Abstract
Background and aims:
We assessed the predictive role of coronary artery calcification (CAC) in clinically relevant cognitive impairment in 148 middle-aged individuals with childhood-onset type 1 diabetes (TID) from the Pittsburgh Epidemiology of Diabetes Complications (EDC) Study.
Methods:
Baseline CAC was measured in 1996–98 and repeated 4–8 years later. Per extensive neuropsychological testing in 2010–15, 28% (41/148) of participants met the study definition of clinically relevant cognitive impairment (two or more of 7 select test scores ≥1.5SD worse than demographically appropriate published norms). Logistic regression models with backward selection were constructed for statistical analysis.
Results:
Mean age and TID duration at first CAC measure were 37 and 29 years, respectively. A greater burden of initial CAC was associated with cognitive impairment determined 14 years later. Compared to Agatston score = 0, odds ratio (OR) and 95% confidence intervals (CI) of 0 < − 100, 100 < −300 and > 300 were 1.4 (0.6, 3.6), 2.3 (0.6, 9.7), and 7.9 (1.6, 38.5), respectively. With both initial and progression of CAC in the multivariable model, backward selection retained only CAC progression, showing it was significantly associated with cognitive impairment (OR [95% CI]: 1.7 [1.1, 2.9]). In those with an initial CAC > 0, CAC density was marginally, inversely, associated with cognitive impairment when controlling for CAC volume (OR [95%CI]: 0.3 (0.1, 1.2), p value = 0.078).
Conclusions:
Greater CAC burden was associated with clinically relevant cognitive impairment in middle-aged adults with childhood-onset TID. CAC progression appears to be a more powerful predictor than initial calcification.
Keywords: Coronary artery calcification, Type 1 diabetes, Cognitive impairment
1. Introduction
Dementia has become the greatest global public health challenge in modern society, and the number affected is projected to triple from 201S to 2050 [1]. Impaired cognitive function is highly prevalent in people with type 1 diabetes [2]. The Pittsburgh Epidemiology of Diabetes Complications (EDC) Study reported a 28% prevalence of clinically relevant cognitive impairment in middle-aged (mean age: 49 years) adults with childhood-onset type 1 diabetes [2], being comparable with that of the general population aged 85 years or older [3]. The pathophysiological basis for the earlier-than-expected occurrence of cognitive dysfunction in people with type 1 diabetes remains unclear, yet its associations with microvascular complications (diabetic retinopathy [4] and diabetic neuropathy [5,6]) and extreme ankle-brachial index (ABI) values suggest a parallel vascular pathogenesis in the brain as well as in other parts of the body [7].
Numerous studies have documented a link between cognitive function and cardiovascular disease and associated risk factors in the general population [8]. Coronary artery calcification (CAC), an established marker of subclinical atherosclerosis in the heart and a predictor of cardiovascular events has been shown to be independently associated with poor cognition and dementia in the general population [10–12].
Diabetes is known to carry a greater burden of vascular calcification, leading to an accelerated vascular aging process, which may, at least partially, explain the increased cardiovascular risk in this patient population. To our knowledge, however, few studies has examined the role of CAC in cognition impairment in those with type 1 diabetes [13]. Thus, the primary aim of this study was to evaluate the relationship of CAC with later clinically relevant cognitive impairment among middle-aged individuals with childhood-onset type 1 diabetes. Initial CAC, annual progression rate, and CAC density were all evaluated.
Vascular calcification occurs in two distinct forms, depending on the location: intima or medial wall. Though CAC may reflect both intimal and medial calcification, existing evidence suggested it mainly represented intimal calcification [14]. Medial arterial calcification is a feature of diabetes, which is characterized by high ankle blood pressure, reflecting poor compressibility of the artery. A high ABI associated with medial calcification is already well established [15]. The secondary aim was to assess the relative roles of CAC and high ABI in the association with impaired cognition. As the CAC was only shown having a modest effect on poor cognition in prior work [10–12], we hypothesized that CAC and high ABI were independently associated with cognitive impairment.
2. Study population and methods
2.1. Study population
The Pittsburgh EDC cohort is an ongoing, prospective study of now middle-aged adults diagnosed with childhood-onset type 1 diabetes (diagnosed before age 17 years). This population has previously been described in detail [16]. In brief, 658 eligible participants were examined between 1986 and 1988, with the 30th year of follow-up currently ongoing. Participants were assessed biennially by surveys and using highly standardized methods by examination for the first 10 years and again in 18, 25 and 30 years. The cohort has been shown to be epidemiologically representative of the type 1 diabetes population in Allegheny County, Pennsylvania [17]. The University of Pittsburgh Institutional Review Board approved the study protocol.
At the EDC 10-year follow-up examination (1996–1998), electron beam computed tomography (EBCT) scans were conducted to measure CAC for all participants aged 218 years; some repeated the scans 4–8 years later (2000–2006). During 2010–2015 (at the EDC 22- and 25-year follow-ups), all locally dwelling EDC participants were invited to participate in an ancillary neurocognitive study to evaluate cognition. A total of 148 participants underwent both of the CAC and the neurocognitive tests and were eligible in the present analysis of CAC and cognitive impairment. The mean (standard deviation, SD) time interval between initial CAC and later cognitive tests was 14.0 (3.5) years (range: 4.7–19.2 years). Of these 148, 116 of them had repeated CAC scans and were included in the subgroup analysis of CAC progression.
2.2. Cognitive assessment
The cognitive assessment of the EDC study has been described in details elsewhere [2,18]. Briefly, extensive neurocognitive tests were performed to assess cognitive function, including verbal intelligence (North American Adult Reading Test [NAART]), learning and working memory (Rey Auditory Verbal Learning Tests, Rey-Osterrieth Complex Figure Delay Task [ROCF-Delay], and Four Word Short Term Memory [4WSTM]), executive function (Verbal Fluency F-A-S [FAS], Trails Making Test-part B [TMTB], Stroop Color-Word [Stroop-CW], and Letter/Number Sequence [LN Sequence]), psychomotor efficiency (Digit Symbol Substitution Test [DSST], Grooved Pegboard [GPI, and Trail Making Test-part A [TMTA]), semantic fluency (Verbal Fluency Animals [Animals]), and visuo-construction skills (Rey-Osterrieth Complex Figure Copy Task [ROCF-copy]). Of these, individual with type 1 diabetes performed significantly worse than non-diabetes controls on the following seven tests: FAS, TMTB, DSST, GP, Stroop-CW, Animals, and ROCF-copy [2]. Clinically relevant cognitive impairment was defined as having two or more of these test scores ≥1.5 SD worse than demographically-appropriate published norms [19,20].
2.3. Assessment of coronary artery calcification (CAC)
CAC was measured using EBCT scans (GE-Imatron C-ISO scanner, South San Francisco, CA, USA). Scans were triggered by electrocardiogram signals at 80% of the R-R interval and obtained in 3 mm contiguous sections of the heart. Calcification was quantified using the method described by Agatston et al. [21] and CAC volume was based on the isotropic interpolation method [22]. The area and density scores were obtained using the algorithms from the Multi-Ethnic Study of Atherosclerosis (MESA) study [23]. Specifically, the area score was obtained by dividing the volume by the scan slice thickness of 3.0 mm. The density was subsequently calculated by dividing the Agatston score by area score. The progression was determined as the annualized difference between the square root of baseline and square root of follow-up CAC score (annual CAC progression = [sqrt (follow-up CAC) - sqrt (baseline CAC)]/[t2-t1 in years]). This calculation was previously applied by Radford et al. [24]. CAC volume might be a better measure of progression as the Agatston score is determined using both CAC density and volume [25]. Our primary analysis of CAC progression was thus based on CAC volume. Progression in CAC Agatston score associated with incident CAD was also examined as Agatston score has been widely used.
2.4. Measurement of covariates
Information regarding demographic, physical covariates, biomedical tests, and diabetes complication status were collected at the time of first CAC measure. Demographic and medical history information was obtained through questionnaires. An ever smoker was defined as a person who had smoked at least 100 cigarettes in the lifetime. Body mass index (BMI) was calculated as the weight in kilograms divided by the square of the height in meters. Blood pressure was measured using a random zero sphygmomanometer according to the Hypertension Detection and Follow-up Program protocol [26]. Hypertension was defined as a systolic blood pressure ≥140 mm Hg, a diastolic blood pressure ≥90 mm Hg, or the use of antihypertensive medications.
A Doppler blood flow detector was used to measure systolic blood pressure in the brachial, posterior tibial, and dorsalis pedis arteries to obtain ABI. An average of two readings greater than 1.3, or non-compressible, was defined as high ABI. The cut-off point of 1.3 were based on published work from this same cohort, showing ABI > 1.3 was highly correlated with medial arterial calcification by X-ray [15]. A meta-analysis documented ABI < 0.9 to be a marker of cognitive impairment and dementia [27]. However, only one participant in this analysis had an ABI < 0.9, which disallowed us to examine associations for extremely low ABI.
Glycosylated hemoglobin was obtained by the automated high-performance liquid chromatography (HPLC) (Diamat, BioRad, Hercules, CA, USA) at the time CAC was first assessed, and subsequently by the DCA 2000 analyzer (Bayer, Tarrytown, NY, USA). Results from the two methods were highly correlated (r = 0.95) [28]. The values were then converted to DCCT-aligned HbA1c using a regression equation derived from duplicate assays [28]. Total cholesterol and triglycerides were determined enzymatically [29,30]. High-density lipoprotein (HDL) cholesterol was measured by a precipitation technique (heparin and manganese chloride) using a modified version of the Lipid Research Clinics method [31]. Non-HDL cholesterol was calculated as total cholesterol minus HDL cholesterol. Urinary albumin was determined by immunonephelometry [32].
Diabetes complications, including coronary artery disease (CAD), proliferative retinopathy, and distal symmetric polyneuropathy (DSP), were evaluated at each follow-up using standard protocols as described previously [33].
2.5. Statistical analysis
Demographics and risk factors were compared between those with and without prevalent CAC (baseline CAC > 0 vs. = 0) and between those with and without clinically relevant cognitive impairment. The Chi-square test (or Fisher’s Exact test) for dichotomous variables and the Student’s t-test (or nonparametric Wilcoxon test) for continuous variables were used for these comparisons, as appropriate.
Logistic regression models were constructed to assess the associations of CAC measures and high ABI with clinically relevant cognitive impairment. Baseline CAC was examined in 4 categories (0, 0 < − 100, 100 < −300 and > 300). Odds ratios (ORs) and confidence intervals (CIS) were presented accordingly. Risk factors previously demonstrated to be important predictors of cognitive impairment in this type 1 diabetes group [2] and factors found to be significantly associated with the outcome events in univariable analyses were considered in the multiple regression analyses. Because age and diabetes duration were highly correlated (r = 0.84) in this cohort, only diabetes duration was used for the adjustment. As education attainment associated with cognitive impairment has been well established [34], year of education was adjusted in all the models.
To limit overfitting of a multivariable model, backward elimination was applied, allowing for a full set of covariates (gender, diabetes duration, years of education, age at cognitive test, apolipoprotein E4 [ApoE4] status [24, 34, or 44], ever smoking, BMI, hypertension, HbA1c, urinary albumin excretion rate[AER], HDL and non-HDL cholesterol, triglycerides, statin use, and presence of diabetes complications: CAD, DSP and proliferative retinopathy). Our model selection strategy was for a parsimonious model that was adjusted for potentially important confounding variables and had a lower Akaike Information Criterion (AIC) to be selected as the final model. Alternative multiple-adjusted models which allowed for different sets of risk factors are available in the Supplementary materials.
The log-likelihood ratio test was applied to assess the improvement of model fit for including CAC and/or high ABI.
Logistic regression with backward elimination was also applied for the tests of CAC density and CAC progression, respectively. Only participants with CAC over 0 were analyzed for CAC density (with controlling for log-transformed CAC volume) because plaque density could only be quantified in those with prevalent CAC [23]. The progression of CAC associated with clinically relevant cognitive impairment was examined in the subset with at least two CAC measures, with offering baseline CAC and other covariates (described above) in the model for backward selection.
Triglycerides and AER were log-transformed prior to statistical testing, given the highly skewed distribution of these variables. A 2-sided p < 0.05 was considered significant. All analyses were performed with SAS 9.3 (SAS Institute, Cary, NC).
3. Results
Among the 148 participants at the first CAC measure (baseline of this analysis), the mean age and diabetes duration were 37.2 (range 18.3–56.1) and 28.6 (range 17.6–50.6) years, respectively. There were 46% (68/148) of participants with prevalent CAC at baseline. The time interval between two CAC scans were 5.5 (1.6) years. In those with baseline CAC = 0, 39% developed somewhat calcification during the follow-up. Compared to participants without CAC, those with prevalent CAC were older, had higher BMI and systolic blood pressure, were more likely to use statins, have hypertension and/or proliferative retinopathy (Table 1).
Table 1.
Clinical characteristics at the first measurement of CAC.a
| Characteristics | Overall | CAC status at baseline | |
|---|---|---|---|
| N = 148 | No calcification (n = 80) | Prevalent CAC (n = 68) | |
| Age, years | 37.2 (7.0) | 34.4 (6.8) | 40.5 (5.7)e |
| Diabetes duration, years | 28.6 (6.9) | 26.3 (6.1) | 31.4 (6.7)e |
| Age at diabetes onset, years | 8.6 (4.0) | 8.1 (4.1) | 9.1 (3.9) |
| Female, % (n) | 51.0 (76) | 50.0 (40) | 52.2 (36) |
| Years of education, years | 14.9 (2.5) | 15.0 (2.5) | 14.8 (2.6) |
| Age at cognitive test, years | 51.1 (7.0) | 49.3 (6.8) | 53.3 (6.7)e |
| ApoE (24, 34, or 44), % (n) | 28.4 (42) | 33.8 (27) | 22.1 (15) |
| Ever smoker, % (n) | 33.6 (50) | 33.8 (27) | 33.3 (23) |
| BMI, kg/m2 | 25.3 (4.0) | 24.1 (3.1) | 26.7 (4.5)e |
| Systolic BP, tnmHg | 114.5 (16.5) | 110.4 (12.1) | 119.2 (19.4)e |
| Diastolic BP, mmHg | 69.7 (11.0) | 68.9 (8.6) | 70.7 (13.3) |
| Hypertension, % (n) | 18.1 (27) | 8.8 (7) | 29.0 (20)e |
| High AB1b, % (n) | 8.7 (13) | 8.8 (7) | 8.7 (6) |
| HbAlc, % | 8.3 (1.4) | 8.4 (1.3) | 8.2 (1.5) |
| AERc, μg/min | 8.2 (4.6, 35.9) | 7.5 (4.9, 21.3) | 11.9 (4.2,80.0) |
| HDL cholesterol, mmol/L | |||
| Male | 47.9 (11.0) | 1.3 (0.3) | 1.2 (0.3) |
| Female | 63.2 (14.2) | 1.7 (0.4) | 1.5 (0.4) |
| Non-HDL cholesterol, mmol/L | 130.7 (36.2) | 3.3 (1.1) | 3.5 (0.8) |
| Triglyceridesc, mmol/L | 78.0 (58.0, 107.0) | 0.8 (0.6, 1.3) | 0.9 (0.7, 1.2) |
| Statin use, % (n) | 7.5 (11) | 0.0 (0) | 15.9(11)e |
| Distal symmetric polyneuropathy, % (n) | 26.2 (39) | 21.3(17) | 31.9 (22) |
| Proliferative retinopathy, % (n) | 45.0 (67) | 36.3 (29) | 55.1 (38)d |
| CAD, % (n) | 13.4 (20) | 10.0 (8) | 17.4 (12) |
| Baseline CAC | |||
| Agatstonc | 0(0, 31.5) | 0 (0, 0) | 42.6 (8.8, 188.9)e |
| Volumec | 0 (0, 36.1) | 0 (0, 0) | 47.2 (14.4, 146.8)e |
| Density | 1.4 (1.9) | 0 (0, 0) | 2.8 (1.8)e |
ABI: ankle-brachial index, AER: urinary albumin excretion rate, ApoE: apolipoprotein E, BMI: body mass index, BP: blood pressure, CAC: coronary artery calcification, CAD: coronary artery disease, HbAlc: hemoglobin Ale, HDL cholesterol: high density lipoprotein cholesterol, SD: standard deviation.
Categorical variable were presented as percentage (number) and continuous variables as mean (SD) or median (1st and 3rd quartiles).
High ABI: ABI > 1.3.
Logarithmically transformed before statistical testing,
p value < 0.05 for the comparison between those with CAC = 0 and > 0.
p value < 0.01 for the comparison between those with CAC = 0 and > 0.
The neurocognitive test battery was conducted a mean (SD) of 14.0 (3.5) years after the first CAC measurement, with 41 (27.7%) participants meeting the definition of clinically relevant cognitive impairment Compared with cognitively normal non-cases, cognitive impairment cases were less educated, older, with longer diabetes duration, and more likely to have a high ABI (>1.3) and/or CAD (Supplementary Table 1).
The proportion of later cognitive impairment increased with increasing baseline CAC: at the follow-up, cognitive impairment was prevalent in 20.7% of those with an Agatston score = 0, in 27.3% of those with a score greater than 0 and up to100, in 36.4% of those with a score greater than 100 and up to 300, and in 72.7% of those with a score > 300 (p trend = 0.004). In multivariable logistic regression with baseline CAC and high ABI in the model (Table 2), a dose-gradient association was observed for CAC associated with later cognitive impairment. Compared to Agatston score = 0, OR (95%CI) of 0 < − 100, 100 < −300 and > 300 were 1.4 (0.6, 3.6), 2.3 (0.6, 9.7), and 7.9 (1.6, 38.5), respectively (final model, after backward elimination). CAC volume showed a similar association with cognitive impairment as that in Agatston scores (Supplementary Table 2). In addition to baseline CAC, high ABI (OR [95%CI]: 6.0 [1.5, 24.3]), fewer years of education, and having CAD were also significantly associated with worse cognition after backward elimination. When evaluating high ABI (Supplementary Table 3) and CAC (Supplementary Table 4) separately, both of these subclinical cardiovascular markers were independently and significantly associated with clinically relevant cognitive impairment.
Table 2.
Associations of baseline CAC Agatston scores and high ABI with having clinically relevant cognitive impairment at follow-up.
| Events = 41/N = 148 | Model 1 | Model 2 | Model 3 | Model 4 (backward selectionb, final model) | ||||
|---|---|---|---|---|---|---|---|---|
| Parameter | OR (95%CI)a | p value | OR (95%CI)a | p value | OR (95%CI)a | p value | OR (95%CI)a | p value |
| CAC Agatston scores | ||||||||
| 0 (17/82) | ref | ref | ref | ref | ||||
| 0 < −100 (12/44) | 1.4 (0.6, 3.6) | 0.173 | 1.3 (0.5, 3.3) | 0.279 | 1.4 (0.5, 3.8) | 0.286 | 1.4 (0.6, 3.6) | 0.236 |
| 100 < −300 (4/11) | 2.1 (0.5, 8.6) | 0.787 | 1.5 (0.3, 6.8) | 0.663 | 1.7 (0.4, 7.6) | 0.734 | 2.3 (0.6, 9.7) | 0.963 |
| > 300 (8/11) | 11.2 (2.4, 51.8) | 0.008 | 7.2 (1.4, 36.7) | 0.026 | 8.1 (1.5, 45.0) | 0.033 | 7.9 (1.6, 38.5) | 0.037 |
| ABI | ||||||||
| ≤ l.3 (33/135) | ref | ref | ref | ref | ||||
| > 1.3 (8/13) | 6.4 (1.7, 24.6} | 0.007 | 6,1 (1.5, 24.1) | 0.011 | 6.5 (1.5, 27.7) | 0.011 | 6.0 (1.5, 24.3) | 0.012 |
| Year of education | 0.6 (0.4, 0.9) | 0.007 | 0.5 (0.4, 0.8) | 0.007 | 0.6 (0.4, 0.9) | 0.013 | 0.6 (0.4, 0.9) | 0.010 |
| Female | NA | 1.1 (0.5, 2.6) | 0.791 | NA | ||||
| Age of cognition tests | NA | 1.2 (0.7, 2.2) | 0.474 | NA | ||||
| Diabetes duration | NA | 1.2 (0.7, 1.9) | 0.673 | NA | ||||
| HbAlc | NA | NA | 1.2 (0.8, 1.9) | 0.295 | ||||
| BMI | NA | NA | 0.8 (0.5, 1.3) | 0.384 | ||||
| Hypertension | NA | NA | 1.6 (0.6, 4.8) | 0.374 | ||||
| Statin use | NA | NA | 2.4 (0.4, 13.8) | 0.334 | ||||
| CAD | NA | NA | 2.9 (0.9, 9.4) | 0.073 | 3.4 (1.0, 10.8) | 0.042 | ||
| Proliferative retinopathy | NA | NA | 1.1 (0.5, 2.7) | 0.810 | ||||
| AIC | 160.433 | 164.599 | 165.335 | 157.976 | ||||
ABI: ankle-brachial index, AIC: Akaike information criterion, BMI: body mass index, CAC: coronary artery calcium, CAD: coronary artery disease, Cl: confidence Interval, HbAlc: hemoglobin Alc, NA: not available, OR: Odds ratio, SD: standard deviation.
Model 1: comprise CAC, high ABI, and year of education.
Model 2: Mode 1 + sex, age of cognition tests, and diabetes duration.
Model 3: Mode I + HbAlc, BMI, Hypertension, statin use, presence of CAD, and presence of proliferative retinopathy.
Model 4: Backward selection with a removal criterion of p value ≥ 0.10. In addition to baseline CAC and ABI, the year of education, sex, age at cognitive tests, diabetes duration, apolipoprotein E4 status, ever smoking, BMI, HbAlc, cholesterols (high density lipoprotein and non-high density lipoprotein), triglycerides, urinary albumin excretion rate, hypertension, proliferative retinopathy, distal symmetric polyneuropathy, and statin use were offered for model selection.
Per SD higher value for continuous variables.
CAC, high ABI, year of education and the existing CAD were retained in the final model after backward selection.
Compared to the base model (comprised of year of education and existing CAD, covariates that were retained after backward selections), the addition of CAC Agatston scores and high ABI significantly improved the fit of the model, respectively (Table 3). The addition of CAC Agatston scores to the model including high ABI did also lead to a significant improvement in model fit. The results were consistent with CAC volume (data not shown).
Table 3.
Likelihood ratio tests for model fit evaluation.
| Model | −2Log likelihood | p value |
|---|---|---|
| Basea | 160.474 | - |
| Basea + high ABI | 152.694 | 0.005 (vs. Base) |
| Basea + CAC Agatston | 151.945 | 0.003 (vs. Base) |
| Basea + CAC Agatston + high ABI | 144.908 | < 0.001 (vs. Base) |
| 0.005 (vs. Base + high ABI) | ||
| 0.008 (vs. Base + CAC Agatston) |
CAC: coronary artery calcification, CAD: coronary artery disease.
Base model was adjusted for year of education and existing CAD.
In the subgroup of 116 participants having repeated CAC measures (Table 4), the backward selection model allowing for the full set of covariates, retained annual progression of CAC, but not baseline CAC. A positive and significant association was found between CAC progression in volume (OR (95%CI) = 1.7 (1.1, 2.9), p value = 0.032) and later clinically relevant cognitive impairment. Besides CAC progression, years of education, diabetes duration, and high ABI were also retained in the final model. A similar association was observed for CAC progression in Agatston scores.
Table 4.
Associations between annual CAC progression and clinically relevant cognitive impairment in participants with repeated CAC measures.
| Events = 32/N = 116 | Model 1 | Model 2 (backward selectionb, final model) | |||
|---|---|---|---|---|---|
| OR (95%CI)a | p value | OR (95%CI)a | p value | ||
| Volume | |||||
| CAC progression | 2.1 (1.1, 3.8) | 0.019 | 1.7 (1.1, 2.9) | 0.032 | |
| Baseline | 0 | ref | |||
| CAC | 0 < −100 | 0.5 (0.2, 1.9) | 0.108 | ||
| 100 < −300 | 1.5 (0.2, 7.1) | 0.661 | |||
| > 300 | 2.2 (0.3, 13.4) | 0.335 | |||
| Year of education | 0.7 (0.5, 1.1) | 0.138 | 0.7 (0.4, 1.1) | 0.094 | |
| High ABI | NA | 5.7(1.3, 24.3) | 0.020 | ||
| Diabetes duration | NA | 1.7 (0.98, 2.8) | 0.061 | ||
| Agatston scores | |||||
| CAC progression | 1.9 (1.1, 3.2) | 0.021 | 1.5 (1.02, 2.4) | 0.039 | |
| Baseline | 0 | ref | |||
| CAC | 0 < −100 | 0.5 (0.1, 1.8) | 0.092 | ||
| 100 < −300 | 1.4 (0.3, 6.7) | 0.672 | |||
| > 300 | 2.0 (0.3, 12.8) | 0.339 | |||
| Year of education | 0.7 (O.5, l.l) | 0.128 | 0.7 (0.4, 1.1) | 0.088 | |
| High ABI | NA | 5.8 (1.4, 24.7) | 0.018 | ||
| Diabetes duration | NA | 1.6 (0.96, 2.8) | 0.068 | ||
ABI: ankle-brachial index, BMI: body mass index, CAC: coronaty artery calcium, CAD: coronary artery disease, Cl: confidence interval, HbAlc: hemoglobin Ale, NA: not available, OR: odds ratio, SD: standard deviation.
Model 1: comprise baseline CAC, CAC progression, and year of education.
Model 2: Backward selection with a removal criterion of p value ≥ 0.10, In addition to baseline and progression of CAC, the high ABI, year of education, sex, age at cognitive tests, diabetes duration, apolipoprotein E4 status, ever smoking, BMI, HbAlc, cholesterols [high density lipoprotein and non-high density lipoprotein,), triglycerides, urinary albumin excretion rate, hypertension, proliferative retinopathy, distaJ symmetric polyneuropathy, and statin use were offered for model selection.
Per SD higher value for continuous variables.
CAC progression, high ABI, year of education and diabetes duration were retained in the final model after backward selection.
Table 5 displays the association of CAC density with cognitive impairment with controlling for CAC volume in participants with CAC over zero. After backward selection of the full set of covariates, both baseline CAC volume and CAC density were retained in the model, along with years of education and high ABI. The CAC density was marginally inversely associated with prevalent cognitive impairment at follow-up (OR [95 CI%]: 0.3 [0.1, 1.2], p value = 0.078).
Table 5.
Association between CAC density and clinically relevant cognitive impairment after controlling for CAC volume in participants with baseline CAC over zero.
| Events = 24/N = 68 | Model 1 | Model 2 (backward selectiorb, final model) | ||
|---|---|---|---|---|
| OR (95%CI)a | p value | OR (95%CI)a | p value | |
| Log (CAC volume) | 2.6 (1.3, 5.4) | 0.010 | 2.9 (1.3, 6.4) | 0.008 |
| CAC density | 0.4 (0.1, 1.5) | 0.161 | 0.3 (0.1, 1.2) | 0.078 |
| Year of education | 0.5 (0.3, 0.9) | 0.018 | 0.4 (0.2, 0.8) | 0.007 |
| High ABI | NA | 36.6 (2.1, 644.0) | 0.014 | |
ABI: ankle-brachial index, BMI: body mass index, CAC: coronary artery calcium, CAD: coronary artery disease, Cl: confidence interval, HbAlc: hemoglobin Alc, NA: not available, OR: odds ratio, SD: standard deviation.
Model 1: comprise CAC volume, CAC density, and year of education.
Model 2: Backward selection with a removal criterion of p ≥ 0.10. In addition to CAC volume and CAC density, the high ABI, year of education, sex, age at cognitive tests, diabetes duration, apolipoprotein E4 status, ever smoking, BMI, HbAlc, cholesterols (high density lipoprotein and non-high density lipoprotein), triglycerides, urinary albumin excretion rate, hypertension, proliferative retinopathy, distal symmetric polyneuropathy, and statin use were offered for model selection.
Per SD higher value for continuous variables.
Baseline CAC volume, CAC density, year of education, and high ABI were retained in the model after backward selection.
No evidence was observed for interaction effect of sex on the associations between CAC measures and cognitive impairment in these data (all the interaction terms of sex with CAC measures were non-significant). When we additionally adjusted for sex and/or diabetes duration (if not retained after the backward elimination) in the final models, the results were consistent and there was an inferior model fit based on AIC (data not shown).
4. Discussion
In a prospective cohort of individuals living with type 1 diabetes since childhood, we found a significantly increased rate of later clinically relevant cognitive impairment among those with a greater burden of baseline CAC. CAC progression appeared a stronger risk factor for later cognitive impairment than initial CAC in this group of participants. An inverse association was also observed, in those with a CAC above zero, between CAC density and impaired cognition when controlling for CAC volume. These associations were independent of years of education, high ABI, and other risk factors that significantly contributed to the prediction models. While a previous report from the Pittsburgh EDC cohort found that a high ABI (>1.3) quadrupled the risk of cognitive impairment [2], the present study extends this prior work by showing that calcified atherosclerotic plaque burden in the heart provides additional information over peripheral arterial disease in predicting impaired cognition in these middle-aged adults with type 1 diabetes.
There were 46% of type 1 diabetes individuals with prevalent CAC at a mean age of 37 years in this cohort, which is tripled compared to the CAC prevalence rate in the general population with similar age [35]. Over a mean interval of 5.5 years between two CAC scans, the incidence of CAC progression was 41 % when defining progression with the method by Hokanson et al. [25] (a square root transformed difference > 2.5). In those with baseline CAC 0, 39% developed somewhat calcification over 5 years, a four-fold increase than that in the general population with a comparable age and a similar time interval between the two CAC scans [36].
There has been increasing interest regarding the role of cardiovascular disease and associated risk factors in age-related cognitive decline and dementia. CAC is an accurate index of subclinical atherosclerosis and the association with poor cognition function has consistently been observed in the general population across middle-to older -aged adults. Both the Cardiovascular Health Study (CHS) Cognition Study, with a population aged 80 + years [11], and the AGES-Reykjavik Study, a general population-based cohort with a mean age of 76 years [37], suggest an increased risk of dementia was associated with a greater burden of CAC in elderly populations. The Coronary Artery Risk Development in Young Adults (CARDIA) study further reported an association between poor cognition and a higher CAC in a middle-aged general population (mean age = 50 years) [12].
It is well recognized that atherosclerosis affects those with diabetes at a higher rate than occurs in the general population [38]. More recently, studies have identified worse cognitive performance in people with diabetes than in age-matched controls without diabetes [2,13]. However, very few studies have evaluated the predictive values of subclinical atherosclerosis on the cognitive complications of diabetes [13]. One study by Hugenschmidt et al. observed a modest association of baseline CAC with poorer performance on cognitive functioning tests in a group of older adults with type 2 diabetes, with average age of 67 years [13]. The current results are thus consistent with the prior work in an aging general population and in older adults with type 2 diabetes. Our work extends this to a middle-aged cohort of adults with childhood-onset type 1 diabetes (mean age: 37 years).
In addition, we observed a stronger magnitude of the effect of CAC progression over initial CAC, in the association with impaired cognition, emphasizing a dynamic nature for the influence of subclinical arthrosclerosis on the pathogenesis of cognitive impairment. Though previous studies has shown the predictive value of CAC progression in cardiovascular disease [24] and mortality [39], very few have studied the relation between dynamic changes in CAC and cognition functioning, to our best knowledge. The pathological pathways between vascular calcification and cognition function are not completely clear. Accelerated vascular calcification may contribute to white matter hyperintensities [40], reflecting cerebral small vessel disease [41], which are correlated to cognitive impairment and progression of amyloid burden [40,42]. Future studies are needed to determine whether a low extent of vascular calcification and/or a slower progression of CAC might reduce the development of diabetes-related cognitive impairment in adults with type 1 diabetes, and whether these findings apply to those with type 2 diabetes or to the general aging population as well.
Vascular calcification is an active and regulated process calcium mineral deposition within arterial intimal and/or medial walls [43]. Unfortunately, conventional non-invasive CT scans cannot distinguish intimal and medial calcification. Existing evidence have shown that CAC primarily represents intimal calcification of focal plaque and atheromatous burden in the coronary arteries [14]. Though CAC may also reflect medial calcification, the degree of medial wall calcification in coronary arteries is difficult to determine and has not been clearly demonstrated to our knowledge. While, the association between high ABI and medial wall calcification has been well described [15].
Being distinct from previous studies, we included both CAC and high ABI in the prediction models. We found that the addition of high ABI to a model already included CAC led to a significant improvement in model fit. The addition of ABI could be explained by a higher plaque burden, involving not only the heart but the peripheral vessels as well. Importantly, however, our results also showed the independent associations of CAC and high ABI with clinically relevant cognitive impairment. We thus believe it is biologically plausible to postulate that CAC, mainly representing intimal calcification, and high ABI, mainly representing medial wall calcification, may respectively contribute to cognitive dysfunction, via distinct vascular pathological pathways, in people with type 1 diabetes. Indeed, animal and human studies have suggested that intimal and medial calcification, referred to as atherosclerosis and arteriosclerosis, respectively, are distinct vascular diseases [44] and the risk factors predicting atherosclerotic burden were not necessarily associated with medial calcification [45]. As such, there may be unique opportunities to identify new therapies targeting medial calcification that ultimately decrease the risk of cognitive impairment and dementia beyond general atherosclerosis prediction.
CAC has been proposed as a measure of “biological age” for its incremental value of risk prediction in addition to chronological age [46]. The present findings support that ABI may also be a useful aging index in addition to CAC for estimating “biological age”, especially for “vascular age”, at least in type 1 diabetes populations.
Our findings also suggest that routine monitoring of CAC and ABI may identify those type 1 diabetes individuals at an increased risk of cognitive impairment, facilitating an earlier implementation of appropriate evaluation and intervention to improve long-term prognosis.
One of our novel findings was that CAC density appears to be borderline protective against impaired cognition when controlling for CAC volume. Criqui et al. also reported an inverse association between CAC density and CVD events in the general population using MESA data [23]. A higher density of calcification might prompt more stable plagues, leading to a lower risk of incident acute coronary artery syndrome [23]. Thomas et al. subsequently identified inverse associations of many CVD risk factors with CAC density, including total cholesterol, blood pressure, smoking, diabetes, BMI, and a series of inflammation biomarkers (interleukin-6, C reactive protein, and Fibrinogen) [47]. To some extent, the factors above are directly or indirectly related to inflammation. As such, a lower degree of inflammation with higher CAC density may also explain the inverse relationship of CAC density and cognitive impairment, consistent with recently reported associations between increased inflammation and cognitive dysfunction and dementia [48,49].
Strengths of this study include the concurrent availability of ABI, CAC, and later neurocognitive evaluations in a well-characterized cohort of individuals with childhood-onset type 1 diabetes. The repeated measurement of CAC allows an assessment of the role of longitudinal change in CAC on cognition. An important limitation of the study was a relatively small sample size, which restricted stratified analysis or testing potential interactions, such as ABI and CAC. Another major consideration is that neurocognitive tests was not conducted at the time of first CAC was measured (the baseline of the present study), which limits the ability to determine causality. However, only one cognitively impaired participant in this analysis was diagnosed less than age 37 years (mean baseline age of the study). It is therefore likely that most of the cognitive impairment cases were truly incident during the follow-up period. Moreover, whether the present findings of vascular calcification with worse cognitive performance will extend to dementia outcome in older age remains unknown.
4.1. Conclusions
A greater burden of CAC is associated with higher odds of having clinically relevant cognitive impairment in middle-aged individuals with long-lasting type 1 diabetes. CAC progression was more informative than initial CAC in predicting impaired cognition. This implies a persistent and accumulating influence of subclinical atherosclerosis on diabetes-related cognition impairment. CAC and high ABI were independently associated with impaired cognition in this study, suggesting that the screening of CAC and ABI may allow early identification of type 1 diabetes patients at an increased risk of developing cognitive impairment.
Supplementary Material
HIGHLIGHTS.
Coronary artery calcification (CAC) is associated with cognitive impairment in middle-aged type 1 diabetes individuals.
CAC progression is more informative than initial CAC in predicting impaired cognition in type 1 diabetes.
CAC and high ankle-brachial index are independently associated with cognitive impairment in type 1 diabetes.
Acknowledgments
The authors thank the staff and the participants of the EDC Study for the contributions.
Financial support
Research reported in this study was supported by the National Institute of Diabetes and Digestive and Kidney Diseases at the National Institutes of Health (grant ROI-DK-034818 and ROI-DK-089028) and the Rossi Memorial Fund.
Footnotes
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.atherosclerosis.2018.11.003.
Conflicts of interest
The authors declared they do not have anything to disclose regarding conflict of interest with respect to this manuscript.
References
- [1].Livingston G, Sommerlad A, Orgeta V, Costafreda SG, Huntley J, Ames D, Ballard C, Banerjee S, Burns A, Cohen-Mansfield J, Cooper, Fox N, Gitlin LN, Howard R, Kales HC, Larson EB, Ritchie K, Rockwood K, Sampson EL, Samus Q, Schneider LS, Selbæk G, Teri L, Mukadam N, Dementia prevention, intervention, and care, Lancet 390 (2017) 2673–2734, 10.1016/S140-6736(17)31363-6. [DOI] [PubMed] [Google Scholar]
- [2].Nunley KA, Rosano C, Ryan CM, Jennings JR, Aizenstein HJ, Zgibor JC, Costacou T, Boudreau RM, Miller R, Orchard TJ, Saxton JA, Clinically relevant cognitive impairment in Middle-Aged adults with childhood-onset type I diabetes, Diabetes Care 38 (2015) 1768–1776, 10.2337/dc15-0041. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [3].Lopez OL, Jagust WJ, DeKosky ST, Becker JT, Fitzpatrick A, Dulberg C, Breitner J, Lyketsos C, Jones B, Kawas C, Carlson M, Kuller LH, Prevalence and classification of mild cognitive impairment in the cardiovascular health study cognition study, Arch. Neurol 60 (2003) 1385, 10.1001/archneur.60.10.1385. [DOI] [PubMed] [Google Scholar]
- [4].Ferguson SC, Blane A, Perros P, McCrimmon RJ, Best JJK, Wardlaw J, Deary IJ, Frier BM, Cognitive ability and brain structure in type I diabetes: relation to microangiopathy and preceding severe hypoglycemia, Diabetes 52 (2003) 149–156.. [DOI] [PubMed] [Google Scholar]
- [5].Brismar T, Maurex L, Cooray G, Juntti-Berggren L, indström PI, Ekberg K, Adner N, Andersson S, Predictors of cognitive impairment in type I diabetes, Psychoneuroendocrinology 32 (2007) 1041–1051, 10.1016/j.psyneuen.2007.08.002. [DOI] [PubMed] [Google Scholar]
- [6].Ryan CM, Williams TM, Orchard TJ, Finegold DN, Psychomotor slowing is associated with distal symmetrical polyneuropathy in adults with diabetes mellitus, Diabetes 41 (1992) 107–113.. [DOI] [PubMed] [Google Scholar]
- [7].Rosano C, Watson N, Chang Y, Newman AB, Aizenstein HJ, Du Y, Venkatraman V, Harris TB, Barillas-Mitchell E, Sutton-Tyrrell K, Aortic pulse wave velocity predicts focal white matter hyperintensities in a biracial cohort of older adults, Hypertension 61 (2013) 160–165, 10.1161/HYPERTENSIONAHA.112.198069. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [8].Stefanidis KB, Askew CD, Greaves K, Summers MJ, The effect of non-stroke cardiovascular disease states on risk for cognitive decline and dementia: a systematic and meta-analytic review, Neuropsychol. Rev (2017), 10.1007/s11065-017-9359-z. [DOI] [PubMed] [Google Scholar]
- [9].Osawa K, Nakanishi R, Budoff M, Coronary artery calcification, Glob. Heart 11 (2016) 287–293, 10.1016/j.gheart.2016.08.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Fujiyoshi A, Jacobs DR, Fitzpatrick AL, Alonso A, Duprez DA, Sharrett AR, Seeman T, Blaha MJ, Luchsinger JA, Rapp SR, Coronary artery calcium and risk of dementia in MESA (Multi-Ethnic study of atherosclerosis)CLINICAL PERSPECTIVE, Circ. Cardiovasc. Imag 10 (2017) e00S349,, 10.1161/CIRCIMAGING.116.005349. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Kuller LH, Lopez OL, Mackey RH, Rosano C, Edmundowicz D, Becker JT, Newman AB, Subclinical cardiovascular disease and death, dementia, and coronary heart disease in patients 80 + years, J. Am. Coll. Cardiol 67 (2016) 1013–1022, 10.1016/j.jacc.2015.12.034. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [12].Reis JP, Launer LJ, Terry JG, Loria CM, Zeki Al Hazzouri A, Sidney S, Yaffe K, Jacobs DR, Whitlow CT, Zhu N, Carr JJ, Subclinical atherosclerotic calcification and cognitive functioning in middle-aged adults: the CARDIA study, Atherosclerosis 231 (2013) 72–77, 10.1016/j.atherosclerosis.2013.08.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Hugenschmidt CE, Hsu F-CC, Hayasaka S, Carr JJ, Freedman BI, Nyenhuis DL, Williamson JD, Bowden DW, The influence of subclinical cardiovascular disease and related risk factors on cognition in type 2 diabetes mellitus: the DHS-Mind study, J. Diabet. Complicat 27 (2013) 422–428, 10.1016/j.jdiacomp2013.04.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Sangiorgi G, Rumberger JA, Severson A, Edwards WD, Gregoire J, Fitzpatrick LA, Schwartz RS, Arterial calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans: a histologic study of 723 coronary artery segments using nondecalcifying methodology, J. Am. Coll. Cardiol 31 (1998) 126–133. [DOI] [PubMed] [Google Scholar]
- [15].Ix JH, Miller RG, Criqui MH, Orchard TJ, Test characteristics of the ankle-brachial index and ankle-brachial difference for medial arterial calcification on X-ray in type 1 diabetes, J. Vasc. Surg 56 (2012) 721–727, 10.1016/j.jvs.2012.02.042. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [16].Orchard TJ, Dorman JS, Maser RE, Becker DJ, Drash AL, Ellis D, Laporte RE, Kuller LH, Prevalence of complications in IDDM by sex and duration. Pittsburgh epidemiology of diabetes complications study II, Diabetes 39 (1990) 1116–112.. [DOI] [PubMed] [Google Scholar]
- [17].Wagener DK, Sacks JM, Laporte RE, Macgregor JM, The Pittsburgh study of insulin-dependent diabetes mellitus. Risk for diabetes among relatives of IDDM, Diabetes 31 (1982) 136–144. [DOI] [PubMed] [Google Scholar]
- [18].Nunley KA, Orchard TJ, Ryan CM, Miller R, Costacou T, Rosano, Statin use and cognitive function in middle-aged adults with type 1 diabetes, World J. Diabetes 8 (2017) 286, 10.4239/wjd.v8.i6.286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [19].Dore GA, Elias MF, Robbins MA, Elias PK, Brennan SL, Cognitive performance and age: norms from the Maine-Syracuse Study, Exp. Aging Res 33 (2007) 205–271, 10.1080/03610730701319087. [DOI] [PubMed] [Google Scholar]
- [20].Saxton J, Snitz BE, Lopez OL, Ives DG, Dunn LO, Fitzpatrick A, Carlson MC, DeKosky ST, GEM Study Investigators, Functional and cognitive criteria produce different rates of mild cognitive impairment and conversion to dementia, J. Neurol. Neurosurg. Psychiatry 80 (2009) 737–743, 10.1136/jnnp.2008.160705. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [21].Agatston AS, Janowitz WR, Hildner FJ, Zusmer NR, Viamonte M, Detrano R, Quantification of coronary artery calcium using ultrafast computed tomography, J. Am. coll. Cardiol 15 (1990) 827–832. [DOI] [PubMed] [Google Scholar]
- [22].Callister TQ, Cooil B, Raya SP, Lippolis NJ, Russo DJ, Raggi P, Coronary artery disease: improved reproducibility of calcium scoring with an electron-beam CT volumetric method, Radiology 208 (1998) 807–814, 10.1148/radiology.208.3.9722864. [DOI] [PubMed] [Google Scholar]
- [23].Criqui MH, Denenberg JO, Ix JH, McClelland RL, Wassel CL, Rifkin DE, Carr JJ, Budoff MJ, Allison MA, Calcium density of coronary artery plaque and risk of incident cardiovascular events, J. Am. Med. Assoc 311 (2014) 271, 10.1001/jama.2013.282535. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [24].Radford NB, DeFina LF, Barlow CE, Lakoski SG, leonard D, Paixao ARM, Khera A, Levine BD, Progression of CAC score and risk of IncideCnt CVD, JACC Cardiovasc. Imag 9 (2016) 1420–1429, 10.1016/j.jcmg.2016.03.010. [DOI] [PubMed] [Google Scholar]
- [25].Hokanson JE, MacKenzie T, Kinney G, Snell-Bergeon JK, Dabelea D, Ehrlich J, Eckel RH, Rewers M, Evaluating changes in coronary artery calcium: an analytic method that accounts for interscan variability, AJR Am. J. Roentgenol 182 (2004) 1327–1332, 10.2214/ajr.182.5.1821327. [DOI] [PubMed] [Google Scholar]
- [26].Borhani N, Kass E, Langford H, Payne G, Remington R, Stamler J, The hypertension detection and follow-up program: hypertension detection and follow-up program cooperative group, Prev. Med 5 (1976) 207–215. [DOI] [PubMed] [Google Scholar]
- [27].Guerchet M, Aboyans V, Nubukpo P, Lacroix P, Clément JP, Preux PM, Ankle-brachial index as a marker of cognitive impairment and dementia in general population. A systematic review, Atherosclerosis 216 (2011) 251–257, 10.1016/j.atherosclerosis.2011.03.024. [DOI] [PubMed] [Google Scholar]
- [28].Prince CT, Becker DJ, Costacou T, Miller RG, Orchard TJ, Changes in glycaemic control and risk of coronary artery disease in type 1 diabetes mellitus: findings from the Pittsburgh Epidemiology of Diabetes Complications Study (EDC), Diabetologia 50 (2007) 2280–2288, 10.1007/s00125-007-0797-7. [DOI] [PubMed] [Google Scholar]
- [29].Bucolo G, David H, Quantitative determination of serum triglycerides by the use of enzymes, Clin. Chem 19 (1973) 476–482. [PubMed] [Google Scholar]
- [30].Allain CC, Poon LS, Chan CS, Richmond W, Fu PC, Enzymatic determination of total serum cholesterol, Clin. Chem 20 (1974) 470–475. [PubMed] [Google Scholar]
- [31].Warnick GR, Albers JJ, Heparin-Mn2+ quantitation of high-density-lipoprotein cholesterol: an ultrafiltration procedure for lipemic samples, Clin. Chem 24 (1978) 900–904. [PubMed] [Google Scholar]
- [32].Ellis D, Coonrod BA, Dorman JS, Kelsey SF, Becker DJ, Avner ED, Orchard TJ, Choice of urine sample predictive of microalbuminuria in patients with insulin-dependent diabetes mellitus, Am. J. Kidney Dis 13 (1989) 321–328. [DOI] [PubMed] [Google Scholar]
- [33].Pambianco G, Costacou T, Ellis D, Becker DJ, Klein R, Orchard TJ, 30-year natural history of type 1 diabetes complications: the Pittsburgh epidemiology of diabetes complications study experience, Diabetes 55 (2006) 1463–1469, 10.2337/db05-1423. [DOI] [PubMed] [Google Scholar]
- [34].Mortamais M, Portet F, Brickman AM, Provenzano FA, Muraskin J, Akbaraly TN, Berr C, Touchon J, Bonafé A, le Bars E, Menjot de Champfleur N, Maller JJ, Meslin C, Sabatier R, Ritchie K, Artero S, Education modulates the impact of white matter lesions on the risk of mild cognitive impairment and dementia, Am. J. Geriatr. Psychiatry 22 (2014) 1336–1345, 10.1016/j.jagp.2013.06.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [35].Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman M, de Ferranti S, Després J-P, Fullerton HJ, Howard VJ, Huffman MD, Judd SE, Kissela BM, Lackland DT, Lichtman JH, Lisabeth LD, Liu S, Mackey RH, Matchar DB, McGuire DK, Mohier ER, Moy CS, Muntner P, Mussolino ME, Nasir, Neumar RW, Nichol G, Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Sorlie PD, Stein J, Towfighi A, Turan TN, Virani SS, Willey JZ, Woo D, Yeh RW, Turner MB, American heart association statistics committee and stroke statistics subcommittee, heart disease and stroke statistics—2015 update: a report from the American heart association, Circulation 131 (2015) e29–322, 10.1161/CIR.0000000000000152. [DOI] [PubMed] [Google Scholar]
- [36].Min JK, Lin FY, Gidseg DS, Weinsaft JW, Berman DS, Shaw LJ, Rozanski TQ Callister, Determinants of coronary calcium conversion among patients with a normal coronary calcium scan, J. Am. Coll. Cardiol 55 (2010) 1110–1117, 10.1016/j.jacc.2009.08.088. [DOI] [PubMed] [Google Scholar]
- [37].Vidal J-S, Sigurdsson S, Jonsdottir MK, Eiriksdottir G, Thorgeirsson G, Kjartansson O, Garcia ME, van Buchem MA, Harris TB, Gudnason V, Launer LJ, Coronary artery calcium, brain function and structure, Stroke 41 (2010) 891–897, 10.1161/STROKEAHA.110.579581. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [38].Gleissner CA, Galkina E, Nadler JL, Ley K, Mechanisms by which diabetes increases cardiovascular disease, Drug Discov. Today Dis. Mech 4 (2007) 131–140, 10.1016/j.ddmec.2007.12.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [39].Budoff MJ, Hokanson JE, Nasir K, Shaw LJ, Kinney GL, Chow D, DeMoss D, Nuguri V, Nabavi V, Ratakonda R, Berman DS, Raggi P, Progression of coronary artery calcium predicts all-cause mortality, JACC Cardiovasc. Imag 3 (2010) 1229–1236, 10.1016/J.JCMG.2010.08.018. [DOI] [PubMed] [Google Scholar]
- [40].Swardfager W, Black SE, Coronary artery calcification: a canary in the cognitive coalmine, J. Am. coll. Cardiol 67 (2016) 1023–1026, 10.1016/j.jacc.2016.01.007. [DOI] [PubMed] [Google Scholar]
- [41].Wardlaw JM, Smith EE, Biessels GJ, Cordonnier C, Fazekas F, Frayne R, Lindley RI, O’Brien JT, Barkhof F, Benavente OR, Black SE, Brayne C, Breteler M, Chabriat H, DeCarIi C, de Leeuw F-E, Doubal F, Duering M, Fox NC, Greenberg S, Hachinski V, Kilimann I, Mok V, van Oostenbrugge R, Pantoni L, Speck O, Stephan BCM, Teipel S, Viswanathan A, Werring D, Chen C, Smith C, van Buchem M, Norrving B, Gorelick PB, Dichgans M, Neuroimaging standards for research into small vessel disease and its contribution to ageing and neurodegeneration, Lancet Neurol. 12 (2013) 822–838, 10.1016/S1474-4422(13)70124-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [42].Grimmer T, Faust M, Auer F, Alexopoulos P, Förstl H, Henriksen G, Perneczky R, Sorg C, Yousefi BH, Drzezga A, Kurz A, White matter hyper-intensities predict amyloid increase in Alzheimer’s disease, Neurobiol. Aging 33 (2012) 2766–2773, 10.1016/J.NEUROBI0LAGING.2012.01.016. [DOI] [PubMed] [Google Scholar]
- [43].Demer LL, Tintut Y, Inflammatory, metabolic, and genetic mechanisms of vascular calcification, Arterioscler. Thromb. Vasc. Biol 34 (2014) 715–723, 10.1161/ATVBAHA.113.302070. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [44].Nakahara T, Dweck MR, Narula N, Pisapia D, Narula J, Strauss HW, Coronary artery calcification: from mechanism to molecular imaging, JACC Cardiovasc. Imag 10 (2017) 582–593, 10.1016/j.jcmg.2017.03.005. [DOI] [PubMed] [Google Scholar]
- [45].Chen NX, Moe SM, Vascular calcification: pathophysiology and risk factors, Curr. Hypertens. Rep 14 (2012) 228–237, 10.1007/s11906-012-0265-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [46].Shaw LJ, Raggi P, Berman DS, Callister TQ, Coronary artery calcium as a measure of biologic age, Atherosclerosis 188 (2006) 112–119, 10.1016/j.atheroscIerosis.2005.10.010. [DOI] [PubMed] [Google Scholar]
- [47].Thomas IC, Shiau B, Denenberg JO, McClelland RL, Greenland P, de Boer I-H, Kestenbaum BR, Lin G-M, Daniels M, Forbang NI, Ritkin DE, Hughes-Austin J, Allison MA, Jeffrey J, Ix JH, Criqui MH, Association of cardiovascular disease risk factors with coronary artery calcium volume versus density, Heart 104 (2018) 135–143, 10.1136/heartjnl-2017-311536. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [48].d’AviIa JC, Siqueira LD, Mazeraud A, Azevedo EP, Foguel D, Castro-Faria-Neto HC, Sharshar T, Chrétien F, Bozza FA, Age-related cognitive impairment is associated with long-term neuroinflammation and oxidative stress in a mouse model of episodic systemic inflammation, J. Neuroinnammation 15 (2018) 28, 10.1186/s12974-018-10S9-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [49].Cervellati C, Trentini A, Bosi C, Valacchi G, Morieri ML, Zurlo A, Brombo G, Passaro A, Zuliani G, Low-grade systemic inflammation is associated with functional disability in elderly people affected by dementia, GeroScience 40 (2018) 61–69, 10.1007/s11357-018-0010-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.