Association of a functional polymorphism in the cholesteryl ester transfer protein (CETP) gene with memory decline and incidence of dementia

Abstract

Context

Polymorphisms in the cholesteryl ester transfer protein gene (CETP) have been associated with exceptional longevity and lower cardiovascular risk but associations with memory decline and dementia risk are unclear.

Objective

To test the hypothesis that a single nucleotide polymorphism at CETP codon 405 (isoleucine to valine; V405; NCBI dbSNP rs5882) is associated with a lower rate of memory decline and lower risk for incident dementia, including Alzheimer’s disease (AD).

Design, Setting and Participants

Prospective cohort study comprising 608 community-dwelling nondemented older adults (≥70 years) from the Einstein Aging Study (EAS; Bronx, NY) with CETP genotype available. 15 individuals with prevalent dementia were excluded; the analytic sample included 593 individuals. Standardized neuropsychological and neurological measures were administered annually from 1994–2009. Linear mixed effects models adjusted for sex, education, race, medical comorbidities, and APOE ε4 examined associations of V405 genotype with longitudinal performance on cognitive tests of episodic memory (Free and Cued Selective Reminding Test; FCSRT), attention (Digit Span), and psychomotor speed (Digit Symbol Substitution). V405 genotype was the main predictor of incident dementia/AD in similarly-adjusted Cox proportional hazards models with age as the time scale.

Main Outcome Measures

Memory decline; incident dementia.

Results

Valine allele frequency was 43.5%. A total of 40 cases of incident dementia occurred during follow-up (mean =[(SD] 4.3 [3.1] years). Compared to isoleucine homozygotes, valine homozygotes had significantly slower memory decline (0.43 points/year of age; 95% confidence interval [CI] −0.58, −0.29 vs. 0.21 points/year of age; 95% CI −.39, −0.04; difference in linear age slope 0.22; 95% CI 0.02–0.41; P=0.03). Valine homozygotes also had lower risk for dementia (HR 0.28; 95% CI 0.10–0.85; P= 0.02) and AD (HR 0.31; 95% CI 0.10–0.95; P=0.03).

Conclusions

This preliminary report suggests that CETP V405 valine homozygosity is associated with slower memory decline and lower incident dementia/AD risk.

As the population ages, the public health and economic burdens of age-associated cognitive decline and dementia will continue to increase. Only APOE (ε4 isoform), has been conclusively associated with increased genetic susceptibility for the most common dementia, sporadic late-onset Alzheimer’s disease (AD).¹ Genes identified through associations with exceptional longevity are logical targets to explore for potentially beneficial associations with cognitive decline and dementia risk.²

Some studies have suggested that the APOE ε2 allele is associated with both increased lifespan and lower dementia risk.^{3, 4} Like APOE, the cholesteryl ester transfer protein gene (CETP) is involved in central nervous system cholesterol homeostasis and has been associated with exceptional longevity.^5–7 At codon 405 (exon 14), a functional single nucleotide polymorphism (SNP) substitutes valine (V) for isoleucine (I; ‘ancestral allele;’ NCBI dbSNP rs5882; V405) and is associated with lower CETP serum concentration and activity, and corresponding increases in HDL levels and lipoprotein (HDL/LDL) particle sizes.^{3, 4} These changes may have additional protective associations with cardiovascular disease, although the precise nature and potential mediators of these relationships are debated.^{3, 5}

CETP was identified as a “longevity gene” in a sample of Ashkenazi Jews by Barzilai and colleagues in 2003: valine homozygosity occurred in 24.8% of centenarians compared to 8.6% among controls. In a cross-sectional follow-up study, Ashkenazi centenarians with good cognitive function had increased frequency of the V405 polymorphism; in an independent cohort valine homozygosity was approximately five-fold higher in nondemented than demented Ashkenazi individuals aged 75–85.² Case-control studies in non-Ashkenazi populations have also reported protective associations between CETP SNPs and dementia prevalence, though replication has been inconsistent.^10–14

Herein, we investigated associations between V405 genotype and longitudinal memory performance and risk for incident dementia/AD in a community-based sample of healthy nondemented older adults. We hypothesized that the valine allele would be associated with less age-associated memory decline and lower incident dementia risk.

METHODS

Participants

The Einstein Aging Study (EAS) is a prospective cohort study designed to identify factors predicting cognitive decline and incident dementia in a racially- and ethnically-diverse community-dwelling population of elderly individuals. Study design and methods for recruitment and annual assessments have been previously described.^{6, 7} Briefly, potential participants were systematically recruited from population lists of Medicare recipients (1994–2004) or Bronx County registered voters (2004–2009). Eligible participants were aged 70 or older, Bronx residents, and had sufficient command of English to participate in neuropsychological testing. Exclusion criteria included audiovisual impairment severe enough to interfere with cognitive testing, inability to ambulate, and institutionalization. Written informed consent was obtained from each individual at study entry according to protocols approved by the institutional review board of the Albert Einstein College of Medicine. Annual in-person evaluations occurred at the EAS Aging Research Center (ARC) located on the Einstein campus.

Since 1994, the EAS has enrolled 1910 individuals; 1215 had blood samples and 608 had DNA extracted and genotyping performed. Availability of CETP and APOE genotypes determined preliminary eligibility for this investigation (n=608). Participants with dementia at baseline (n=15) were excluded from all analyses (n=593). Those without follow-up visits (n=70) were additionally excluded from the incident dementia analysis (n=523). Compared to 523 included individuals, the 70 excluded for lack of follow-up did not differ statistically in age (78.5 vs. 78.4 years, P=0.81), sex (61% vs. 63% female, P=0.76), race (69% vs. 67% Caucasian, P=0.75; 26% vs. 27% African-American, P=0.78), education (13.9 vs. 14.4 years, P=0.64), baseline cognition (BIMC 2.1 vs. 1.9, P=0.88), or genotype frequency for APOE ε4 (22% vs. 28%, P=0.34) or CETP (valine homozygotes 21% vs. 14% and valine heterozygotes 45% vs. 60%, P=0.06). Reasons for lack of follow-up included interval occurrence of an exclusion criterion after study entry (n=3), moved (n=4), became too ill to continue (n=6), loss of contact (n=6), less than a year as a study participant (n=7), death (n=12), and declined/unavailable to continue (n=32).

Of the 607 individuals with blood samples but no genotype, some newly-recruited individuals have had only their baseline visit and consequently no follow-up data is yet available for them. We plan to make DNA on these individuals and reassess our results as follow-up time accumulates. In addition, for about 150 participants long-banked buffy coat did not yield high quality DNA and so these individuals were excluded from the current analysis, as were a small number of individuals with prevalent dementia. In the total EAS sample, statistical comparison between individuals with and without genotyping revealed that the two groups were similar in sex and racial distribution, but those with genotyping were significantly younger, better educated and had better global cognitive function at baseline.

Demographic and Clinical Information

Trained research assistants used structured questionnaires to obtain sociodemographic information (age, sex, race, Ashkenazi heritage, and years of education) and medical history at each annual visit. We obtained information on race necessary to explore potential confounding by genetic admixture. Participants self-identified their race and religious background to the research assistants, who asked “to which race group do you belong?” and “What is your religious preference?” Response options for race were Caucasian, Black, Hispanic/White, Hispanic/Black, Asian, or other. For religious background, options were Protestant, Catholic, Jewish, or other, or individuals could decline to endorse a preference. Using baseline medical history, we tabulated a Medical Comorbidity Index score (range 0–10) from dichotomous self-report (present vs. absent) of hypertension, diabetes, angina, myocardial infarction, congestive heart failure, stroke, Parkinson’s disease, rheumatoid arthritis, chronic obstructive pulmonary disease, and depression.⁸ Depressive symptoms were assessed using the 15-item Geriatric Depression Scale (GDS).⁹ Functional status was assessed using the Lawton-Brody scale of instrumental activities of daily living (range 0–8; higher scores better).¹⁰

Neuropsychological Assessment

At baseline and at each annual evaluation, cognitive status was assessed by a comprehensive neuropsychological battery comprising tests of global cognition and the specific cognitive domains of attention, episodic and visual memory, executive function, language, and visuoconstructional ability.¹¹ Administration by trained neuropsychological assistants was standardized according to EAS protocols based on published guidelines. For this study, we report pre-morbid intelligence, global cognitive status, and results from representative domain-specific cognitive tests of attention, psychomotor speed, and episodic memory. The Vocabulary subtest from the Wechsler Adult Intelligence Scale (WAIS-R) served as a ‘hold’ test, estimating crystallized intelligence less vulnerable to the effects of aging and not expected to decline as a consequence of dementia.¹² Global cognition was assessed via the Blessed-Information-Memory-Concentration test (BIMC), which correlates well with AD neuropathology.^{13, 14} Domain-specific cognitive tests were the Digit Span and Digit-Symbol-Substitution WAIS-R¹² subtests, and “free recall” from the Free and Cued Selective Reminding Test (FCSRT).^{15, 16} We selected the FCSRT to test our a priori hypothesis of a protective association between V405 genotype and episodic memory based on its demonstrated operating characteristics.^17–19 Digit Span was used to assess attention, unlikely to decline as a result of dementia and, when impaired, more commonly a consequence of cerebrovascular disease.²⁰ Digit-Symbol-Substitution was chosen as a test of psychomotor processing speed, which may exhibit age-associated decline but is less sensitive to dementia.²¹

Dementia Diagnosis

The EAS neuropsychologist and neurologist met monthly in diagnostic consensus case conferences that included review of all available clinical and neuropsychological data.²² Dementia diagnoses were based on standardized clinical criteria from the Diagnostic and Statistical Manual, Fourth Edition (DSM-IV)²³ and required impairment in memory plus at least one additional cognitive domain, accompanied by evidence of functional decline. AD was diagnosed in demented participants meeting clinical criteria for probable or possible disease established by the National Institute of Neurological and Communication Disorders and Stroke and the Alzheimer Disease and Related Disorders Association (NINCDS-ADRDA).²⁴

Genotyping

All blood samples for genotyping were obtained with informed consent. The phlebotomist drew 20cc of whole blood at the ARC; samples were centrifuged and stored in aliquots in a −70C freezer. Subsequently, DNA was isolated directly from buffy coat using the Puregene DNA Purification System (Gentra System, Minnesota). Amplification and sequencing primers for genotyping of the target V405 SNP (rs5882) and for the two APOE SNPs, rs429358 (position 112) and rs7412 (position 158), were designed using PSQ version 1.0.6 software (Biotage); in each case the reverse primer was biotinylated. Genotyping was performed using a Pyrosequencing PSQ HS 96A system (http://www.pyrosequencing.com) according to manufacturer’s instructions.

Statistical Analysis

We classified participants dichotomously based on whether incident dementia occurred during follow-up. Time of dementia diagnosis was assigned the visit date immediately preceding the consensus conference making the diagnosis. V405 genotype was dichotomized into homozygote (V-V) and heterozygote (I–V) groups; isoleucine homozygotes served as the reference group in all longitudinal analyses. Baseline demographic and clinical characteristics were compared using either the Wilcoxon Rank Sum or Kruskal-Wallis tests for continuous variables; Chi-square or Fischer’s exact test were used as appropriate for categorical variables.

Associations between V405 genotype and longitudinal performance in episodic memory, attention, and psychomotor speed were examined with linear mixed effects models using age (centered at 85) as the time scale and adjusted for sex, education, race/ethnicity, and presence of an APOE ε4 allele. Indicator variables coded for African-American race and Ashkenazi heritage. V405 genotype groups were separately compared with the isoleucine reference group. The main outcome of interest was repeated measures of individual free recall scores on the FCSRT. For each V405 group the B coefficient for the difference in linear age slope estimated the change in cognitive score per additional year of age, compared to the reference group. A quadratic age term tested whether an accelerated non-linear trend was present. Performance on tests of attention and psychomotor speed were similarly modeled.

To estimate the risk of incident dementia and AD as a function of V405 group, we employed nested Cox proportional hazard models with delayed entry and age as the time scale.³³ In cohort studies, age is preferred to follow-up time as the time scale because the hazard function can be directly interpreted as the age-specific incidence function for dementia; inclusion of age as a non-parametric term in this manner provides a more flexible and effective control than treating age as a covariate.²⁵ Hazard ratios were estimated using V405 groups as categorical predictor variables, with isoleucine homozygotes as the reference. Time-to-event was the interval between age at baseline and age at dementia/AD diagnosis or last contact, as appropriate. In Model 1 we adjusted for sex, race/ethnicity (coded as above), and years of education. Model 2 included the covariates in Model 1 plus an additional adjustment for the medical comorbidity index. Model 3 included all covariates from Models 1 and 2, plus an additional adjustment for presence of an APOE ε4 allele. To assess potentially-competing cardiovascular morbidity, we reran the final model substituting combined history of hypertension, myocardial infarction, and stroke for the comorbidity index. All covariates were pre-selected based on biological plausibility and/or potential for confounding and were used as time-constant covariates. Proportional hazards assumptions of the models were examined using methods based on scaled Schoenfeld residuals and were adequately met.²⁶

To determine whether missing data on genotype information biased our results, we performed a multiple imputation analysis that gave unbiased estimates under a missing-at-random (MAR) missing data mechanism, allowing the missingness of CETP data to depend on incident dementia and covariates but not on unobserved genotype.

We performed a posthoc calculation of the study’s power to detect incident dementia and memory decline. Given 21% prevalence for valine homozygosity, the sample of 523 participants with 40 incident dementia cases provided 75% power to detect a hazard ratio of 0.3 between valine homozygotes and the isoleucine reference group. For smaller effect sizes, such as hazard ratios closer to one, power would be lower. For memory decline, the data provide 57% power to detect a difference of 0.2 points/year in linear age slope on FCSRT.

The statistical software packages SAS version 9.1 (SAS Institute Inc., Cary, NC) and S-Plus 8.0 (Insightful Corp.) were used for all analyses. Two-sided probability values less than 0.05 were considered statistically-significant in all tests, including Fisher’s exact test.

RESULTS

Demographic Characteristics

In the study sample of 523 individuals, 40 incident cases of dementia occurred (Table 1). Frequency of the APOE-ε4 allele (23%) was similar to other racially- and ethnically-diverse cohorts in U.S. urban centers.^{27, 28} At first evaluation, compared with individuals who remained nondemented throughout follow-up, those who developed dementia were older, less educated, and had poorer performance in global cognition, psychomotor speed and episodic memory. The groups did not differ by sex, race, Ashkenazi heritage, CETP/APOE genotype or medical comorbidity burden.

Table 1.

Baseline characteristics by dementia status at follow-up

Variable	All Individuals (N=523)	Individuals with no Incident Dementia (N=483)	Individuals with Incident Dementia Subjects (N=40)	P value
≥ 1 CETP Valine allele, n (%)	345 (66)	324 (67)	21 (53)	0.06
≥ 1 APOE ε4 allele, n (%)	117 (23)	104 (22)	13 (33)	0.11
Age, median (IQR), years [range]	78.2 (8.1) [65.3–96.2]	77.3 (8.3) [66–95]	79.5 (6.9) [70.9–96.2]	0.02
Female sex, n (%)	319 (61.0)	296 (61.6)	23 (57.5)	0.64
Caucasian, n (%)	361 (69.0)	337 (69.8)	24 (60.0)	0.20
African American, n (%)	134 (25.6)	121 (25.0)	13 (32.5)	0.30
Other Race, n (%)	28 (5.4)	25 (5.2)	3 (7.5)	0.47
Ashkenazi Jewish, n (%)	157 (30.0)	149 (30.9)	8 (20.0)	0.15
Education, median (IQR), years [range]	14.0 (4.0) [3–22]	14.0 (4.0) [3–22]	12.0 (4.5) [4–20]	0.02
At Least Skilled or Professional Prior Occupational Attainment,a n (%)	459 (88)	428 (89)	31 (78)	0.12
Follow-up time, median (IQR), years [range]	3.2 (4.9) [0.9–15.5]	3.2 (4.9) [0.9–15.5]	3.1 (5.5) [1.0–10.7]	0.94
Medical Comorbidity Index, self-report, median (IQR) [range]	1.0 (1) [0–6]	1.0 (1) [0–6]	1.0 (1) [0–5]	0.50
Instrumental Activities of Daily Living, self-report, median (IQR) [range]b	7 (3) [1–6]	7 (3) [1–8]	6 (3) [4–8]	0.21
Geriatric Depression Scale, median (IQR), [range] c	2 (2) [0–10]	2 (2) [0–10]	2 (2) [0–9]	0.55
Blessed IMC test, median (IQR), [range] d	1 (3) [0–10]	1 (3) [0–8]	4.5 (4) [0–10]	<0.001
WAIS-R Vocabulary raw score, median (IQR), [range] e	48 (18) [9–69]	48 (18) [9–69]	42 (28.5) [15–69]	0.16
WAIS-R Digit Span raw score, median (IQR), [range]f	14 (4) [6–27]	14 (4) [6–27]	12.5 (3.5) [7–27]	0.02
WAIS-R Digit Symbol Substitution raw score, median (IQR), [range] g	42 (18) [1–85]	42 (17) [1–85]	31 (19) [13–61]	<0.001
Free and Cued Selective Reminding median (IQR), [range] h	31 (7) [11–31]	32 (7) [14–43]	25 (7) [11–41]	<0.001

Abbreviations: CETP, cholesteryl ester transfer protein; APOE, apolipoprotein E (ε4 isoform); IQR, interquartile range; Blessed I-M-C, Blessed Information-Memory-Concentration test; WAIS-R, Wechsler Adult Intelligence Scale-Revised.

Values are median (interquartile range unless otherwise noted). Percentages might not equal 100% due to rounding. P-values are from the Wilcoxon rank-sum test for continuous variables; Chi-square or Fischer’s Exact test were used as appropriate for categorical variables. Group comparisons were between nondemented and incident dementia groups.

^aStandard Occupational Classification Manual scale, US Department of Commerce, range 0–9, higher scores better; proportion of sample in highest two tertiles of occupational attainment scores shown.

^bLawton-Brody scale, range 0–8, higher scores better

^c15-item; >6 indicates significant depressive symptoms

^dPossible score range 0–33 higher scores worse; ≥8 considered impaired

^ePossible theoretical raw score range 0–66; mean ± 2SD range in 80–84 year-olds 9–60

^fPossible theoretical raw score range 0–30; mean ± 2SD range in 80–84 year-olds 5–21

^gPossible theoretical raw score range 0–133; mean ± 2SD range in 80–84 year-olds 11–74

^hPossible score range 0–48; higher scores better; ≤24 considered memory impaired

In the study sample of 523 participants, allele frequency for valine was 43.5%: 235 (44%) heterozygotes, 110 (21%) valine homozygotes; 178 (34%) were isoleucine homozygotes (34%). Genotype frequencies differed marginally from Hardy-Weinberg equilibrium (χ²=3.86,1 df=1, P=0.05), possibly consistent with a longevity effect on allele distribution.^{29, 30} Demographic characteristics and baseline neuropsychological test results among the three genotype groups were similar except for pre-morbid intelligence and race (Table 2). Compared to 37% in Caucasians, valine frequency was 60% in African Americans, similar to frequencies previously reported elsewhere.³¹ Median age was similar for African-American (77.2 years [IQR 7.6]) and non-African-American participants (78.3 [IQR 8.2]. There were more women in the African-American group (79% vs. 55%, P<0.0001), and African Americans had less education (median [IQR; range] 12.5 [4.0; 3–20] years vs. 14.0 [4.0; 4–22] years, P=0.001), worse baseline BIMC scores (median [IQR; range] 3.0 [4.0; 0–8] vs. 1.0 [3.0; 1–10], P<0.0001) and a higher percentage of individuals with an APOE ε4 allele (31% vs. 23%, P=0.008) compared to non-African-Americans. Baseline medical comorbidity index scores were similar among African-American (median [IQR; range] 1 [1–6]) and non-African-American participants (1 [1–5]). African Americans had greater frequency of self-reported hypertension compared to non-African-Americans (n=93, 69%; vs. n=224, 58%; P=0.02), and lower frequency of myocardial infarction (n=4, 3%; vs. 39, 10%; P=0.01). There was no statistically-significant difference in self-reported stroke (n=13, 10%; vs. n=32, 9%; P=0.60). Crude dementia incidence rates among African Americans (2.04/100 person-years) were higher than in non-African-Americans (1.69/100 person-years) but the difference was not statistically significant (P=0.18).

Table 2.

Characteristics of All Individuals at First Evaluation, Stratified by Cholesteryl Ester Transfer Protein (CETP) V405 Genotype

Variable	CETP I-I (N=178)	CETP I–V (N=235)	CETP V-V (N=110)	P value
Crude Incidence of Dementia, per 100 person-years (95% CI)	2.48 (1.6–4.0)	1.56 (0.8–2.5)	1.05 (0.3–2.5)	0.12
≥ 1 APOE ε4 allele, n (%)	43 (23)	47 (20)	27 (25)	0.49
Age, median (IQR), years [range]	78.3 (8.1) [70.4–96.2]	77.7 (8.2) [66.2–91.7]	77.7 (9.0) [69.7–94.5]	0.81
Female sex, n (%)	106 (59.6)	141 (60.0)	72 (65.5)	0.56
Caucasian, n (%)	151 (84.8)	158 (67.2)	52 (47.3)	<0.001
African-American, n (%)	21 (11.8)	64 (27.2)	49 (44.6)	<0.001
Other Race, n (%)	6 (3.4)	13 (5.5)	9 (8.2)	0.21
Ashkenazi Jewish, n (%)	53 (29.8)	77 (32.8)	27 (24.6)	0.30
Education, median (IQR), years [range]	14 (4) [4–20]	13 (4) [3–22]	14 (4) [5–21]	0.14
At Least Skilled or Professional Prior Occupational Attainment,a n (%)	161 (90)	205 (87)	93 (83)	0.50
Follow-up time, median (IQR), years [range]	3.1 (4.1) [0.9–15.5]	3.2 (5.0) [0.9–12.7]	3.1 (5.1) [0.9–14.6]	0.63
Medical Comorbidity Index, self-report, median (IQR) [range]	1 (2) [2–5]	1 (1) [0–6]	1 (1) [0–5]	0.89
Instrumental Activities of Daily Living, self-report, median (IQR) [range] b	7 (2) [0–8]	7 (3) [1–8]	8 (3) [1–8]	0.43
Geriatric Depression Scale, median (IQR), [range] c	2 (2) [0–10]	2 (2) [0–10]	2 (2) [0–10]	0.80
Blessed IMC test, median (IQR), [range] d	1 (2) [0–10]	1 (2) [0–10]	2 (3) [0–8]	0.39
WAIS-R Vocabulary raw score, median (IQR), [range] e	48 (16) [12–69]	48 (16) [9–69]	45.5 (18) [15–69]	0.04
WAIS-R Digit Span raw score, median (IQR), [range] f	14 (3) [6, 27]	14 (4) [6–27]	13 (4) [7–25]	0.46
WAIS-R Digit Symbol Substitution raw score, median (IQR), [range] g	42 (16) [16–20]	41 (17) [10–81]	43 (20) [5–85]	0.57
Free and Cued Selective Reminding,e median (IQR), [range] h	32 (7) [11–42]	31 (8) [15–43]	31 (7) [16–41]	0.12

Abbreviations: CETP, cholesteryl ester transfer protein; APOE, apolipoprotein E (ε4 isoform); IQR, interquartile range; Blessed I-M-C, Blessed Information-Memory-Concentration test; WAIS-R, Wechsler Adult Intelligence Scale-Revised.

Values are median (interquartile range unless otherwise noted). Percentages might not equal 100% due to rounding. P-values are from comparison of medians among the three genotype groups for continuous variables; Chi-square or Fischer’s Exact test were used as appropriate for categorical variables.

^aStandard Occupational Classification Manual scale, US Department of Commerce, range 0–9, higher scores better; proportion of sample in highest two tertiles of occupational attainment scores shown.

^bLawton-Brody scale, range 0–8, higher scores better

^c15-item; >6 indicates significant depressive symptoms

^dPossible score range 0–33 higher scores worse; ≥8 considered impaired

^ePossible theoretical raw score range 0–66; mean ± 2SD range in 80–84 year-olds 9–60

^fPossible theoretical raw score range 0–30; mean ± 2SD range in 80–84 year-olds 5–21

^gPossible theoretical raw score range 0–133; mean ± 2SD range in 80–84 year-olds 11–74

^hPossible score range 0–48; higher scores better; ≤24 considered memory impaired

Associations between CETP Genotype and Longitudinal Cognitive Performance

Baseline performance on the domain-specific cognitive tests of FCSRT, Digit Span, and Digit-Symbol-Substitution was similar among the three genotype groups (Table 2). In the linear mixed effects model (Table 3), after adjustments for sex, race, education, medical comorbidities, and APOE status, valine homozygosity was independently associated with slower decline on the FCSRT. A quadratic trend for age was found for cognitive decline, but was common for homo-and heterozygotes. Compared to an absolute linear age slope of 0.43 points/year of age (95% CI −0.58, −0.29) decline in FCSRT score in the reference group, the linear age slope in valine homozygotes was 0.21 points/year of age (95% CI −0.39, 0.04). Based on the difference in linear age slopes, valine homozygosity was independently associated with 0.21 points/year of age less decline on FCSRT, slower by 51%. Memory decline in valine heterozygotes was similar to that of the reference group. V405 genotype was not associated with cognitive performance on domain-specific tests of Digit Span or Digit-Symbol-Substitution.

Table 3.

Estimated Mean Change in Free and Cued Selective Reminding Score with Increasing Age, According to CETP V405 genotype group

	FCSRT		WAIS-R Digit Span		WAIS-R Digit Symbol
	B (95% CI)	P Value	B (95% CI)	P Value	B (95% CI)	P Value
Quadratic age trend	−0.02 (−0.03, −0.01)	<.0001	−0.005 (−0.009, −0.002)	0.002	−0.01 (−0.02, 0.001)	0.07
Linear age slope (per year) in reference group	−0.43 (−0.58, −0.29)	<.0001	0.05 (−0.01, 0.11)	0.10	−0.93 (−1.11, −0.75)	<.0001
Difference in Linear age slope between Valine Heterozygotes and Reference Group	0.07 (−0.09, 0.23)	0.39	0.01 (−0.05, 0.08)	0.67	0.06 (−0.11, 0.27)	0.42
Difference in Linear age slope between Valine Homozygotes and reference group	0.22 (0.02, 0.41)	0.03	−0.03 (−0.10, 0.06)	0.55	0.10 (−0.14, 0.36)	0.40

Abbreviations: CI, Confidence Interval; FCSRT, Free and Cued Selective Reminding Test; WAIS-R, Wechsler Adult Intelligence Scale-Revised.

Linear mixed effect model: Isoleucine homozygotes (I-I) were used as the reference group for each cognitive test. The model was centered at 85 and adjusted for sex, race, years of education, medical comorbidities, and presence of an APOE ε4 allele. The quadratic term evaluates for presence of accelerated non-linear decline. All parameter estimates are unstandardized regression coefficients. The regression coefficients for the difference in linear age slope estimates the change in cognitive score per additional year of age in each V405 genotype group compared to the reference group.

Associations between CETP Genotype and Incidence of Dementia and Alzheimer’s Disease

Of the 40 individuals with incident dementia, 35 met criteria for probable or possible Alzheimer’s disease. Based on information in Table 2, the absolute attributable rate of incident dementia (difference in incidence rates) in valine compared to isoleucine homozygotes was −1.43 cases/100 person-years. In the fully-adjusted models (Table 4; Figures 1A–B), valine homozygosity was associated with lower risk for developing both dementia and AD. The magnitude of the hazard ratios for heterozygotes were also substantially less than one for both dementia and AD, but the upper boundaries of the confidence intervals included one. Substituting combined history of hypertension, myocardial infarction, and stroke for the comorbidity index in models for dementia slightly improved the magnitude of protective associations (valine homozygotes: HR 0.26, 95%CI 0.09–78; P=0.02; heterozygotes: HR 0.54, 95%CI 0.27–1.1, P=0.09).

Table 4.

CETP V405 Genotype and Risk for Dementia and Alzheimer’s Disease^a

Category	Model 1b		Model 2c		Model 3d
	HR (95% CI)	P Value	HR (95% CI)	P Value	HR (95% CI)	P Value
Risk for Dementia Comparing Valine Heterozygotes, Carriers and Homozygotes to Isoleucine Homozygotes
Valine Heterozygotes n=16	0.52 (0.26 – 1.06)	0.07	0.53 (0.26 – 1.09)	0.08	0.57 (0.28 – 1.15)	0.12
Valine Homozygotes n=5	0.29 (0.10 –0.85)	0.02	0.28 (0.09 – 0.84)	0.02	0.28 (0.10 – 0.85)	0.02
Risk for Alzheimer’s Disease Comparing Valine Heterozygotes, Carriers and Homozygotes to Isoleucine Homozygotes
Valine Heterozygotes n=14	0.52 (0.24 – 1.13)	0.10	0.53 (0.25 – 1.2)	0.11	0.56 (0.26 –1.2)	0.14
Valine Homozygotes n=5	0.31 (0.10 – 0.96)	0.04	0.30 (0.10 – 0.94)	0.04	0.31 (0.10 – 0.95)	0.04

Abbreviations: CI, Confidence Interval; HR, Hazard Ratio.

^aP values from Cox proportional hazard models with delayed entry and age as the time scale. There were 40 incident cases of dementia and 35 incident cases of Alzheimer’s disease. There were 19 individuals in the reference group for dementia and 16 individuals in the reference group for Alzheimer’s disease.

^bAdjusted for sex, years of education, non-Ashkenazi white race, black race.

^cAdjusted for the covariates in Model 1 plus an additional adjustment for medical comorbidities as measured by the Medical Comorbidity Index.

^dAdjusted for the covariates in Model 2 plus an additional adjustment for presence of an APOE ε4 allele.

Figure 1.

Figure 1A. Dementia Incidence by CETP Genotype Group

Survival curves are based on the Cox Proportional Hazards regression analysis.

Figure 1B. Alzheimer’s Disease (AD) Incidence by CETP Genotype Group

Survival curves are based on the Cox Proportional Hazards regression analysis.

To address the possibility that racial admixture was confounding our observations, we reran the linear mixed effects and the final Cox proportional hazards models (minus the indicator variable for African-American race) in a Caucasian-only subgroup (n=361 individuals, 24 incident dementia cases). No association between CETP and memory decline was detected. In the Cox model, the magnitude of the association lessened slightly compared to main results and the model lost statistical significance (homozygote HR 0.45, 95% CI 0.10–2.03, P=0.30; heterozygote HR 0.56, 95% CI 0.23–1.38, P=0.21), possibily a consequence of lower power in this smaller group.

Because genotype results were not available for all individuals with blood samples, we compared dementia-free survival in those with and without genotype information. Adjusted for sex, education, race, and baseline BIMC, the genotyped individuals survived longer without dementia than those without genotype (HR 0.45; P<0.0001), suggesting nonrandom missingness of genotype. In the multiple imputation analysis performed to allay the resulting concerns about the representativeness of our sample and the generalizability of our results, the HR for valine homozygotes was 0.44 (95% CI 0.20–0.96; P=0.04), similar to results observed in the primary analysis.

COMMENT

Compared to a reference group homozygous for the ancestral allele in the full sample, valine homozygosity was independently associated with slower age-associated memory decline and lower risk for incident dementia and Alzheimer’s disease. Main findings were statistically significant in adjusted models, including APOE. In valine homozygotes, memory declined 51% slower than in the reference group; homozygosity was associated with lower risk for both dementia (72%) and Alzheimer’s disease (69%). Risk for dementia and Alzheimer’s disease was nonsignificantly lower in valine heterozygotes compared to individuals homozygous for isoleucine, although the magnitude of the association was less than that observed in homozygotes, suggesting a gene-dose relationship.

In the cognitive decline analysis, our primary hypothesis was that the valine allele would be associated with slower episodic memory decline. To our knowledge, only one study has evaluated the CETP V405 polymorphism in the context of cognitive change. Johnson et al examined V405 associations with childhood IQ and life-long change in global cognition in a group of Scottish older adults.³² Specific cognitive domains, including memory, were examined cross-sectionally at age 79. They reported that CETP was associated with cardiovascular disease rate, but detected no significant global longitudinal or domain-specific cross-sectional cognitive associations. In the present report, valine homozygosity was not associated with cross-sectional cognitive function, but protected against memory decline. Thus, the longitudinal design of our study may account for the different results we found. Population differences may also contribute, as valine was at considerably higher frequency (69%) than in our sample.

To our knowledge, this is the first study to examine associations between CETP genotype and dementia risk in an incidence study. Of 10 case-control studies published since 2004, three investigated the V405 polymorphism. The single population-based case-control study reported that the V405 polymorphism was associated with an elevated odds ratio (OR) of 1.67 for AD in non-APOE ε4 carriers.³³ A 2005 Spanish study reported that ε4 carriers homozygous for the minor allele of the C-629 A CETP polymorphism had a lower OR (2.33) for AD than ε4 non-carriers (OR 7.12) but found no association between V405 and AD.³⁴ In 2009, Qureischie and colleagues reported lack of association between V405 and AD, although a CETP haplotype including valine was associated with cholesterol profiles in plasma and CSF.³⁵ Studies investigating other CETP polymorphisms reveal similar heterogeneous results.

Methodological differences among studies may contribute to the disparate results presented here. We suggest that selection bias in case-control studies may be particularly problematic for polymorphisms associated with longevity. Case-control studies may be biased by factors associated with duration of disease in prevalent cases with longevity-favoring genotypes, as they preferentially include dementia cases with longer survival. Based on enhanced survival, such individuals might be over-represented among cases, spuriously attenuating any protective associations. For this reason, case-control studies likely underestimate associations between longevity genes and prevalent dementia. Because we focused on incident cases, our cohort study may have circumvented this potential bias. Our results suggest that prospective studies may offer methodological advantages in assessing associations between longevity genes and risk for incident dementia.

Two recent genome-wide association studies comparing AD cases to non-AD controls failed to identify the CETP V405 polymorphism.^{36, 37} This is less surprising than it might appear: in general, GWAS look for polymorphisms with elevated odds ratios, indicating a strong associated with increased risk of disease. Any effect associated with V405 would be protective, with an odds ratio significantly below one and therefore likely to be dismissed as insignificant by current GWAS practices and making it difficult to compare our results with those from traditional dementia GWAS studies.

We detected no bivariate association between Ashkenazi heritage and genotype but race was associated with genotype: though comparable to previously published reports, prevalence of valine was higher in African Americans than Caucasians. African-American and non-African-American participants were similar in median age and crude dementia incidence rates, but African Americans had disadvantaged profiles for several known indices of dementia risk, including education, baseline global cognition, and APOE ε4 frequency. These results might be interpreted as relative protection insofar as there appeared to be higher risk without higher incidence. Conversely, they could be paradoxical, given increased valine allele frequency in African Americans without lower incidence, possibly indicating a genotype-race interaction we were not powered to detect. Alternatively, African Americans may possess a third factor that increases dementia susceptibility despite higher population frequencies of the putatively protective allele. We attempted to account for the potential effects of race by including it as a confounder in our models. After excluding African Americans in the Cox models for dementia, the magnitude of the resulting association changed only slightly, although the upper boundary of the confidence interval included one. The low number of incident dementia cases prevented exploration of interactions in race-stratified analyses. These observations merit further investigation.

Our study cannot directly address the nature of causal relationships underlying the associations reported here, limiting us to informed speculation. Nonetheless, several of the criteria for causation are satisfied. While it remains unknown whether CETP polymorphisms cause specific differential rates of memory decline or dementia incidence, the basic temporal sequence of the genetic factor preceding downstream effects is indisputable. The gene-dose character of the association between V405 genotype and memory decline and dementia in our sample also would seem to favor causality. CETP is synthesized³⁸ and expressed³⁹ in brain but sparse knowledge limits speculation about CETP’s role in neurodegenerative pathophysiological mechanisms that might influence dementia risk. Growing evidence posits a mechanistic pathway linking cerebral cholesterol metabolism and AD pathology; hypothetical descriptions include genetic susceptibility conferred by cholesterol-metabolism genes such as CETP and APOE.^{5, 41} Future studies investigating intermediate steps in the putative causal pathway – mediation by endophenotypic biomarkers, such as CETP plasma levels and activity, for example – might provide useful first steps toward eventual resolution of the causality question.

There are several caveats to our results. Though APOE frequency in our sample was comparable to other aging cohorts, our participants were nonetheless community-residing relatively healthy older adults living in the Bronx and we acknowledge the need to evaluate CETP in other longitudinal studies with greater numbers of incident dementia cases. We used well-established procedures and standardized criteria to diagnose dementia/AD but some misclassification may have occurred. Like any longitudinal study, ours may have had selective attrition, although the duration of follow-up helps allay this concern. At 40, the number of incident dementia cases was small.

Despite the small number of incident dementia cases, this preliminary report suggests that CETP valine homozygosity is associated with slower memory decline and lower risk for incident dementia or Alzheimer’s disease. This potentially-protective association is supported by several observations. First, prior work has shown that at cross-section valine homozygosity was associated with better mental status.² Second, valine homozygosity is associated with a slower rate of memory decline in our entire sample, not just those who developed dementia. Third, in this sample the influence of CETP on risk for dementia and AD is gene-dose dependent. Finally, an association between CETP and cognition and dementia is biologically plausible because other genes involved in lipid metabolism, including APOE, are associated with dementia risk.