F6 (5′-TAAGCTTGGCACGGCTGTCCAAGGA-3′). The PCR products were digested with 5 units of Hha I and the fragments separated by electrophoresis on 8% polyacrylamide non-denaturing gel. The specific allelic fragments were: E2; E3; and E4. ApoE4 was defined by E4/4, E3/4 or E2/4 20.
The blood samples were centrifuged immediately after the blood draw. The sandwich Aβ ELISA was used. Plates were coated with 2G3 (anti-Aβ40) and 21F12 (anti-Aβ42) antibodies overnight at 4°C. Samples were then loaded and incubated overnight at 4°C followed by incubation with a biotinylated monoclonal anti-N terminus Aβ antibody (3D6B) for 2 hrs. Finally, streptavidin-conjugated alkaline phosphatase (Promega, USA) was added and incubated, and the signal was amplified by adding alkaline phosphatase fluorescent substrate (Promega, USA), which was then measured. The lowest detection for both Aβ peptides was 1.6 pg/ml in the standard curves with %CV between 1.1 to 7.2. However, we used 3.1 pg/ml for both Aβ 1-40 (2 sample) and 1-42 (10 samples) as a low cut-point. Aβ samples were treated as the cut-points for comparison and regression if their levels were below the cut-point of detection. The samples with higher levels than the standard curve were repeated with dilutions for measurement. The intra-correlations with two other laboratories, which have published the results of the Aβ measurement 21, 22, showed R = 0.63 and 0.84 for Aβ40 and R = 0.90 and 0.96 for Aβ42.
Depressive symptoms were assessed by using the Center for Epidemiological Studies Depression scale (CES-D) 23; a CES-D score of ≥ 16 was used as the cut-off point for clinical depression 24. This CES-D cut-off point had a sensitivity of 0.90 and a specificity of 0.83 for the DSM-IV diagnosis of major depression by a board-certified psychiatrist in our study. Vascular depression was defined by a CES-D score ≥ 16 and the presence of CVD.
Research assistants, trained by a board certified neuropsychologist, administered the cognitive tests. The subjects were screened for severe cognitive impairment using the Mini Mental State Examination (MMSE) 25. Those with MMSE < 24 were defined for cognitive impairment and MMSE ≥ 24 for intact cognition. WMS-III Logical Memory: Two stories (A and B) were read aloud to the subject; the subject was then asked to repeat each story to assess immediate recall. After 30 minutes, the subject was asked to repeat both stories, which served as a measure of delayed recall.
Subjects were classified as having CVD according to whether they had been previously informed by a doctor that they had congestive heart failure, coronary heart disease, angina pectoris or a heart attack. Stroke history was recorded. Body mass index (BMI) was measured and calculated as body mass (in kg) over the squared height (in m2). Current hypertension was defined by the average of systolic blood pressure > 140 mm Hg or diastolic blood pressure > 90 mm Hg at two determinations. Renal function, which is associated with plasma Aβ 26, was assessed by measurements of serum creatinine.
Statistical analysis was performed using SAS (version 9.1). Normally distributed variables such as age were presented as mean ± SD and compared using T-test. Variables with skewed distributions (plasma Aβ40 and Aβ42) were presented as median (25th, 75th percentiles) and compared using Wilcoxon rank sum test 4. The Chi-Square test was used to compare proportions for binary endpoints. Both Aβ40 (Log Aβ40) and Aβ42 (Log Aβ42) were transformed to log10 for multivariate regression due to skewed distributions. Logistic regression was used to examine associations between depression (CES-D score ≥ 16 vs. not) or intact cognition (MMSE score > 24 vs. not) and Log Aβ40 or LogAβ42 while adjusting for CVD, ApoE4 and other confounders of age, race, gender, school and creatinine. The interactions between LogAβ42 and ApoE4 or between LogAβ42 and depression were applied to the logistic models. For all analyses, the two-sided significant level of 0.05 was used.
One thousand and sixty subjects from the ongoing NAME study characterized for ApoE were used in this analysis. The average age (mean ± SD) of this population was 75.3 ± 8.4 years old, and 76% were female. It was multi-ethnic with 61% Caucasian, 35% African American and 4% other ethnicities. Sixty two percent of subjects had at least high school education. Depression, defined as a CES-D score ≥ 16, was observed in 34% (365/1060) of the subjects. ApoE allele frequencies were ApoE2/2 = 15/1060 (1%); ApoE2/3 = 141/1060 (13%); ApoE2/4 = (32/1060 (3%); ApoE3/3 = 649/1060 (61%); ApoE3/4 = 204/1060 (19%) and ApoE4/4 = 19/1060 (2%). Overall, there were 255 subjects (24%) carrying at least one ApoE4 allele.
Table 1 shows that other than ethnicity there were no differences in demographics in those with and without ApoE4 allele. There were a higher proportion of African American subjects in those carrying the ApoE4 allele than those without ApoE4 (49% vs. 33%, P < 0.0001). The rates of cardiovascular disease (CVD) and other vascular diseases of diabetes, stroke and current hypertension were similar in those with and without ApoE4. Compared to non-ApoE4 carriers, ApoE4 carriers were found to have a comparable rate of depression (Table 1) or average CES-D score (Mean ± SD: 13.3 ± 10.2 vs. 12.7 ± 10.2, P = 0.42). The proportions of those with cognitive impairment, as defined by MMSE score < 24, were similar in those with and without the ApoE4 allele.
ApoE4 carriers had a lower concentration of plasma Aβ42 (median: 17.4 vs. 19.2, P = 0.02) and a higher Aβ40/Aβ42 ratio (median: 7.9 vs. 6.9, P = 0.002) than non-ApoE4 carriers (Table 2). In contrast, no differences in plasma Aβ40 (median: 137.3 vs. 130.6, P = 0.34) and creatinine (mean ± SD: 1.3 ± 1.5 vs. 1.1 ± 0.8, P = 0.54) were found in those with and without ApoE4.
Since depression is associated with low plasma Aβ42 4, in this study we examined the effects of the ApoE4 allele on depression and plasma Aβ peptides. Subjects were further divided into depression and no depression subgroups in both ApoE4 carriers and non-ApoE4 carriers (Table 3). In the absence of the ApoE4 allele, depressed subjects had significantly lower plasma Aβ42 [median (Q1, Q3): 17.1 (11.6, 27.8) vs. 20.2 (12.9, 32.9) pg/ml, P = 0.006] (Figure 1) and a higher Aβ40/Aβ42 ratio [median (Q1, Q3): 7.1 (4.6, 11.3) vs. 6.9 (3.4, 9.7), P = 0.03] than those without depression. When subjects with CVD were excluded from this subset of non-ApoE4 carriers, the differences in plasma Aβ42 concentration (median: 15.7 vs. 20.7, P < 0.0001) and Aβ40/Aβ42 ratio (median: 7.4 vs. 6.3, P = 0.002) in those with and without depression became more significant. In contrast, in the presence of ApoE4, no difference in plasma Aβ42 was found in those with and without depression (Figure 1). Nevertheless, it was noted that ApoE4 carriers regardless of depression status and non-ApoE4 carriers with depression had lower plasma Aβ42 compared to non-ApoE4 carriers without depression (Figure 1 and Figure 3). No difference in plasma Aβ40 was found between the subgroups with and without depression regardless of ApoE status (data not shown).
Additionally, in the absence of ApoE4, depressed subjects had significantly lower scores on MMSE (mean ± SD: 24.5 ± 3.1 vs. 25.5 ± 3.3, P < 0.0001), and a greater number had cognitive impairment defined by a MMSE score lower than 24 (39% vs. 22%, P < 0.0001) (Figure 2) compared to those without depression. Furthermore, in the absence of ApoE4, depressed subjects had lower scores on logical memory (mean ± SD: 17.3 ± 9.7 vs. 19.4 ± 9.6, P = 0.002) than those without depression. In contrast, in the presence of ApoE4 no difference in MMSE (mean ± SD: 25.0 ± 3.8 vs. 25.0 ± 3.8, P = 0.88) was found in those with and without depression in this population (Figure 1B).
To further explore the relationship between depression and plasma Aβ42 in the presence and absence of ApoE4, multivariate logistic regression was performed (Table 4). Since it is well documented that late life depression is highly related to vascular diseases, such as CVD, the analytic models were built based on demographic variables in addition to vascular diseases to characterize any potential confounders, which might be expected. Low plasma Aβ42 (OR = 0.63, 95% CI = 0.42, 0.94, P = 0.02) and CVD (OR = 1.79, 95% CI = 1.35, 2.37, P < 0.0001) were independently associated with depression after adjusting for confounders of age, race, gender, diabetes, creatinine and Aβ40. There was no relationship between the ApoE4 allele and depression (Table 4: Model I). Since the relationship between depression and plasma Aβ42 was strengthened after stratification with ApoE4 (Figure 1A), we, therefore, analyzed the interaction between plasma Aβ42 and ApoE4. This interaction, denoting low plasma Aβ42 in the absence of ApoE4, was found to be positively associated with depression with OR = 3.94, 95% CI = 1.50, 10.33, P = 0.005 (Table 4: Model II), and the relationship was further strengthened after adjusting for MMSE with OR = 4.65, 95% CI = 1.73, 12.5, P = 0.002 (Table 4: Model III).
The proportions of depressed African American subjects compared to the non-depressed subgroups were different in the presence (41% vs. 53%, P = 0.01) and absence (36% vs. 31%, P = 0.07) of ApoE4 (Table 3). Ethnicity, however, was not associated with depression (Table 4). While in the absence of ApoE4 more depressed subjects had diabetes compared to those without depression, no difference was found in rates of diabetes between these subgroups in the presence of ApoE4 (Table 3). Again, diabetes status was not found to be associated with depression in this study sample (Table 4). The rates of stroke between those with and without depression were similar in the presence or absence of ApoE4, and stroke was not found to be associated with depression using multivariate regression analysis (data not shown).
ApoE4 is the major risk factor of AD 9, and is associated with low level of Aβ42 in cerebral spinal fluid (CSF) even among the elderly with normal cognition27. Late life depression has also been shown to occur prior to AD 1-3. Using a homebound elderly population, the subset of depressed non-ApoE4 carriers was found to have low plasma Aβ42 and a high Aβ40/Aβ42 ratio, as would be expected had they been ApoE4 carriers (Tables (Tables22 and and4,4, Figure 1 and Figure 3). And these depressed subjects had cognitive impairment, including poor memory (Figure 2 and Figure 3). As a high plasma Aβ40/Aβ42 ratio has been found to increase the risk of AD in two longitudinal studies6, 7, we hypothesize that amyloid-associated depression, characterized by low plasma Aβ42 and high Aβ40/Aβ42 ratio, might represent another preclinical stage or a risk of AD in addition to ApoE4.
We recently reported 4 that there are at least two separate depression subtypes in this homebound elderly population: 1) vascular depression 8, which is linked with CVD and not necessarily associated with low plasma Aβ42 28, and 2) amyloid-associated depression, which is associated with low plasma Aβ42 28. In this study, we found that depression in the absence of ApoE4 was associated with low plasma Aβ42 and high Aβ40/Aβ42 ratio (Figure 1). In the presence of the ApoE4 gene, however, low plasma Aβ42 and high Aβ40/Aβ42 ratio were found regardless of depression status (Table 2 and Figure 1). Plasma Aβ42 levels decline significantly prior to the onset of AD and in its early stages 15, leading to a high plasma Aβ40/Aβ42 ratio which significantly increases the risk of AD longitudinally 6, 7. Depression may also precede AD 29-33. Therefore, while ApoE4 might lead to AD regardless of depression status, amyloid-associated depression accompanied by a decline in plasma Aβ42 may herald AD even in the absence of the ApoE4 allele.
Studies have shown that plasma Aβ peptides cannot be used to diagnose AD 34, 35. Our findings in this study were consistent with these previous studies in that we did not find a relationship between cognition and plasma Aβ42 alone in the homebound elderly (data not shown). We found, however, that cognitive impairment was associated with the co-occurrence of low plasma Aβ42 and depression in the absence of ApoE4 (Figure 2), even after adjusting for confounders (data not shown). This result suggests that, despite its lack of usefulness in diagnosing the disease in its later stage, low plasma Aβ42 might represent a prognostic biomarker indicating a pre-clinical stage or an early stage of AD 6, 7, especially in those who have depression. Low Aβ42 in CSF, a consensus diagnostic test for AD 36, correlates with plasma Aβ42 at the pre-clinical stage of AD, but this relationship disappears in the later stage of the disease 37.
Multiple etiologies of AD involving multiple pathways can result in the brain AD pathologies: Aβ plaques and neurofibrillary tangles. ApoE4 is one major AD risk factor, and its role in AD pathogenesis is probably mediated through increased aggregation of Aβ or impaired clearance of Aβ in the brain (reviewed by Holtzman DM, 2004) 38. It is not yet clear how late life depression influences the degenerative process of AD. A history of depression correlates with an increase in both amyloid plaques and neurofibrillary tangles in AD 39. On the other hand, approximately 80% of the depressed elderly do not develop AD after a 4-year follow-up 3. This suggests that there are different subtypes of depression in the elderly including 1) early-onset depression; 2) post-stroke depression 40; 3) vascular depression related to CVD and other vascular risk factors that lead to executive dysfunction 8; 4) pre-clinical depression of AD and 5) comorbid depression in AD. In our study, after adjusting for MMSE the relationship between depression and plasma Aβ42 in the absence of ApoE4 was strengthened (Table 4: Model III), suggesting that different depression subtypes have different cognitive patterns and prognoses. Since these depression subtypes may have different underlying pathologies, not all would be expected to be related to plasma Aβ, or plasma Aβ might bear upon the subtypes in different ways. The finding that depression associated with a high plasma Aβ40/Aβ42 ratio present with significant poorer memory compared to those with a low plasma Aβ40/Aβ42 ratio argues that amyloid-associated depression is probably a preclinical stage of AD 5.
Our study did not find a difference in depression rates between those with and without ApoE4 (Table 1). This result was consistent with previously published studies that showed that the ApoE4 allele does not increase the risk of late life depression 10, 11. Other studies showed that in subjects aged 80 and older, a relationship between ApoE4 and depression is observed 12, 13, probably suggesting a prodromal depression of AD. It is noted that the average age of our study sample was 75. Longitudinal studies have shown that ApoE4 increases the risk of cognitive decline in the elderly 41-43. In our study, those non-ApoE4 carriers with depression had lower cognitive function compared to those who did not have depression (Figure 2). At cross-sectional level, those ApoE4 carriers tended to have more cognitive impairment than those non-ApoE4 carriers without depression, but did not reach the statistical significance.
Although our study and others 6,7 have shown that low plasma Aβ42 levels have a potential to be used as a prognostic biomarker of AD, another population study reports that high plasma Aβ42 is associated with the risk of AD, but plasma Aβ42 declines before the onset of the disease44, which could lead to low levels of plasma Aβ42 at the prodromal stage. Unlike late-onset AD, both early-onset AD 45 and Down syndrome 46 present with high plasma Aβ42 levels at the pre-clinical stage. One study shows that MCI patients have higher levels of plasma Aβ42 than the controls only in women 47. It is not yet known whether plasma Aβ status is different at the pre-clinical stage of AD with and without depression. Importantly, despite that our assays in measurements of Aβ40 vs. Aβ42 in plasma had high sensitivity and specificity, the field lacks standard assays for inter-laboratory comparisons. Therefore, the complexity of plasma Aβ 48, 49 requires brain investigation and a longitudinal follow-up to validate amyloid-associated depression as a unique depression subtype. Other limitations of this study include: 1) the cross-sectional design prevents the study from exploring the possibility that depressed elderly with ApoE4 have shorter life spans and would therefore be less likely to be included in this study; 2) depression was based on the CES-D score rather than DSM-IV criteria, and we have no information about its onset and the course; 3) we are unable to determine whether amyloid-associated depression is a precursor of AD without longitudinal data of the study. Nevertheless, our findings link amyloid-associated depression characterized by low plasma Aβ42 and a high Aβ40/42 ratio to cognitive impairment specifically in non-ApoE4 carriers. Prospective studies and brain investigations are needed to determine whether amyloid-associated depression is an independent risk factor for AD.
The authors are grateful for Dr. Dennis J. Selkoe for his support and collaboration. We thank the NAME study staff and the Boston homecare agencies for their hard work and acquisition of subjects, and Xian Adiconis from Dr. Jose Ordovas’s laboratory for her characterization of ApoE alleles. This work was supported by grants from NIA, AG-022476 for W.Q.Q and AG-21790 for I.R. Support was also provided through the General Clinical Research Center funded by the National Center for Research Resources of the NIH under grant no. MO1-RR00054.