Article Text
PDF
Abstract
Introduction There is conflicting evidence whether lower extremity arterial calcification coincides with coronary arterial calcification (CAC). The aims of this study were to investigate the associations between (1) femoral and crural calcification with CAC, and (2) femoral and crural calcification pattern with CAC.
Research design and methods This cross-sectional study included 405 individuals (74% men, 62.6±10.9 years) from the ARTEMIS cohort study at high risk of cardiovascular disease (CVD) who underwent a CT scan of the femoral, crural and coronary arteries. High CVD risk was defined as history/presence of cerebrovascular disease, coronary artery disease, abdominal aortic aneurysm, renal artery stenosis, peripheral artery disease or CVD risk factors: diabetes mellitus type 2, hypertension, hyperlipidemia. Calcification score within each arterial bed was expressed in Agatston units. Dominant calcification patterns (intimal, medial, absent/indistinguishable) were determined via a CT-guided histologically validated scoring algorithm. Multivariable-adjusted multinomial logistic regression analyses were used. Replication was performed in an independent population of individuals with diabetes mellitus type 2 (Early-HFpEF cohort study).
Results Every 100-point increase in femoral and crural calcification score was associated with 1.23 (95% CI=1.09 to 1.37, p<0.001) and 1.28 (95% CI=1.11 to 1.47, p=0.001) times higher odds of having CAC within tertile 3 (high) versus tertile 1 (low), respectively. The association appeared stronger for crural versus femoral arteries. Moreover, the presence of femoral intimal (OR=10.81, 95% CI=4.23 to 27.62, p<0.001), femoral medial (OR=10.37, 95% CI=3.92 to 27.38, p<0.001) and crural intimal (OR=6.70, 95% CI=2.73 to 16.43, p<0.001) calcification patterns were associated with higher odds of having CAC within tertile 3 versus tertile 1, independently from concomitant calcification score. This association appeared stronger for intimal versus medial calcification patterns. The replication analysis yielded similar results.
Conclusions Higher femoral and crural calcification scores were associated with higher CAC. Moreover, the presence of femoral intimal, femoral medial and crural intimal calcification patterns was associated with increased CAC. It appears that arterial calcification is a systemic process which occurs simultaneously in various arterial beds.
- Atherosclerosis
- Coronary Circulation
- Lower Extremity
- Epidemiology
Data availability statement
Data are available upon reasonable request. The informed consent that was signed by the study participants is not compliant with publishing individual data in an open-access institutional repository or as supporting information files with the published paper. However, a data request can be sent to the SMART Steering Committee at uccdatarequest@umcutrecht.nl.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Individuals with diabetes mellitus type 2 or other cardiovascular disease (CVD) risk factors have a high burden of arterial calcifications, which pose them at high risk of atherosclerotic events, such as infarction and stroke, and arteriosclerotic events such as lower limb amputation.
WHAT THIS STUDY ADDS
This study shows that arterial calcification within the lower extremities coincides with coronary arterial calcification, which indicates that arterial calcification is a systemic process which occurs simultaneously throughout the whole body. This may explain why individuals with diabetes mellitus type 2 or other CVD risk factors with arterial calcification have a threefold to fourfold increased risk of CVD compared with individuals without arterial calcification.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This research underlines the importance of curbing the burden of arterial calcification within individuals with diabetes mellitus type 2 or other CVD risk factors in order to reduce CVD mortality and morbidity. Future research should focus on prevention and treatment of arterial calcification, while practice should focus on detection, prevention and treatment.
Background
Arterial calcification in any arterial wall accounts for a threefold to fourfold increased risk of cardiovascular disease (CVD) and mortality.1 The majority of research on calcification morphology has been devoted to the coronary arteries, carotid arteries and aorta, but less is known on calcification morphology in arteries of the lower extremities and their contribution to total calcification burden. It is known that calcification of the intimal layer is more often observed in the coronary arteries, common carotid arteries and aorta, and is associated with atherosclerotic outcomes such as myocardial infarction and stroke.2 In contrast, calcification of the medial layer is more often observed in arteries of the breast, carotid siphon and lower extremities, and is associated with vascular stiffness which may give rise to lower limb amputation among others.3
So far, it is unclear whether lower extremity arterial calcification (LEAC) coincides with coronary arterial calcification (CAC). A few studies investigated the association between LEAC and CAC. However, the validity of these study results was limited by the small sample size4–7 and the use of specific study populations such as hemodialysis patients and patients with type 1 diabetes mellitus (T1DM).5 6 Moreover, no prior studies investigated whether the presence of either intimal arterial calcification (IAC) or medial arterial calcification (MAC) patterns in the lower extremities was associated with CAC, independently from concomitant calcification score. Consequently, there is a need for a large-scale study investigating the association between LEAC, in terms of score and pattern, with CAC in a population at high risk of CVD. Understanding these associations may help to establish whether LEAC contributes to calcification burden and CVD risk independently from CAC. In addition, it may shed light on identification of individuals with pronounced CAC based on the presence of LEAC.
It is hypothesized that shear stress and atherosclerotic risk factors (eg, smoking and hyperlipidemia) have systemic consequences and therefore calcification of the femoral and crural arteries should be associated with higher CAC. Moreover, since IAC and MAC share various risk factors, it is thought that the presence of either IAC or MAC in the lower extremities is associated with increased CAC, with the association being stronger for IAC compared with MAC.8
This study aims to investigate the associations between (1) femoral and crural calcification with CAC and (2) femoral and crural calcification patterns with CAC, independently from concomitant calcification score, in a population at high risk of CVD. Cross-sectional data will be used from the ARTEMIS Study, which is a substudy with participants from the ongoing Second Manifestations of ARTerial Disease (SMART) cohort study and the ongoing Hoorn Diabetes Care System (DCS) cohort study.
Materials and methods
Study population and design
Data were used from two ongoing cohort studies, the DCS cohort9 and the SMART cohort,10 together forming the ARTEMIS Study.8
The DCS cohort consists of over 15 000 individuals aged 50–75 years with type 2 diabetes mellitus (T2DM) from the West-Friesland region in the Netherlands. These individuals were referred to the DCS Study centre by their general practitioner. The DCS Study centre provides monitoring of glycemic control and diabetes-related risk factors and complications. Details of the DCS are described elsewhere.9 Individuals from the DCS cohort are representative of Western-European semiurban individuals with T2DM and are therefore considered to be at increased risk of CVD.9
The SMART cohort consists of over 13 000 individuals aged 18–79 years who were referred to the University Medical Centre Utrecht since 1996 with or at high risk of CVD. High risk of CVD was defined as having a history of cerebrovascular disease, coronary artery disease (CAD), abdominal aortic aneurysm, renal artery stenosis, peripheral artery disease (PAD) or having CVD risk factors such as T2DM, hypertension, hyperlipidemia, renal insufficiency and/or family history of CVD. Details of the SMART cohort are described elsewhere.10 Attributable to the wide age range and the liberal definition of high CVD risk, individuals from the SMART cohort are representative for a nationwide or even Western-European population of individuals at high CVD risk.11
A total of 718 individuals (DCS N=198, SMART N=520) were included in the ARTEMIS Study as they underwent a high-resolution quantitative CT (HR-qCT) scan of the femoral arteries and crural arteries in both lower limbs between 2015 and 2017. Because of logistic and financial reasons, only a subset of individuals were invited to participate in the ARTEMIS Study. However, baseline characteristics were similar for SMART and SMART-ARTEMIS participants,8 11 as well as for DCS and DCS-ARTEMIS participants.9 12 Of these 718 individuals, 405 individuals (DCS N=198, SMART N=207) also received an HR-qCT scan of the coronary arteries and were included in the analysis. Individuals with bilateral lower limb amputation at the moment of inclusion into the ARTEMIS Study were excluded. Figure 1 portrays the flow diagram.
Flow diagram of the Diabetes Care System (DCS) and Second Manifestations of ARTerial Disease (SMART) cohorts forming the ARTEMIS cohort study with replication by the Early-HFpEF cohort study.
In a replication analysis, CT data of 600 individuals from the Early-HFpEF Study were used. The Early-HFpEF Study consists of 848 individuals with T2DM ranging 50–75 years. These participants also originated from the DCS cohort and are therefore at high risk of CVD. Main reasons for replication were the availability of reliability data on calcification measurements and validation of conclusions drawn from the main analyses.
Measurement of calcification score
The calcification score was derived from a CT scan of the femoral, crural and coronary arteries using a locally developed software tool (iX Viewer, Utrecht University Medical Centre). All participants underwent an unenhanced CT scan of either the full body or the legs (slice thickness 1 mm, increment 0.7 mm, resampled to 5 mm slices with 4 mm increment). A calcific lesion was considered as calcification when the CT density exceeded 130 Hounsfield units. The amount of calcification in the coronary arteries was measured using the Agatston method.13 The Agatston method is the most commonly accepted method for determination of CAC. Femoral and crural calcification scores were also measured using the Agatston method after resampling to a higher slice thickness. Subsequently, the calcification scores of the left and right femoral arteries were averaged, just as the calcification scores of the left and right crural arteries. The main outcome of this study was CAC and due to the non-normal distribution of calcification scores, CAC scores were normalized by categorization into tertiles. The quantitative scoring of LEAC and CAC was performed by an experienced researcher (EJB) who was blinded to patient characteristics. This researcher was trained by an experienced radiologist (PAdJ). Data on reliability parameters within the ARTEMIS Study were unknown.
Within the Early-HFpEF Study, determination of LEAC score was performed by a researcher (RM) and two research assistants who were trained by PAdJ. The inter-rater correlation coefficient was 0.922 (0.853–0.959) for the femoral arteries and 0.890 (0.793–0.942) for the crural arteries, which both reflect excellent reliability. Finally, determination of CAC score was done by a researcher (RM) and a research assistant, who were trained by PAdJ. The inter-rater correlation coefficient of the CAC measurement was 0.997 (0.994–0.999), which reflects nearly perfect reliability.
Measurement of calcification pattern
The calcification pattern was derived from the femoral and crural arteries. A CT-guided scoring algorithm was used to identify the pattern of calcification in these arteries.14 Points were assigned to three domains of the calcification: circularity (0=absent, 1=dot, 2=<90°, 3=90–270° or 4=270–360°), thickness (0=absent, 1=≥1.5 mm, or 3=<1.5 mm) and morphology (0=indistinguishable, 1=irregular/patchy or 4=continuous). The points per domain were summed for determination of the total pattern score. Dominant calcification pattern was categorized as: ‘absent’ (0 points), ‘indistinguishable’ (0 points for morphology, regardless of points for circularity and thickness), ‘dominant IAC’ (<7 points) or ‘dominant MAC’ (≥7 points). The qualitative scoring of calcifications was performed by PAdJ who was blinded to patient characteristics. The scoring algorithm was developed by Kockelkoren et al and was validated in the intracranial internal carotid artery.14 This study showed that the inter-rater reliability was 0.72 (0.60–0.84) and the intrarater reliability was 0.82 (0.73–0.89).14 In 2022, the scoring algorithm was validated versus histology in the crural arteries and more than 70% of the arteries were correctly classified as ‘absent’, ‘dominant IAC’ or ‘dominant MAC’.15 Moreover, the absolute agreement for IAC was 0.47, while this was 0.42 for MAC.15
Within the Early-HFpEF Study, determination of calcification pattern was done by a researcher (RM) and two research assistants who were trained by PAdJ. The inter-rater kappa was 0.59 (0.55–0.64) and 0.57 (0.53–0.62) for the femoral and crural arteries, respectively. Both agreements reflect moderate reliability.
Covariates
Medical history on cerebrovascular disease, CAD, aneurysm of the abdominal aorta and PAD (all yes/no) was obtained via a comprehensive screening at baseline (SMART) or via self-report during routine care visits (DCS). Smoking status was self-reported (never/former/current smoker). Moreover, data on the use of antihypertensive medication and lipid-lowering medication (both yes/no) were collected via self-report and medication lists. Data on diabetes status (yes/no) were collected in the SMART cohort when there was a referral diagnosis or self-report of diabetes mellitus, the participant used glucose-lowering medication and/or when there was a baseline fasting plasma glucose level ≥7 mmol/L. Physical examinations were performed in order to obtain weight, height, systolic blood pressure (SBP) and diastolic blood pressure (DBP). The weight (kg) and height (m) were used to calculate the body mass index (BMI) (kg/m2). The SBP (≥140 mm Hg) and DBP (≥90 mm Hg) were used along with the use of antihypertensive medication (yes/no) to define hypertension (yes/no) if at least one of these three criteria was fulfilled.
Within the Early-HFpEF Study, only data on age and sex were available, as the study was still ongoing in 2023.
Biochemical measurements
Fasting blood samples were drawn and analyzed for measurement of total cholesterol (mmol/L), high-density lipoprotein-cholesterol (mmol/L), low-density lipoprotein (LDL)-cholesterol (mmol/L), triglycerides (mmol/L), glucose (mmol/L), HbA1c (mmol/mol), C reactive protein (mg/L) and creatinine levels (µmol/L) using standard laboratory techniques.9 10 The CKD-EPI formula was used to estimate the glomerular filtration rate (eGFR) (mL/min/1.73 m2).16 Hyperlipidemia was dichotomously defined as having a total cholesterol ≥5.0 mmol/L, an LDL-cholesterol ≥3.2 mmol/L and/or using lipid-lowering medication.
Statistical analysis
Baseline characteristics were described for the full cohort and per CAC tertile. Continuous variables were expressed as mean (SD) or median (IQR) depending on their distribution. Categorical variables were presented as frequencies (group percentage). Differences between these CAC tertiles were tested with a Χ2 test for categorical variables and with an analysis of variance (ANOVA) (or Kruskal-Wallis test in case of violation of assumptions for ANOVA) for continuous variables.
The associations between femoral and crural calcification scores with CAC score were examined using a multinomial logistic regression analysis. Calcifications in the femoral and crural arteries were modeled continuously with increments of 100 points in calcification score, while CAC score was normalized by categorization into tertiles 1 (low; reference), 2 (medium) and 3 (high). No normalization of femoral and crural scores was needed as multinomial regression analysis does not require normalization of exposure variables.
The associations between calcification pattern in the femoral and crural arteries with CAC score were evaluated using a multinomial logistic regression analysis. Both the femoral and crural calcification patterns were regarded as exposure and were categorized as ‘absent/indistinguishable’ (reference), ‘dominant IAC’ and ‘dominant MAC’. Again, CAC score was categorized into tertiles.
Confounders were selected based on literature. Model 1 adjusted for age, sex, cohort, and femoral and crural Agatston score (per 1-point increment; only for analysis on calcification pattern). Model 2 included adjustment for variables in model 1 along with smoking status and BMI. Finally, model 3 included adjustment for model 2 along with diabetes mellitus, hypertension, hyperlipidemia and eGFR (per 1-point increment).
Effect modification by sex, smoking status, history of CVD, history of diabetes, use of antihypertensive medication and use of lipid-lowering medication was checked. This was evaluated by adding an interaction term to the final model. If the p-for-interaction values were below 0.1, then the analysis was stratified accordingly.
Effect estimates consisted of ORs, the respective 95% CIs, and the corresponding p values. Moreover, goodness of fit of the final model was tested using the Hosmer-Lemeshow test and described with the McFadden’s R2.
The two main analyses were replicated using data from the Early-HFpEF Study. Only adjustment for age and sex was performed as data on all other covariates were still unavailable within the Early-HFpEF Study at moment of data analysis.
The correlation between calcification scores in each arterial bed was described using the Spearman’s rank-order correlation test and was visually presented with two-way scatterplots. Moreover, the correlation between calcification pattern in the femoral arteries with calcification pattern in the crural arteries was described using the Cramér’s V test. For the latter correlation, the calcification patterns were categorized as ‘absent’, ‘indistinguishable’, ‘dominant IAC’ and ‘dominant MAC’.
Imputation of calcification scores was considered infeasible as most of the missing scores were attributable to technical issues or presence of prostheses. As such, complete case analysis was applied on data from participants who had calcification data on both femoral and coronary arteries (n=383), on both crural and coronary arteries (n=395) and on all three arteries (n=380). Two participants were excluded because they had missing data on smoking status and BMI. All analyses were carried out using STATA software (V.17.0, StataCorp, College Station, USA). A probability value of 0.05 was used as a cut-off value to assess statistical significance, unless stated otherwise.
Results
Baseline and calcification characteristics
The total cohort consisted of 76.9% males with a mean age of 62.0±10.6 years (table 1). Moreover, 59.6% of the cohort had hypertension, while 34.5% had hyperlipidemia. Across increasing CAC categories, the percentage of male sex and mean age increased. Within the full cohort, 38.0% and 27.9% showed a dominant IAC and a dominant MAC pattern in the femoral arteries, respectively (table 2). The prevalence of a dominant IAC and a dominant MAC pattern was somewhat lower in the crural arteries (30.4% and 25.2%, respectively). The prevalence of both IAC and MAC increased across increasing CAC categories in both the femoral and crural arteries. Furthermore, the full cohort showed a median calcification score of 143 (16–645) and 74 (9–364) in the femoral and crural arteries, respectively. Both calcification scores showed a tendency to increase along increasing CAC categories.
Baseline characteristics described for the full cohort and per coronary arterial calcification (CAC) category
Femoral and crural calcification characteristics described for the full cohort and per coronary arterial calcification (CAC) category
The study population characteristics of the Early-HFpEF cohort are shown in online supplemental table 1. The Early-HFpEF Study population (61.5% males, mean age 67.8±6.1 years) had slightly less males and a slightly higher age compared with the study participants from the ARTEMIS Study. The prevalence of IAC (femoral=22.0%; crural=12.2%) was lower, while the prevalence of MAC (femoral=40.0%; crural=28.2%) and median calcification scores (femoral=658 (197–2269); crural=223 (69–985)) were higher in the Early-HFpEF Study compared with the ARTEMIS Study.
Supplemental material
Association between femoral and crural calcification scores with CAC score
Adjusted according to model 3, every 100-point increase in femoral calcification score was significantly associated with 1.17 (95% CI 1.05 to 1.31, p=0.004) and 1.23 (95% CI 1.09 to 1.37, p<0.001) times higher odds of having a CAC score within middle and highest tertile, versus lowest tertile, respectively (table 3). The Hosmer-Lemeshow p value was 0.419 and the McFadden’s R2 was 0.214, indicating a good fit.
Associations between every 100-unit increase in calcification score within the femoral arteries (upper half) and within the crural arteries (lower half) with coronary arterial calcification (CAC) categories
Similarly, every 100-point increase in crural calcification score was associated with 1.25 (95% CI 1.08 to 1.44, p=0.002) and 1.28 (95% CI 1.11 to 1.47, p=0.001) times higher odds of having a CAC score within middle and highest tertile, versus lowest tertile, respectively (table 3: model 3). The Hosmer-Lemeshow p value was 0.946 and the McFadden’s R2 was 0.189, again indicating a good fit.
No effect modification by sex, smoking status, manifest CVD, history of diabetes, use of antihypertensive medication and use of lipid-lowering medication was observed (data not shown).
Association between femoral and crural calcification patterns with CAC score
Significant associations were observed for calcifications of both the femoral and crural arteries, where the effect estimates were slightly stronger, although not significantly, for IAC compared with MAC (table 4).
Associations between calcification pattern (IAC, MAC, absent/indistinguishable) within the femoral and crural arteries with coronary arterial calcification (CAC) categories
Having an IAC or an MAC pattern in the femoral arteries was significantly associated with higher odds of having a CAC score within tertile 3 (high) versus tertile 1 (low) (OR=10.81; 95% CI 4.23 to 27.62, p<0.001; OR=10.37, 95% CI 3.92 to 27.38, p<0.001) in model 3, respectively. These effect estimates were weaker, although still statistically significant when comparing IAC and MAC in tertile 2 (medium) with tertile 1. The Hosmer-Lemeshow p value was 0.174 and the McFadden’s R2 was 0.267, indicating a good fit.
Similar associations were observed in the crural arteries. The presence of an IAC or an MAC pattern in the crural arteries was significantly associated with higher odds of having a CAC score within tertile 3 versus tertile 1 (OR=6.70; 95% CI 2.73 to 16.43, p<0.001; OR=5.10, 95% CI 2.03 to 12.80, p=0.001) in model 3, respectively. When comparing tertile 2 with tertile 1, only the association for IAC remained significant (OR=2.35; 95% CI 1.05 to 5.29, p=0.038), whereas this was statistically insignificant for MAC (OR=1.54; 95% CI 0.63 to 3.74, p=0.339). Again, effect estimates were stronger for IAC compared with MAC. The Hosmer-Lemeshow p value was 0.235 and the McFadden’s R2 was 0.250, again indicating a good fit.
No effect modification by sex, smoking status, manifest CVD, history of diabetes, use of antihypertensive medication and use of lipid-lowering medication was observed (data not shown).
Correlations between femoral, crural and coronary calcification scores
The Spearman’s rank-order correlation test showed that femoral calcification score was strongly correlated with CAC score (r=0.627; p<0.001) and with crural calcification score (r=0.758; p<0.001). Moreover, crural calcification score was positively associated with CAC score (r=0.575; p<0.001). The corresponding two-way scatterplots are shown in online supplemental figure 1.
Supplemental material
Correlation between femoral calcification pattern and crural calcification pattern
The Cramér’s V test showed that femoral calcification pattern was strongly correlated with crural calcification pattern. The corresponding Cramér’s V was 0.517.
Replication analysis
According to the Early-HFpEF data, every 100-point increase in femoral calcification score was significantly associated with 1.04 (95% CI 1.02 to 1.06, p<0.001) and 1.07 (95% CI 1.05 to 1.09, p<0.001) times higher odds of having a CAC score within second and third tertile versus first tertile (online supplemental table 2: model 1). Similarly, every 100-point increase in crural calcification score was significantly associated with 1.07 (95% CI 1.02 to 1.11, p=0.003) and 1.09 (95% CI 1.04 to 1.14, p<0.001) times higher odds of having a CAC score within second and third tertile versus first tertile (online supplemental table 2: model 1). Similar to the ARTEMIS data, these associations were stronger within the crural compared with the femoral arteries.
Supplemental material
Having an IAC or an MAC pattern in the femoral arteries was significantly associated with higher odds of having a CAC score within tertile 3 versus tertile 1 (OR=6.26; 95% CI 3.02 to 12.97, p<0.001; OR=3,45, 95% CI 1.81 to 6.56, p<0.001), respectively (online supplemental table 3: model 1). These effect estimates were weaker, although still statistically significant, when comparing IAC and MAC in tertile 2 (medium) with tertile 1. Similar associations were observed in the crural arteries. The presence of an IAC or an MAC pattern in the crural arteries was significantly associated with higher odds of having a CAC score within tertile 3 versus tertile 1 (OR=13.41; 95% CI 4.30 to 41.84, p<0.001; OR=2.68, 95% CI 1.36 to 5.28, p=0.005), respectively. When comparing tertile 2 with tertile 1, only the association for IAC remained significant (OR=4.65; 95% CI 1.50 to 14.41, p=0.008), whereas this was statistically insignificant for MAC (OR=0.78; 95% CI 0.40 to 1.54, p=0.479). Similar to the ARTEMIS data, effect estimates were stronger for IAC compared with MAC.
Supplemental material
Discussion
Main findings
This study showed that increased calcification in both the femoral and crural arteries was associated with increased CAC, with the association appearing stronger within the crural versus femoral arteries. Moreover, the presence of either a dominant IAC pattern or a dominant MAC pattern in the femoral arteries and crural arteries was associated with increased CAC, independently of concomitant calcification score, with associations appearing stronger for IAC compared with MAC.
Comparison with previous literature
The findings of the association between femoral and crural calcification score with CAC are partially in agreement with a previous study conducted by Aly et al. This research group investigated in patients with CAD whether calcification score in various arterial beds was correlated with the Gensini score, which is a measure of luminal narrowing within the coronary arteries (with a higher score reflecting more luminal narrowing).7 This study showed that femoral calcification score positively correlated with Gensini score (r=0.3, p=0.007). However, the same study showed that there was no significant correlation between crural calcification score and Gensini score (r=0.1, p=0.386). Although the study by Aly et al did not specifically use CAC as outcome, the Gensini score may be considered a proxy for CAC as these two variables are found to be strongly correlated and both variables characterize coronary artery health.17 However, Aly et al showed that the association between LEAC and CAC was stronger for the femoral arteries compared with the crural arteries, whereas within the present study, this was the opposite. This discrepancy may be attributable to differences in study populations and data collection methods.
Moreover, the study findings seem to be corroborated by a study conducted by Shin et al in patients with PAD.4 This study showed that patients with PAD with single-vessel or multivessel CAD had significantly more calcification in the femoral arteries and crural arteries compared with individuals with non-significant CAD. Similar to our study, the associations were stronger in the crural arteries compared with the femoral arteries.
In addition, a study by Patsch et al investigated whether crural calcification score correlated with CAC score in patients with chronic kidney disease treated with hemodialysis.5 This study demonstrated that the crural calcification score highly correlated with CAC score (r=0.6, p<0.001), which was similar to our findings.
Finally, a longitudinal study showed that LEAC was positively but not significantly associated with CAC over a period of 6 years.6 This study was conducted in patients with T1DM. However, no distinction was made between the upper and lower leg.
To our knowledge, this is the first study to examine the association between the presence of IAC and MAC-dominant calcification patterns in the lower extremities with CAC, independently from calcification score in the lower extremities. Therefore, we cannot compare the results of these analyses with previous literature. We have observed that the presence of a dominant IAC (in both femoral and crural arteries) and dominant MAC (only in the femoral artery) pattern was associated with increased CAC, independently of total calcification in these arteries. One study by Lilly et al showed that an elevated ankle-brachial index (>1.4), as a proxy for MAC, was associated with higher CAC in individuals with T2DM.18 However, the study by Hoek et al showed that an elevated ankle-brachial index is associated with MAC but cannot be used as a proxy for the presence of MAC in a group of individuals at high risk of CVD.19 Nevertheless, the current study implies that either the presence of IAC or MAC within the femoral and crural arteries is associated with increased CAC after calcification score has been taken into account. We have observed that the association between calcification pattern with CAC was slightly stronger for IAC compared with MAC, which seems to be corroborated by the fact that the coronary arteries show a dominant IAC pattern.20
Relevance
Results of this study may be useful in the search for new markers of CVD. Arterial calcification is often present in early stages of atherogenesis21 and may thus pose as an early marker of atherosclerotic outcomes within the coronary arteries.7 The study results imply that calcifications within the femoral arteries and crural arteries may act as a marker of increased CAC and thus an increased risk of coronary events. Moreover, the study results imply that the presence of a dominant IAC pattern in the upper and lower leg and a dominant MAC pattern in the upper leg may be indicative of increased CAC. Future research should point out whether there is indeed a causal association between calcifications in these two arterial beds and whether the assessment of LEAC is a valid tool for predicting CAC. In such instance, data obtained from CT scans of the lower extremities may provide vital information on CAC without subjecting the individual to additional testing. The results could, in addition to systemic atherosclerosis burden, also be explained by variation between persons in their tendency to calcify. Results from a study by Eelderink et al suggest that some individuals have a high tendency to calcify both in the intima and the media and in multiple vascular beds beyond atherosclerosis burden.22
Strengths and limitations
This study has several strengths. To start, this study had a large sample size of individuals with or at high risk of CVD. Another strength of the study was the measurement of calcification pattern and calcification score within the same population, which is rare in studies on arterial calcification. Moreover, we used a CT-guided scoring algorithm to characterize dominant calcification patterns, which was histologically validated in the femoral and crural arteries. Finally, there were comprehensive data on confounders (0.5% missing) and replication in an independent population.
This study also has limitations. First, this study used cross-sectional data, which provide no evidence for potential causal associations. Due to logistic reasons, 313 out of the 718 individuals in the ARTEMIS Study did not receive an HR-qCT scan of the coronary arteries, which limited the power for the analyses. Moreover, LEAC score data were missing in 25 individuals out of 405 with CAC data (6%), mostly attributable to insufficient quality of the scans (due to technical issues or scatter caused by prostheses), and data on reliability of the available LEAC score data were unknown. However, the low amount of missing data (6% of calcification data; 0.5% on confounder data) makes selection bias minimal. Finally, history of manifest CVD was self-reported along with smoking status and medication use. This data collection method is subjective to recall bias and social desirability bias. As such, true prevalences of these characteristics may be higher than observed.
Conclusion
In conclusion, this study showed that increased femoral and crural calcification were associated with higher CAC. Moreover, the presence of intimal or medial calcification patterns in the femoral arteries and the presence of an intimal calcification pattern in the crural arteries were associated with higher CAC, independently of calcification scores within these arteries. It appears that arterial calcification is a systemic process which is reflected by simultaneous calcification in various arterial beds. Additional research focusing on causality and the diagnostic importance of this association is warranted to establish the exact role of arterial calcifications within the lower extremities with regard to coronary calcification burden.
Data availability statement
Data are available upon reasonable request. The informed consent that was signed by the study participants is not compliant with publishing individual data in an open-access institutional repository or as supporting information files with the published paper. However, a data request can be sent to the SMART Steering Committee at uccdatarequest@umcutrecht.nl.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants. The cross-sectional ARTEMIS Study was approved by the medical ethics review board of the University Medical Centre Utrecht (14/444). All participants provided written informed consent before participation in the ARTEMIS Study.
Acknowledgments
We gratefully acknowledge the contribution of the Utrecht Cardiovascular Cohort-Second Manifestations of ARTerial Disease Study group (UCC-SMART Study group): MJ Cramer, HM Nathoe and MG van der Meer (co-PI), Department of Cardiology; GJ de Borst and M Teraa (co-PI), Department of Vascular Surgery; ML Bots and M van Smeden, Julius Center for Health Sciences and Primary Care; MH Emmelot-Vonk, Department of Geriatrics; PA de Jong, Department of Radiology; AT Lely, Department of Gynaecology and Obstetrics; NP van der Kaaij, Department of Cardiothoracic Surgery; LJ Kappelle and YM Ruigrok, Department of Neurology; MC Verhaar, Department of Nephrology & Hypertension; JAN Dorresteijn (co-PI) and FLJ Visseren (PI), Department of Vascular Medicine, UMC Utrecht, University Medical Centre Utrecht and Utrecht University. Finally, we want to thank S Fekecs, TT Wong, I Oughzou and ML Romero Prats (Department of Epidemiology & Data Science, Amsterdam UMC) for their help in data collection on calcification within the Early-HFpEF Study.
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Collaborators Utrecht Cardiovascular Cohort-Second Manifestations of ARTerial Disease Study group (UCC-SMART Study group): MJ Cramer, HM Nathoe, MG van der Meer, GJ de Borst, M Teraa, ML Bots, M van Smeden, MH Emmelot-Vonk, PA de Jong, AT Lely, NP van der Kaaij, LJ Kappelle, YM Ruigrok, MC Verhaar, JAN Dorresteijn, FLJ Visseren.
Contributors RM contributed to conception and design, data analysis and interpretation, drafted the manuscript and critically revised it. PAdJ, PJME and JB contributed to conception and design, acquisition, data analysis and interpretation and critically revised the manuscript. AGH and EJB contributed to conception and design and acquisition, and critically revised the manuscript. TD contributed to conception and acquisition. In order to ensure integrity and accuracy, all authors gave final approval and agreed to be accountable for all aspects of this work. JB is guarantor of this work.
Funding JB is supported by a ZonMw NWO-Vidi grant (91718304).
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.