FOXO3 longevity genotype attenuates the impact of hypertension on cerebral microinfarct risk

Abstract

Objective:

The G-allele of FOXO3 SNP rs2802292, which is associated with human resilience and longevity, has been shown to attenuate the impact of hypertension on the risk of intracerebral haemorrhage. We sought to determine whether the FOXO3 G-allele similarly attenuates the impact of hypertension on the risk of cerebral microinfarcts (CMI).

Methods:

From a prospective population-based cohort of American men of Japanese ancestry from the Kuakini Honolulu Heart Program (KHHP) and Kuakini Honolulu-Asia Aging Study (KHAAS) that had brain autopsy data, age-adjusted prevalence of any CMI on brain autopsy was assessed. Logistic regression models, adjusted for age at death, cardiovascular risk factors, FOXO3 and APOE-ε4 genotypes, were utilized to determine the predictors of any CMI. Interaction of FOXO3 genotype and hypertension was analyzed.

Results:

Among 809 men with complete data, 511 (63.2%) participants had evidence of CMI. A full multivariable model demonstrated that body mass index (OR:1.07, 95% CI=1.01–1.14, P = 0.015) was the only predictor of CMI, while hypertension was a borderline predictor (OR:1.44, 95% CI=1.00–2.08, = 0.052). However, a significant interaction between FOXO3 G-allele carriage and hypertension was observed (P = 0.020). In the stratified analyses, among the participants without the longevity-associated FOXO3 G-allele, hypertension was a strong predictor of CMI (OR:2.25, 95% CI=1.34–3.77, P = 0.002), while among those with the longevity-associated FOXO3 G-allele, hypertension was not a predictor of CMI (OR:0.88, 95% CI: 0.51–1.54, P = 0.66).

Conclusions:

The longevity-associated FOXO3 G-allele mitigates the impact of hypertension on the risk of CMI.

INTRODUCTION

Cerebral microinfarcts (CMIs) are microscopic vascular lesions of the brain that are often “invisible” to the conventional structural neuroimaging studies. CMIs are detected mainly by histologic examination [1] with a mean diameter of approximately 0.2 mm [2] and characterized as regions of cellular death or tissue necrosis [3–5]. Prevalence of one or more detectable CMIs ranges from 16% to 46% among elderly populations who died from all causes [1], and CMI severity is associated with the development of dementia [6,7]. Although CMIs share similar histopathological characteristics of macroscopic infarcts, the pathophysiology of CMIs is not understood completely. Epidemiological studies have identified hypertension as an important risk factor for CMI [8,9]. How other genetic or environmental factors modulate the association between hypertension and CMI, however, has not been adequately studied.

Minor alleles of multiple single nucleotide polymorphisms (SNPs) located in the forkhead winged-helix box O type 3 gene (FOXO3) – particularly the G allele of SNP rs2802292 – have been strongly associated with human longevity [10–12]. In the Kuakini Honolulu Heart Program (KHHP) cohort, carriers of the longevity-associated FOXO3 G-allele lived longer than subjects who were homozygous for the major (T) allele [10]. We subsequently showed that G-allele carriage was associated with protection against coronary artery disease (CAD)-associated mortality [13]. Another study also demonstrated the protective effect of FOXO3 G-allele against heart disease among older Japanese men [14]. Although the exact mechanism by which FOXO3 genotype is associated with healthy aging and increased lifespan remains to be fully clarified, it has been postulated that the longevity-associated FOXO3 G-allele may serve as a “resilience” gene by mitigating the adverse effects of chronic cardiometabolic stress on intracellular processes, thereby reducing the risk of life-threatening cardiovascular events [12,15].

More recent studies using the KHHP cohort demonstrated that the longevity-associated FOXO3 G-allele attenuates the deleterious impact of chronic hypertension on the long-term risk of spontaneous intracerebral haemorrhage (ICH) by possibly providing cerebrovascular resilience and protection against the adverse effects of chronic hypertension [16], and that FOXO3 G-allele carriage was associated with protection against Alzheimer’s disease in men with late-life hypertension [17]. Although FOXO3 has never been identified as a stroke-specific gene, its encoded protein, the transcription factor FOXO3, may nevertheless protect against CMIs by activating intracellular pathways in vessel walls so as to reduce the cumulative burden of hypertension on the cerebral vasculature. Assessing gene-environment interaction can improve understanding of how a known genotype (i.e., FOXO3 genotype) may modulate the end-organ manifestation of chronic cardiovascular diseases [18]. To date, the impact of the longevity-associated FOXO3 G-allele on hypertension-related risk of CMI has not been studied. Therefore, we sought to assess whether FOXO3 G-allele carriage protects against CMIs by mitigating the deleterious impact of hypertension on cerebral arterioles.

METHODS

Study Design and Subjects

The KHHP is a prospective population-based study of cardiovascular disease among American men of Japanese ancestry living in Hawaii. Starting in the mid-1960s, the KHHP began following 8,006 men of Japanese ancestry living on the island of Oahu for the development of CAD and stroke [19–21]. Participants were identified using World War II Selective Service Registration files. They were 45 to 68 years old at baseline examination after recruitment between 1965 and 1968. They have been followed since then with periodic examinations, and continuous hospital surveillance for selected morbidity and all mortality through December 2000. Details of the selection process for the cohort have been published [21]. Morbidity and mortality have been assessed since the beginning of the study by monitoring hospital discharge records, death certificates, and local obituaries. Data collection is believed to be complete for all-cause mortality. Attrition in this cohort was very small – at the fourth examination only five men were lost to follow-up. The Kuakini Honolulu-Asia Aging Study (KHAAS) was a continuation of the KHHP that started in 1991 with the fourth examination of the KHHP and focused on cognitive disorders and dementia. All KHAAS participants were eligible for inclusion in the autopsy sample. Consent for each autopsy was provided by a next-of-kin family member or a legally authorized representative. Procedures performed were in accord with institutional guidelines and were approved by the Institutional Review Board of Kuakini Medical Center. Written informed consent was also obtained at all examination cycles. The study adhered to the observational cohort guideline.

Data Collection

Data on cardiovascular risk factors were obtained at the KHAAS baseline examination (examination 4 in 1991–1993). Hypertension at baseline was defined by systolic/diastolic blood pressure (SBP/DBP) of ≥140/90 mmHg or the self-reported use of anti-hypertensive medications. Body mass index (BMI) was defined as weight in kilograms divided by height in meters squared. Diabetes was defined by modified American Diabetes Association (ADA) criteria, as fasting glucose/two-hour oral glucose tolerance test result of ≥126/200 mg/dl or reported use of insulin or oral hypoglycemic medications. Smoking was defined as pack-years by self-report. Physical activity index (PAI) was quantified as metabolic output during a typical 24-hour period by multiplying a weighting factor by the number of hours spent in 5 activity levels (no activity=1.0, sedentary=1.1, slight=1.5, moderate=2.4, and heavy=5.0) [22]. Serum cholesterol and high-density lipoprotein (HDL) were measured in fasting blood samples. Alcohol intake was measured by self-report as ounces per month. Cognitive function was measured by the 100-point Cognitive Abilities Screening Instrument (CASI) [23]. Age at death was used for the analyses. Controlled hypertension was defined as a diagnosis of baseline hypertension on Examination 4 and a measurement of SBP ≤ 140 mmHg and DBP ≤ 90 mmHg at the follow up (Examination 5: 1994–1996).

Genotyping

Genotyping of FOXO3 and apolipoprotein E (APOE) variants was performed on blood samples that had been frozen at –70ºC. For men who participated in Examination 4 (1991–1993), DNA for genotyping was obtained from the blood sample buffy coat [24]. After DNA isolation, PCR was used for amplification of a suitable region of each gene using a combination of QIAmp cell-free DNA isolation followed by REPLI-g Single-Cell WGA & WTA amplification (QIAGEN Sciences, Germantown, MD, USA). Genotyping was performed using TaqMan on an Applied Biosystems QuantStudio 12K Flex system (ThermoFisher Scientific, Waltham, MA, USA).

Outcome measures

The current analysis was based on participants with autopsy and genotyping data. Microscopic neuropathologic assessments were performed by 1 of 3 senior neuropathologists who had trained together for standardization. The neuropathologist who carried out the assessment was blind to the clinical features. Methodology and definitions of cerebrovascular lesions have been described in previously published studies [7,25,26]. The brain autopsy data were coded in two distinct periods (1992–2003 and 2003–2012) using two different coding forms. To standardize both batches of data, brain locations that were used commonly in both periods were included for the study, namely bilateral (right/left) sections from each of the following areas: caudate, putamen, globus pallidus, thalamus, frontal lobe, temporal lobe, parietal lobe, occipital lobe, hippocampus, and brainstem. CMI was defined as presence of microinfarct, microlacune, and other focal ischemic changes. A dichotomized variable, defined as presence of any CMI regardless of the number of lesions, was defined as the primary outcome for the analysis.

Statistical Analysis

General linear model (GLM) was used to compare the baseline variables between subjects with CMI and without any CMI. Logistic regression models were used to assess the association of hypertension with prevalence of any CMI. Odds ratios (OR) and 95% confidence intervals (CI) were estimated for the whole cohort and then stratified by FOXO3 genotype (G-allele carriage versus TT genotype) adjusting for age at death, and baseline variables at exam 4 (1991–1993), namely: BMI, fasting glucose, type 2 diabetes, total cholesterol, HDL, smoking, PAI, alcohol consumption, history of prevalent clinical stroke, and APOE ε4 genotype. The interaction term of hypertension with FOXO3 G-allele in a multivariate full logistic model was tested for any genetic modulating effect to the present. Levels of P <0.05 were considered statistically significant. All statistical analyses were performed using the Statistical Analysis System (SAS) version 9.4 (Cary, North Carolina, USA).

RESULTS

Among a total of 8,006 participants, 809 men with complete brain autopsy and information from KHHP examination 4 on hypertension and FOXO3 genotype were included in the analysis. Overall, 551 (63.2%) participants had evidence of any CMI. Characteristics of those with and without any CMI are shown in Table 1. Those with any CMI were significantly more likely to have baseline hypertension, higher BMI, higher fasting glucose, and prevalent clinical stroke. Furthermore, there was no difference in the proportion of controlled hypertension between those with and without CMI (36.8% vs. 29.6% respectively, P = 0.13). Prevalence of APOE ε4 allele and total cholesterol levels was slightly higher in those with any CMI but did not reach statistical significance.

TABLE 1.

Baseline characteristics of participants with and without any cerebral microinfarcts (CMI)

	No CMI	Any CMI	P *
n	258 (%)	551 (%)
Hypertension	177 (68.6)	423 (76.8)	0.013
Age at death (years)	88.7 ± 6.2	88.9 ± 5.7	0.73
BMI (kg/m2)	23.1 ± 3.07	23.8 ± 3.22	0.008
Fasting glucose (mg/dL)	108.4 ± 21.2	113.0 ± 27.9	0.021
Diabetes	63 (24.3)	153 (27.8)	0.31
APOE ε4 genotype	40 (15. 6)	114 (20.7)	0.082
Total cholesterol (mg/dL)	187.3 ± 32.0	192.0 ± 33.2	0.063
HDL cholesterol (mg/dL)	51.3 ± 13.1	51.3 ± 13.2	0.97
Smoking (pack-years)	24. 5 ± 31.9	25.5 ± 35.5	0.70
Physical activity index	30. 9 ± 4.97	30.6 ± 4.28	0.40
Alcohol intake (oz/month)	17.8 ± 40.3	14.6 ± 26.2	0.19
Prevalent clinical stroke	6 (2.3)	33 (6.0)	0.023
CASI at baseline	82.2 ± 16.1	81.7 ± 18.0	0.72
FOXO3 G-allele carriage	125 (48. 5)	259 (47.0)	0.70

^*Statistically significant P values are shown in bold.

Abbreviations: CMI, cerebral microinfarct; Hypertension = SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or use of anti-hypertensive medications at baseline (Examination 4); BMI, body mass index; Diabetes = Type 2 diabetes by modified ADA criteria, with and without use of medications; APOE, apolipoprotein E; HDL, high-density lipoprotein; FOXO3, forkhead winged-helix box O type 3; CASI, Cognitive Abilities Screening Instrument. Data are n (%) and mean ± SD.

Multivariable models were created (Table 2) to assess the predictors of any CMI. In models 1 and 2, hypertension was a significant predictor of CMI. However, in the full covariate adjusted multivariable model, only BMI (OR: 1.07, 95% CI: 1.01–1.14, P = 0.015) was a predictor of CMI, while hypertension was only a borderline predictor (OR: 1.44, 95% CI: 1.00–2.08, P = 0.052). The interaction effect of the FOXO3 G-allele with hypertension on any CMI was estimated by adding the interaction term FOXO3 x hypertension to model 4. Doing so resulted in a significant interaction being seen (P = 0.020). The significant interaction indicated that FOXO3 genotype modulated the effect of hypertension on CMI.

TABLE 2.

Logistic regression models for any cerebral microinfarcts (CMI)

	Model 1 OR (95% CI)	Model 2 OR (95% CI)	Model 3 OR (95% CI)	Model 4 OR (95% CI)
Hypertension	1.51 (1.09, 2.10) *	1.45 (1.03, 2.05) *	1.40 (0.99, 1.98)	1.44 (1.00, 2.08)
Age at death	1.00 (0.98, 1.03)	1.01 (0.99, 1.04)	1.01 (0.98, 1.04)	1.01 (0.99, 1.04)
BMI (kg/m2)		1.05 (1.00, 1.11) *	1.06 (1.01, 1.11) *	1.07 (1.01, 1.13) *
Fasting glucose (mg/dL)		1.01 (1.00, 1.02)	1.01 (1.00, 1.02)	1.01 (1.00, 1.02)
Diabetes		0.86 (0.56, 1.31)	0.85 (0.55, 1.30)	0.84 (0.54, 1.33)
APOE ε4 genotype			1.44 (0.95, 2.19)	1.45 (0.93, 2.25)
Total cholesterol (mg/dL)				1.00 (1.00, 1.01)
HDL cholesterol (mg/dL)				1.01 (0.99, 1.02)
Smoking (pack-years)				1.00 (1.00, 1.01)
Physical activity index				0.98 (0.94, 1.01)
Alcohol intake (oz/month)				1.00 (0.99, 1.00)
Prevalent clinical stroke				2.32 (0.82, 6.55)
CASI at baseline				1.01 (0.99, 1.02)
FOXO3 G-allele carriage				0.92 (0.66, 1.27)

^*Statistically significant P values are shown in bold.

Abbreviations: Hypertension = SBP ≥ 140 mmHg and/or DBP ≥ 90 mmHg or use of anti-hypertensive medications; BMI, body mass index; Diabetes = Type 2 diabetes by modified ADA criteria, with and without use of medications; APOE, apolipoprotein E; HDL, high-density lipoprotein; FOXO3, forkhead winged-helix box O type 3. *P <0.05.

In the stratified analyses, among the participants without the longevity-associated FOXO3 G-allele (i.e., those who had the TT genotype), hypertension was a strong predictor of CMI (OR: 2.25, 95% CI: 1.34–3.77, P = 0.002), while among those with the longevity-associated FOXO3 G-allele, hypertension was not a predictor of CMI (OR: 0.88, 95% CI: 0.51–1.54, P = 0.660) (Table 3).

TABLE 3.

Logistic regression models for any cerebral microinfarcts (CMI), stratified by FOXO3 genotype

	FOXO3 TT (n=425)		FOXO3 G-allele (n=384)
	OR (95% CI)	P	OR (95% CI)	P
Hypertension	2.30 (1.37, 3.87)	0.002	0.89 (0.51, 1.56)	0.69
Age at death	1.01 (0.97, 1.06)	0.53	1.02 (0.98, 1.06)	0.41
BMI (kg/m2)	1.04 (0.96, 1.13)	0.30	1.10 (1.01, 1.20)	0.029
Fasting glucose (mg/dL)	1.01 (1.00, 1.02)	0.27	1.01 (1.00, 1.03)	0.08
Diabetes mellitus	0.98 (0.51, 1.88)	0.94	0.62 (0.32, 1.22)	0.17
APOE ε4 genotype	1.74 (0.90, 3.35)	0.10	1.17 (0.63, 2.17)	0.63
Total cholesterol (mg/dL)	1.00 (1.00, 1.01)	0.35	1.0 (1.00, 1.01)	0.25
HDL cholesterol (mg/dL)	1.02 (1.00, 1.04)	0.056	0.99 (0.98, 1.01)	0.53
Smoking (pack-years)	0.99 (0.99, 1.00)	0.13	1.01 (1.00, 1.01)	0.20
Physical activity index	0.94 (0.89, 0.99)	0.030	1.01 (0.96, 1.06)	0.81
Alcohol intake (oz/month)	1.00 (0.99, 1.01)	0.87	0.99 (0.99, 1.00)	0.13
Prevalent clinical stroke	5.34 (1.11, 25.7)	0.037	1.21 (0.26, 5.74)	0.81
CASI at baseline	1.02 (1.00, 1.04)	0.058	0.99 (0.97, 1.01)	0.44

DISCUSSION

Among subjects homozygous or heterozygous for the longevity-associated G-allele of FOXO3 SNP rs2802292 (i.e., who were G-allele carriers), the impact of hypertension on the risk of any CMI found during brain autopsy was attenuated compared with those lacking the longevity-associated FOXO3 genotype (i.e., were TT). This is the first reported evidence indicating that FOXO3 longevity associated variants also provide resilience to CMI. We speculate that longevity associated FOXO3 variants exert an effect on cerebrovascular resilience so as to protect against the adverse effects of chronic hypertension (Figure 1). FOXO3 SNP rs2802292 G-allele carriage-related resilience to chronic cardiovascular stress was shown previously in KHHP subjects having a cardiometabolic disease (CMD) [15] and against intracerebral haemorrhage [27].

Graphical Abstract/Figure 1.

Conceptual framework illustrating the protective effect of the resilience allele on the cerebral vessels compared to the common allele with existing chronic hypertension. In subjects having the common FOXO3 genotype (TT), hypertension increased the risk of cerebral microinfarct (CMI). In subjects with a longevity-associated FOXO3 genotype (G-allele carriers), the deleterious impact of hypertension was attenuated and helped preserve brain health.

The precise pathophysiology of CMI is not entirely understood but the pathological appearance of CMI shares similar characteristics as the macroscopic ischaemic infarctions present, and thus the mechanism is speculated to be related to ischemia. When CMIs are identified in early stages, they show the predicted acute ischemic appearance of red neurons (if cortical) and sometimes with vacuolization from cytotoxic edema. In subacute stages, the lesions show an influx of macrophages followed by surrounding gemistocytic astrocytosis at about 10 days. Chronic lesions show cavitation with surrounding fibrillary gliosis [1]. The presence of cerebral vessel pathologies such as atherosclerosis, arteriosclerosis and/or amyloid angiopathy is independently associated with CMI [28].

Although hypertension is a well-established risk factor for stroke and cerebrovascular disease, the exact mechanism whereby hypertension contributes to the formation of CMI is uncertain. It is speculated that chronic hypertension increases mechanical stress on the endothelium, leading to degenerative changes in the penetrating arterioles, such as fibrinoid necrosis and deposition of plasma proteins, including fibrin, in the arteriolar wall. Such endothelial damage may lead to altered blood cell-endothelial interaction, microthrombus formation, and ultimately microscopic ischemic lesions [29]. Furthermore, chronic hypertension can accelerate the arteriosclerotic process and increase the likelihood of cerebral large artery and small artery stenosis [29], which may further worsen cerebral perfusion and increases the risk of CMI when a significant decline in blood pressure occurs [30].

Since the risk of stroke is proportional to the blood pressure level [31], one would assume that there may be a direct relationship between the severity of small vessel disease pathologies and blood pressure level. However, some studies have found fibrinoid necrosis of the vessel walls in those with only mild or benign hypertension, suggesting that other factors may modulate the susceptibility of cerebral vessel injuries in response to chronically elevated blood pressure [32].

It has been proposed that the FOXO3 transcription factor may protect blood vessels by effects on pathways that result in inhibition of vascular smooth muscle cell proliferation and neointimal hyperplasia [33], and thereby provide protection from vascular aging processes [12]. This may be particularly important during events of acute stress. Activation of FOXO3 transcription in human embryonic stem cells resulted in reinforcement of human vascular cell homeostasis, delayed aging, and increased resistance to oxidative injury compared with wild-type cells [34]. Loss-of-function studies have shown that FOXO3 expression helps to maintain homeostasis of a diverse array of vascular cell types [35,36]. We therefore speculate that elevation in intracellular FOXO3 levels, as occurs for G-allele carriers [37], may protect against progressive injuries to the cerebral arterioles of individuals with chronic hypertension, thereby protecting them from CMI. Future studies are needed to further assess how FOXO3 and its encoded protein may impact small vessel wall injuries of cerebral arterioles in response to chronic hypertension.

We acknowledge that our study has some limitations. Our study’s main dichotomized outcome variable was a presence or absence of one or more CMIs, and the total number of CMIs present in the collected specimen were not measured. Therefore, we were unable to assess the impact of the FOXO3 G-allele on the severity of the underlying cerebrovascular disease. Our study population of American men of Japanese ancestry living on Oahu in Hawaii limits generalizability of the study findings to other ethnic groups and to women. The study did not include women, the reason being that in 1965–1968 when recruitment took place, heart disease was uncommon in middle-aged women, and particularly in American women of Japanese ancestry. However, our study also has many strengths, including a large overall sample size, its longitudinal prospective study design, and the very long follow-up period which was accompanied by multiple periodic examinations involving collection of a wide array of key clinical information. Although the cohort only included men of Japanese descent, the population is unique in that it is genetically more homogenous than other racial populations, and that the latter and other similar Japanese populations have not been studied extensively. We had, moreover, a considerable amount of data on other cardiovascular risk factors, allowing us to adjust for these factors to minimize confounding. Our surveillance system for incident stroke was thorough, facilitated by ours being an island population, and follow up was meticulous.

In conclusion, the present study found that FOXO3 longevity genotype attenuates the impact of hypertension on the risk of CMI in American men of Japanese ancestry living on the island of Oahu in Hawaii. Future studies should attempt to replicate these findings in other large populations, including in other ethnic groups, and in women.