Association of aortic perivascular adipose tissue density with aortic calcification in women with systemic lupus erythematosus

Accepted Manuscript Association of aortic perivascular adipose tissue density with aortic calcification in women with systemic lupus erythematosus Kelly J. Shields, Samar R. El Khoudary, Joseph M. Ahearn, Susan Manzi PII:

S0021-9150(17)30185-5

DOI:

10.1016/j.atherosclerosis.2017.04.021

Reference:

ATH 15041

To appear in:

Atherosclerosis

Received Date: 12 January 2017 Revised Date:

17 April 2017

Accepted Date: 28 April 2017

Please cite this article as: Shields KJ, El Khoudary SR, Ahearn JM, Manzi S, Association of aortic perivascular adipose tissue density with aortic calcification in women with systemic lupus erythematosus, Atherosclerosis (2017), doi: 10.1016/j.atherosclerosis.2017.04.021. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT

1

Association of aortic perivascular adipose tissue density with aortic calcification in women with

2

systemic lupus erythematosus

RI PT

3 Kelly J Shields*, Samar R. El Khoudaryb, Joseph M Ahearna, Susan Manzi,a

4 5

7 8 9

Lupus Center of Excellence, Autoimmunity Institute, Department of Medicine, Allegheny Health

SC

a

Network, PA, USA b

University of Pittsburgh, Graduate School of Public Health, Department of Epidemiology, PA,

M AN U

6

USA

10

*Corresponding author: 320 E North Avenue, Allegheny Health Network, Allegheny General Hospital

12

8th Floor South Tower, Pittsburgh, Pennsylvania 15212. Email address: [email protected] (K. J

13

Shields)

TE D

11

EP

14 Abstract

16

Background and aims: Women with systemic lupus erythematosus (SLE) have an increased risk of

17

cardiovascular disease (CVD) . Perivascular adipose tissue (PVAT) surrounds the vascular wall, is

18

associated with CVD risk factors, and may contribute to premature CVD in SLE. We previously found

19

greater volumes of aortic PVAT associated with aortic calcification (AC) in female SLE patients. There is

20

recent evidence that not only volume but adipose density may also be indicative of inflammation. We

21

hypothesized that female SLE patients would have a difference in aPVAT quality associated with AC that

22

is independent of volume.

AC C

15

ACCEPTED MANUSCRIPT

Methods: Aorta PVAT quality was evaluated using the average radiodensity (density) of adipose tissue-

2

specific Hounsfield Units (-190 to -30 HU) within each clinical CT scan of CVD-free, age-/race- matched

3

SLE women (n=143) and healthy controls (HC, n=143).

4

Results: Aorta PVAT density was significantly higher in SLE (mean (SD): (-83.6 (1.9) HU) versus HC (-

5

84.1 (1.8) HU) , p=0.03). Increasing aPVAT volume was correlated with denser aPVAT in SLE (ρ, p-value:

6

0.75,<0.0001) and HC (0.74,<0.0001). Increasing AC score (log) was correlated with denser aPVAT for

7

SLE (0.31, 0.0005) and HC (0.23, 0.008). In linear regression, denser aPVAT was more strongly

8

associated with AC score in SLE (β (SE) : 0.445 (0.11) , p<0.0001) versus HC (0.335 (0.12), p=0.006)

9

independent of age, circulating inflammatory markers, CVD risk factors and BMI (p<0.05), but was

M AN U

SC

RI PT

1

attenuated with aPVAT volume (p=0.3).

11

Conclusions: Denser aPVAT is associated with aPVAT volume and AC in SLE women. Adjusting for aPVAT

12

volume attenuated the detected association between aPVAT density and AC, which may be indicative of

13

adipose dysfunction.

14

TE D

10

Key Word: Subclinical cardiovascular disease, Women, Computerized tomography, Vascular

16

calcification, Perivascular adipose tissue, Adipose quality, Epidemiology.

AC C

17

EP

15

2

ACCEPTED MANUSCRIPT

1 Introduction

3

Systemic lupus erythematosus (SLE, lupus) is an autoimmune disease characterized by chronic systemic

4

inflammation. Women with SLE have a greater risk of developing cardiovascular disease (CVD) or

5

experiencing a cardiovascular (CV) event when compared to healthy controls.1 The elevated CVD risk is

6

only partially explained by traditional CVD risk factors.2

7

Adipose tissue is a heterogeneous mix of immune regulating cells and adipocytes, containing either

8

single, large lipid droplets or many small lipid droplets along with differences in vascularity and

9

mitochondrial concentration. Given this variance in adipose tissue content, not only is the volume

10

relevant to CVD pathogenesis, but the quality of the adipose may be just as important (Fig. 1A). The

11

quality of adipose may be quantified using clinical CT scans by thresholding for adipose tissue specific

12

radiodensity (density) (adipose tissue: -190 to -30 Hounsfield Units (HU)) (Fig. 1A and B). Previous

13

studies evaluating abdominal visceral and subcutaneous quality using density measures obtained from

14

CT scans have shown less dense adipose tissue associated with cardiometabolic risk3, 4 and incident CVD

15

in age- and sex-adjusted models5.

16

The unique aspect of perivascular adipose tissue (PVAT) versus abdominal visceral or subcutaneous

17

adipose is the proximity to the vascular wall and the absence of any fascial boundary between the PVAT

18

and the vascular wall adventitia. More recent studies have begun to evaluate the association between

19

pericoronary (epicardial and paracardial) adipose tissue volume and density quantified using clinical CT

20

scans with high risk coronary plaques and coronary artery calcification (CAC) ultimately endeavoring to

21

establish clinically relevant differences in PVAT volume and density.6-8 An inflammatory PVAT

22

environment may have a direct, localized effect on the vascular wall participating in CVD pathogenesis.

23

Adipose density and the association with CVD may be location dependent. Perivascular adipose tissue

AC C

EP

TE D

M AN U

SC

RI PT

2

3

ACCEPTED MANUSCRIPT

density may be uniquely different when compared to the abdominal visceral and subcutaneous adipose

2

tissue density.

3

We have previously found that clinically CVD-free women with SLE have greater volumes of aortic

4

perivascular adipose tissue (aPVAT) surrounding the descending thoracic aorta when compared to

5

healthy controls, which is associated with subclinical vascular calcification.9 Greater volumes of other

6

small visceral adipose depots including epicardial adipose tissue have been associated with CVD risk

7

factors, CAC, future CVD events, and SLE.10-14 The quality of aPVAT measured by density and the

8

association with aorta calcification (AC) has never been characterized in SLE patients. Based on previous

9

studies examining the association between subcutaneous and visceral abdominal adipose tissue density

M AN U

SC

RI PT

1

10

and CVD, we hypothesized that female SLE patients would have less dense aPVAT associated with AC,

11

independent of aPVAT volume.

12 Materials and methods

14

Inclusion/exclusion criteria

15

This retrospective case-control study included female SLE patients and their respective age- and race-

16

matched controls (1:1) as previously described.15 Briefly, women who fulfilled the 1987 revised American

17

College of Rheumatology criteria for SLE but with no prior history of CV event (myocardial infarction,

18

angina, stroke, cerebrovascular event, transient ischemic attack or coronary revascularization)

19

confirmed through medical records were non-selectively recruited from the Pittsburgh Lupus Registry to

20

participate in this study. Healthy, non-SLE women were to be matched on gender, age (±5 years) , race

21

and recruited based on: 1) Voters registration list or Motor Vehicles License list depending on which list

22

the SLE woman was found or 2) direct sample neighborhood control (geographic location) if the case

23

(SLE woman) was not on the previous lists. The women were participants of the “Heart Effects on

AC C

EP

TE D

13

4

ACCEPTED MANUSCRIPT

Atherosclerosis and Risk of Thrombosis in SLE” (HEARTS) study funded by the National Institutes of

2

Health.15, 16 Data for this cross-sectional study were collected between March 2002 and September

3

2005.17 All participants completed informed consent procedures. The study was approved by the

4

University of Pittsburgh Institutional Review Board.

5

Those with diabetes mellitus as defined by a history of diabetes, fasting glucose ≥7 mmol/L (≥126

6

mg/dL), or hypoglycemic therapy were excluded from the current analysis because of their inherent CVD

7

risk. After the exclusion of participants with diabetes mellitus and confirming a 1:1 match, the current

8

study population included 143 SLE patients and 143 Controls.

M AN U

SC

RI PT

1

9 Data collection/measures

11

Information on patient demographics and potential risk factors was collected at the time of Electron

12

Beam Computed Tomography (EBT) scan as previously described.9, 15 Briefly, the study visit included

13

anthropomorphic measurements (height, weight, and waist and hip circumferences), two consecutive

14

blood pressure readings, and a blood draw after a required fast, which was uniform between groups.

15

Blood samples were used to measure total cholesterol, triglycerides, and high-density lipoprotein

16

cholesterol (HDL-C). Low-density lipoprotein cholesterol (LDL-C) was calculated from measured total

17

cholesterol, HDL-C and triglycerides (Friedewald equation) .18 The homeostatic model assessment of

18

insulin resistance (HOMA-IR) was calculated by [insulin (mU/liter) × glucose (mmoles/liter) ] ÷22.5.

19

Hypertension was defined by physician diagnosis, measured mean blood pressure ≥140/90 mmHg, or

20

antihypertensive medication use.

AC C

EP

TE D

10

21 22 23 5

ACCEPTED MANUSCRIPT

EBT scans

2

The EBT scans were performed using an Imatron C-150 scanner (Imatron, San Francisco, CA) using

3

standard imaging procedures for all participants. The same scanner and methodology were used for SLE

4

participants and healthy controls.15 Aortic scans (6 mm transaxial images, 300 ms exposure time) were

5

obtained from the aortic arch to the iliac bifurcation. All scan data were saved to optical disc and

6

calcification scoring (Agatston19) was performed using a DICOM work station and software by AcuImage,

7

Inc (San Francisco, CA).

SC

RI PT

1

9

M AN U

8 Aortic PVAT density

The aPVAT volume was previously quantified by calculating the area of adipose surrounding the

11

descending thoracic aorta in each CT slice within the HU attenuation range for adipose tissue (-190 to -

12

30 HU) from the carina (proximal starting point) through T12 (distal end point) using the commercially

13

available software Slice-O-Matic (Tomovision, Montreal, Canada) and multiplying by the slice thickness

14

(Fig. 2A) .9 The adipose density (aPVAT density) was calculated by taking the average HU for each slice

15

(within the range: -190 to -30 HU) over the entire volume of the descending thoracic aorta (Fig. 2B).

16

EP

TE D

10

Statistical analysis

18

Univariate comparison of demographic variables, cardio-metabolic factors, AC, and aPVAT density

19

between SLE and HC were conducted using t-tests or Wilcoxon rank-sum for continuous variables and

20

Chi-squared for categorical variables. Fisher’s exact test was utilized for categorical variables if sample

21

size per cell was <5. The distributions of aPVAT volume were log transformed to achieve normality. AC

22

score was analyzed as a continuous (log (score+1)) variable. Spearman correlations were used to assess

23

associations between continuous cardio-metabolic factors and adipose density and volumes along with

AC C

17

6

ACCEPTED MANUSCRIPT

AC score. To assess whether aPVAT density differs by SLE status independent of aPVAT volume,

2

multivariable linear regression modeling was used with aPVAT density as the main dependent variable

3

and SLE status as the main independent variable. Linear regression models were also used to test the

4

association between AC score and aPVAT density. All models were further adjusted for age and either

5

CVD risk factors (CVD risk: including the presence of hypertension, postemenopausal status, and

6

cholesterol ratio) or circulating inflammatory markers (cInflamm: including CRP, homocysteine,

7

eSelectin, and sICAM) . Traditional CVD risk factors and circulating inflammatory markers were kept

8

separate during the analysis to determine their respective influence over SLE status and adipose tissue

9

density. Adjusted R2 values along with p-values were used to define which factors produced greater

M AN U

SC

RI PT

1

variation in the associations of SLE with adipose density or adipose volume with AC score. The models

11

were also adjusted for the 1:1 matching between SLE and HC participants; however, the matched term

12

was not significant, did not have great influence on the other covariates within the models and was

13

therefore removed. Surrogate measures of adiposity including BMI and waist-to-hip ratio were not

14

included in the multivariable regression analyses. We did run both surrogate measures of adiposity in

15

separate linear regression analyses and found that they did not attenuate the association between SLE-

16

status and PVAT density like PVAT volume. Therefore, in order to simplify the analysis, create

17

parsimonious models, and directly address the relationship between aorta PVAT density, volume and

18

aortic calcification in female SLE patients, we declined to include these measures in the models. A

19

variance of inflation (VIF) was calculated for all multivariable regressions to assess multicollinearity. All

20

analyses were performed using Stata (Stata/IC 12.1, StataCorp LP, College Station, TX). The level of

21

statistical significance was set at a 2-sided p-value of <0.05.

AC C

EP

TE D

10

22 23

7

ACCEPTED MANUSCRIPT

Results

2

Demographic, clinical and laboratory characteristics, aortic calcification (AC) and aPVAT density

3

comparisons between SLE and HC

4

Despite no significant difference in BMI (p=0.17) between SLE and HC, significantly greater waist-to hip

5

ratio (p=0.002) was detected in female SLE patients, suggesting a disparity in body mass distribution

6

when compared to their age- and race-matched HC counterparts. Interestingly, significant differences in

7

total cholesterol (p=0.008) and LDL (p=0.006) were reported, with HC having higher median values

8

compared to SLE. Not surprisingly, the circulating inflammatory markers, including CRP (p=0.044) ,

9

sICAM (p=0.0002) , eSelectin (p=0.0018) and homocysteine (p=0.025) were significantly higher in the

M AN U

SC

RI PT

1

female SLE patients (Table 1).

11

Despite no clinical manifestations of CVD within this cohort, AC was detected in more female SLE

12

patients (≈80%) than HC women (≈61%) (p<0.0001). AC score was significantly greater in female SLE

13

patients when compared to HC (p=0.0006) (Table 1).

14

There was a statistically significant difference in aPVAT density (p=0.03) with a higher mean density in

15

female SLE patients when compared to HC women (Table 1). As shown previously, aPVAT volume

16

(p=0.014) was significantly different between SLE and HC, with greater volume in female SLE patients.9

EP

AC C

17

TE D

10

18

SLE status is associated with aPVAT density

19

To determine the association of SLE status with aPVAT density, we performed linear regression analyses

20

with SLE status as an independent, categorical variable (HC or SLE) and aPVAT density as the dependent

21

variable. In univariate analysis, SLE status was associated with aPVAT density (p=0.032) (Table 2). The

22

positive association of aPVAT density with SLE was maintained after adjusting for age (p=0.013) or age

23

and CVD risk factors (p=0.046). However, when including circulating inflammatory markers (p=0.15) or 8

ACCEPTED MANUSCRIPT

1

aPVAT volume (p=0.44) the association was attenuated. The adjusted R2-value, suggests that the

2

aPVAT volume has more influence over the attenuation when compared to the circulating inflammatory

3

markers (R2: 0.47 versus 0.18, respectively) (Table 2).

RI PT

4

Increased aPVAT density is correlated with increasing aPVAT volume and AC score

6

We also found that aPVAT density was positively correlated with aPVAT volume (ρ, p-value: 0.75,

7

<0.0001). The high correlation was observed in both groups (SLE: (0.75,<0.0001) ; HC: (0.74,<0.0001)).

8

More dense aPVAT was significantly correlated with increased AC score for both SLE and HC women,

9

with stronger correlations in female SLE patients (ρ, p-value: 0.31 vs 0.23, p<0.01 for both) (not shown).

M AN U

SC

5

10

Linear regression analysis: continuous AC (log) scores are associated with increased aPVAT density

12

independent of CVD risk factors and circulating inflammatory markers in female SLE patients

13

The association of aPVAT density with continuous AC score (log) was stronger for SLE (β (SE) ,p-value:

14

0.45 (0.11) , <0.0001) when compared to HC (0.34 (0.12), 0.006) women. The adjusted R2-value

15

indicates that the SLE aPVAT density had more influence over AC score (adjusted R2=0.10) when

16

compared to the HC aPVAT density (adjusted R2=0.05) .

17

There were no interactions found between disease status (HC vs. SLE) and aPVAT density (p=0.53)

18

indicating that SLE status did not modify the association between aPVAT density and aortic calcification.

19

However, in order to determine which group was driving significant associations between aPVAT density

20

and aortic calcification, we analyzed each model separately within the HC group and within the SLE

21

group. For the female SLE patients, the adipose aPVAT density maintained a significant association with

22

AC score when adjusting for age (p=0.003) and CVD risk factors (p=0.02) or circulating inflammatory

23

markers (p=0.01) (Table 3). The inclusion of aPVAT volume, however, attenuated the association

AC C

EP

TE D

11

9

ACCEPTED MANUSCRIPT

between SLE aPVAT density and AC score (p=0.30). There were no interactions detected between

2

aPVAT density and aPVAT volume (p=0.84).

3

Contrastingly, within the HC women, the significant association between aPVAT density and AC score

4

was attenuated when adjusting for age alone and subsequently for CVD risk factors and circulating

5

inflammatory markers (p>0.1 for all) (Table 3).

RI PT

1

SC

6 Discussion

8

This is the first study to evaluate the density of adipose surrounding the aorta as related to AC in

9

clinically CVD-free SLE women. Female SLE patients had a higher aPVAT density when compared to HC

M AN U

7

women and both SLE and HC women showed a direct correlation between higher aPVAT density, greater

11

aPVAT volume and higher AC scores. Our findings are not consistent with our hypothesis or some

12

previous reports that suggested that lower density abdominal visceral and subcutaneous adipose was

13

indicative of greater CVD risk. We found that higher density aortic perivascular adipose is more strongly

14

associated with the presence of aortic calcification, but not after accounting for aPVAT volume. It raises

15

the question as to whether aPVAT depot is uniquely different than visceral and subcutaneous adipose

16

depots.

17

Higher density adipose tissue has been associated with more metabolically active adipose tissue also

18

known as brown adipose tissue, which is associated with leanness, female gender 20 and has been

19

viewed as more favorable when compared to large, single droplet lipid laden adipocytes known as white

20

adipose tissue.21-27 Contrary to this belief, brown adipose tissue activation through cold exposure, in a

21

widely used murine model of atherosclerosis (ApoE-/- and Ldlr-/-) , actually accelerated plaque growth

22

and compromised plaque stability28. However, others have presented evidence that continues to

23

support the beneficial aspects of brown adipose tissue, including Berbee et al., arguing that a functional

AC C

EP

TE D

10

10

ACCEPTED MANUSCRIPT

hepatic clearance pathway is not maintained in the ApoE-/- and Ldlr-/- strain thus contributing to the

2

opposing brown adipose tissue findings of Dong, et al.29 However, in the human, there may be instances

3

where chronic adipose inflammation, which potentially activates brown adipose tissue, could allow

4

plasma levels of lipoprotein remnants to exceed hepatic clearance capacity and thus contribute to CVD

5

progression.

6

Another possibility for the higher density perivascular adipose tissue associated with calcification would

7

be that we are actually quantifying fibrotic adipose tissue. In addition to the association with

8

calcification, the more positive aPVAT density was also highly correlated with greater aPVAT volume for

9

both HC and female SLE patients. Evaluating the adipose volume and quality within the Framingham

M AN U

SC

RI PT

1

Heart Study (FHS), Alvey et al. also found more dense visceral and subcutaneous adipose independently

11

associated with CAC and AC, which was also contrary to their initial hypothesis.30 They theorized this

12

higher density adipose tissue resulted from excess collagen production in response to the chronic

13

inflammatory condition caused by obesity and adipose hypertrophy. Increased volumes of adipose

14

tissue are inextricably linked to hypoxic conditions, which can signal for excess extracellular matrix

15

production including collagen thus creating fibrotic adipose tissue.31-33 Although speculative, adjusting

16

for aPVAT volume attenuated the detected association between aPVAT density and AC reported in our

17

study, suggesting aPVAT dysfunctionality.

18

Previous studies have shown less dense abdominal visceral and subcutaneous adipose tissue associated

19

with cardiometabolic risk and incident CVD.3-5 It is well known that aortic calcification precedes

20

coronary artery calcification.34 Just as separate vascular beds experience varied CVD progression,

21

perhaps the different adipose depots are more susceptible to changes in quality as measured by

22

radiodensity. Additionally, localized PVAT inflammation may alter adipose quality and the PVAT of

23

different vascular beds may be more susceptible to premature hypertrophy or fibrosis. Future studies

24

evaluating adipose quality at various anatomic locations with several subclinical CVD measures such as

AC C

EP

TE D

10

11

ACCEPTED MANUSCRIPT

carotid intima media thickness and pulse wave velocity will aid in isolating the effects of adipose

2

location and density with CVD pathology.

3

From a clinical perspective, the small, but significant difference in PVAT density that we observed is not

4

unprecedented. Franssens et al. recently reported a 7% increase in the risk of being classified in a higher

5

CAC score per each 5 HU decrease in epicardial adipose tissue (EAT) among women at high risk for CVD,

6

while Lu et al reported small differences in both EAT (-88.1 vs. -86.9 HU, p=0.008) and paracardial

7

adipose tissue (-106 vs. -103 HU, p<0.0001) quality related to high risk coronary plaques. 6, 7 These

8

recent findings support a potential role for relatively low density threshold differences associated with

9

high risk CVD. Our clinically CVD-free population may further explain our smaller reported differences

10

along with the important results from our multivariate model analysis reveal that aPVAT density may

11

carry more weight in the SLE women, which supports further studies to clarify the pathophysiological

12

implications of these differences, mainly among SLE women.

13

Our findings should be interpreted within the context of some limitations, which include the cross-

14

sectional design preventing us from understanding the temporal association between a potentially

15

changing adipose quality and volume with aortic calcification progression or regression. More recently,

16

Lee JJ, et al, followed participants within the Framingham Heart Study Third Generation cohort over 6

17

years and found increasing volumes of visceral and subcutaneous adipose tissue associated with

18

incident hypertension, hypertriglyceridemia, and metabolic syndrome with similar trends involving a

19

decrease in adipose density.35 Additionally, we were not able to analyze the density or volume of

20

subcutaneous or visceral adipose depots within this HEARTS cohort. The original CT scanning protocol

21

was only to capture vascular calcification and lacked purposeful positioning of the participants to obtain

22

proper images for analysis at the level of the umbilicus. Although the HC participants in this cohort have

23

been found to have a similar metabolic and lipid profile as our SLE group, which may contribute to the

24

high CAC prevalence 15, 16 and the smaller observed differences in adipose quality, we were able to

AC C

EP

TE D

M AN U

SC

RI PT

1

12

ACCEPTED MANUSCRIPT

detect significant associations with aortic calcification and aPVAT density in adjusted models within

2

female SLE patients while no significant associations in adjusted models within the HC women thus

3

lending to the power of our study.

4

Conclusions

5

Higher aPVAT density is associated with increased aPVAT volume and AC in female SLE patients. Given

6

that the association of aPVAT density with AC was attenuated after adjusting for aPVAT volume, the

7

denser aPVAT, which correlates with greater aPVAT volume in our study, may be an indicator of fibrosis

8

instead of a better adipose tissue quality. Aorta PVAT may be susceptible to fibrosis prior to other

9

visceral or subcutaneous adipose depots due to localized inflammation. Further studies are needed to

M AN U

SC

RI PT

1

10

better define the relevance of adipose tissue density in the premature vascular disease observed in

11

patients with SLE.

12 Conflict of interest

14

The authors declared they do not have anything to disclose regarding conflict of interest with respect to

15

this manuscript.

EP

16

TE D

13

Financial support

18

This study was supported by HEARTS RO1.

19

AC C

17

20 21

13

ACCEPTED MANUSCRIPT

Author contributions

2

KJS: developed CT scan thresholding protocol, collected data, primary author, statistical analysis,

3

created Fig. 1, manuscript review and submission.SRE: Statistical analysis, manuscript editing/content

4

contribution. JMA: manuscript editing/content contribution. SM: RO1 funded the collection of data for

5

this HEARTS cohort, manuscript editing/content.

RI PT

1

TE D

M AN U

[1] Manzi, S, Meilahn, EN, Rairie, JE, et al., Age-specific incidence rates of myocardial infarction and angina in women with systemic lupus erythematosus: comparison with the Framingham Study, American journal of epidemiology, 1997;145:408-415. [2] Ross, R, Atherosclerosis is an inflammatory disease, American heart journal, 1999;138:S419-420. [3] Rosenquist, KJ, Pedley, A, Massaro, JM, et al., Visceral and subcutaneous fat quality and cardiometabolic risk, JACC. Cardiovascular imaging, 2013;6:762-771. [4] Lee, JJ, Pedley, A, Hoffmann, U, et al., Cross-Sectional Associations of Computed Tomography (CT) -Derived Adipose Tissue Density and Adipokines: The Framingham Heart Study, Journal of the American Heart Association, 2016;5. [5] Rosenquist, KJ, Massaro, JM, Pedley, A, et al., Fat quality and incident cardiovascular disease, allcause mortality, and cancer mortality, The Journal of clinical endocrinology and metabolism, 2015;100:227-234. [6] Lu, MT, Park, J, Ghemigian, K, et al., Epicardial and paracardial adipose tissue volume and attenuation - Association with high-risk coronary plaque on computed tomographic angiography in the ROMICAT II trial, Atherosclerosis, 2016;251:47-54. [7] Franssens, BT, Nathoe, HM, Visseren, FL, et al., Relation of Epicardial Adipose Tissue Radiodensity to Coronary Artery Calcium on Cardiac Computed Tomography in Patients at High Risk for Cardiovascular Disease, The American journal of cardiology, 2017. [8] El Khoudary, SR, Shields, KJ, Janssen, I, et al., Postmenopausal Women With Greater Paracardial Fat Have More Coronary Artery Calcification Than Premenopausal Women: The Study of Women's Health Across the Nation (SWAN) Cardiovascular Fat Ancillary Study, J Am Heart Assoc, 2017;6. [9] Shields, KJ, Barinas-Mitchell, E, Gingo, MR, et al., Perivascular adipose tissue of the descending thoracic aorta is associated with systemic lupus erythematosus and vascular calcification in women, Atherosclerosis, 2013;231:129-135. [10] Rosito, GA, Massaro, JM, Hoffmann, U, et al., Pericardial fat, visceral abdominal fat, cardiovascular disease risk factors, and vascular calcification in a community-based sample: the Framingham Heart Study, Circulation, 2008;117:605-613. [11] Ding, J, Hsu, FC, Harris, TB, et al., The association of pericardial fat with incident coronary heart disease: the Multi-Ethnic Study of Atherosclerosis (MESA) , The American journal of clinical nutrition, 2009;90:499-504.

EP

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

References

AC C

7

SC

6

14

ACCEPTED MANUSCRIPT

EP

TE D

M AN U

SC

RI PT

[12] Cheng, VY, Dey, D, Tamarappoo, B, et al., Pericardial fat burden on ECG-gated noncontrast CT in asymptomatic patients who subsequently experience adverse cardiovascular events, JACC. Cardiovascular imaging, 2010;3:352-360. [13] Lipson, A, Alexopoulos, N, Hartlage, GR, et al., Epicardial adipose tissue is increased in patients with systemic lupus erythematosus, Atherosclerosis, 2012;223:389-393. [14] Liu, J, Fox, CS, Hickson, D, et al., Pericardial adipose tissue, atherosclerosis, and cardiovascular disease risk factors: the Jackson heart study, Diabetes care, 2010;33:1635-1639. [15] Kao, AH, Wasko, MC, Krishnaswami, S, et al., C-reactive protein and coronary artery calcium in asymptomatic women with systemic lupus erythematosus or rheumatoid arthritis, The American journal of cardiology, 2008;102:755-760. [16] Kao, AH, Lertratanakul, A, Elliott, JR, et al., Relation of carotid intima-media thickness and plaque with incident cardiovascular events in women with systemic lupus erythematosus, The American journal of cardiology, 2013;112:1025-1032. [17] Greco, CM, Li, T, Sattar, A, et al., Association between depression and vascular disease in systemic lupus erythematosus, J Rheumatol, 2012;39:262-268. [18] Friedewald, WT, Levy, RI and Fredrickson, DS, Estimation of the concentration of low-density lipoprotein cholesterol in plasma, without use of the preparative ultracentrifuge, Clinical chemistry, 1972;18:499-502. [19] Agatston, AS, Janowitz, WR, Hildner, FJ, et al., Quantification of coronary artery calcium using ultrafast computed tomography, Journal of the American College of Cardiology, 1990;15:827-832. [20] Cypess, AM, Lehman, S, Williams, G, et al., Identification and importance of brown adipose tissue in adult humans, The New England journal of medicine, 2009;360:1509-1517. [21] Baba, S, Jacene, HA, Engles, JM, et al., CT Hounsfield units of brown adipose tissue increase with activation: preclinical and clinical studies, Journal of nuclear medicine : official publication, Society of Nuclear Medicine, 2010;51:246-250. [22] Cinti, S, The adipose organ at a glance, Disease models & mechanisms, 2012;5:588-594. [23] Murano, I, Barbatelli, G, Giordano, A, et al., Noradrenergic parenchymal nerve fiber branching after cold acclimatisation correlates with brown adipocyte density in mouse adipose organ, Journal of anatomy, 2009;214:171-178. [24] Lowell, BB and Spiegelman, BM, Towards a molecular understanding of adaptive thermogenesis, Nature, 2000;404:652-660. [25] Nedergaard, J, Bengtsson, T and Cannon, B, Unexpected evidence for active brown adipose tissue in adult humans, American journal of physiology. Endocrinology and metabolism, 2007;293:E444452. [26] Shabalina, IG, Kramarova, TV, Nedergaard, J, et al., Carboxyatractyloside effects on brown-fat mitochondria imply that the adenine nucleotide translocator isoforms ANT1 and ANT2 may be responsible for basal and fatty-acid-induced uncoupling respectively, The Biochemical journal, 2006;399:405-414. [27] Ahmadi, N, Hajsadeghi, F, Conneely, M, et al., Accurate detection of metabolically active "brown" and "white" adipose tissues with computed tomography, Academic radiology, 2013;20:14431447. [28] Dong, M, Yang, X, Lim, S, et al., Cold exposure promotes atherosclerotic plaque growth and instability via UCP1-dependent lipolysis, Cell metabolism, 2013;18:118-129. [29] Berbee, JF, Boon, MR, Khedoe, PP, et al., Brown fat activation reduces hypercholesterolaemia and protects from atherosclerosis development, Nature communications, 2015;6:6356. [30] Alvey, NJ, Pedley, A, Rosenquist, KJ, et al., Association of fat density with subclinical atherosclerosis, Journal of the American Heart Association, 2014;3.

AC C

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

15

ACCEPTED MANUSCRIPT

RI PT

[31] Pasarica, M, Sereda, OR, Redman, LM, et al., Reduced adipose tissue oxygenation in human obesity: evidence for rarefaction, macrophage chemotaxis, and inflammation without an angiogenic response, Diabetes, 2009;58:718-725. [32] Divoux, A, Tordjman, J, Lacasa, D, et al., Fibrosis in human adipose tissue: composition, distribution, and link with lipid metabolism and fat mass loss, Diabetes, 2010;59:2817-2825. [33] Khan, T, Muise, ES, Iyengar, P, et al., Metabolic dysregulation and adipose tissue fibrosis: role of collagen VI, Molecular and cellular biology, 2009;29:1575-1591. [34] Kuller, LH, Matthews, KA, Sutton-Tyrrell, K, et al., Coronary and aortic calcification among women 8 years after menopause and their premenopausal risk factors : the healthy women study, Arteriosclerosis, thrombosis, and vascular biology, 1999;19:2189-2198. [35] Lee, JJ, Pedley, A, Hoffmann, U, et al., Association of Changes in Abdominal Fat Quantity and Quality With Incident Cardiovascular Disease Risk Factors, Journal of the American College of Cardiology, 2016;68:1509-1521.

SC

1 2 3 4 5 6 7 8 9 10 11 12 13 14

AC C

EP

TE D

M AN U

15

16

ACCEPTED MANUSCRIPT

Fig. Legends

2

Fig. 1: Schematic representations of CT radiodensity ranges of biological tissue and within CT slice aorta

3

PVAT density differences.

4

(A)

5

relative position of adipose density within the HEARTS cohort, (iii) proposed density differences within

6

aorta PVAT. (B) Aorta lumen with surrounding PVAT (green and gray areas): (i) the area per slice is

7

calculated by including any pixels between -190 to -30 HU (green + gray), (ii) the density per slice is

8

calculated by taking the average HU.

RI PT

1

M AN U

SC

Hounsfield Unit scale bar highlighting (i) adipose tissue specific HU range (-190 to -30 HU), (ii)

9

Fig. 2: Clinical CT image and resulting 3D aorta PVAT volume.

11

(A) Clinical CT image thresholded for aorta PVAT (red) (-190 to -30 HU), (B) 3D reconstruction of the

12

aorta PVAT volume resulting from the thresholded aorta PVAT in (A).

TE D

10

13 Table Legends

15

Table 1: Demographic, clinical and laboratory characteristics.

16

Median (25th-75th%) or mean (SD). Depending on normality, t test or Wilcoxon rank-sum were run

17

along with Chi-squared (n≥5) or Fisher’s exact test (n<5) for frequency.

18

*Three participants were not able to be matched on race, but were matched for gender, age (±5 years)

19

and geographic location.

AC C

EP

14

20 21

17

ACCEPTED MANUSCRIPT

Table 2: Association of aPVAT density with SLE adjusted for age and CVD risk factors.

2

CVD risk factors (CVD risk): hypertensive status, cholesterol ratio, postmenopausal status; circulating

3

inflammatory markers (cInflamm): CRP, homocysteine, eSelectin, sICAM, or aPVAT volume. VIF was

4

calculated for each model and all were <1.5.

RI PT

1

5

Table 3: Linear regression models aPVAT density regressed on AC score (log).

7

The CVD risk factors include: age, hypertension, cholesterol ratio, postmenopausal status; circulating

8

inflammatory markers (cInflamm) include: CRP, homocysteine, eSelectin, sICAM. VIF was calculated for

9

each model and all were <1.6.

AC C

EP

TE D

M AN U

SC

6

18

ACCEPTED MANUSCRIPT

SLE

Healthy control p-value

n=143

n=143

125 (87%)

128 (90%)

Age

49.6 (45-55)

51.2 (45-56)

0.28

Postmenopausal status, n (%)

84 (59%)

66 (46%)

0.03

Current smoker, n (%)

16 (11%)

18 (13%)

0.72

Systolic BP, log

119 (106-133)

120 (110-135)

0.11

Hypertensive

73 (51%)

44 (31%)

0.001

Waist-to-Hip ratio

0.833 (0.78-0.88)

0.802 (0.77-0.84)

0.002

BMI

27.0 (23-31)

27.2 (24-32)

0.17

Total cholesterol

185 (162-214)

203 (173-219)

0.008

3.41 (2.9-4.3)

3.52 (2.9-4.3)

0.87

Triglyceride, log

111 (75-153)

102 (73-140)

0.19

HDL-C

54.2 (43-62)

55.9 (46-64)

0.14

LDL-C

107 (88-126)

118 (100-140)

0.006

HOMA-IR (mmol/L x µU/mL)

6.71 (4.9-10)

6.79 ( 5.4-9.4)

0.49

2.35 (0.93-5.5)

1.61 (0.61-3.6)

0.044

SC

M AN U

TE D

Cholesterol ratio

0.60

RI PT

Caucasian*, n (%)

AC C

EP

(Total cholesterol/HDL)

Circulating inflammatory factors CRP (mg/L), log

264 (229-319)

246 (209-278)

0.0002

eSelectin

46.8 (30-62)

36.5 (26-53)

0.0018

Homocysteine (µmol/L), log

9.70 (8.0-12)

9.0 (7.6-10)

0.025

AC score (log)

3.45 (1.5-5.4)

1.79 (0-4.3)

0.0006

Any AC (AC>0)

115 (80%)

<0.0001

Calcification

87 (61%)

M AN U

Adipose quality (HU) and volumes (cm3)

RI PT

sICAM (ng/ml)

SC

ACCEPTED MANUSCRIPT

-83.6 (1.9)

-84.1 (1.8)

0.03

aPVAT volume, cm3 (log)

31.7 (26-42)

28.5 (23-37)

0.014

AC C

EP

TE D

aPVAT density

ACCEPTED MANUSCRIPT

aPVAT density (n=264)

Models β (SE)

adjusted R2

0.496 (0.23)

0.014

0.032

+ Age

0.554 (0.22)

0.087

0.013

+ Age, CVD risk

0.453 (0.23)

0.15

0.046

+ Age, cInflamm

0.318 (0.22)

0.18

0.15

AC C

EP

RI PT

M AN U

TE D

+ Age, aPVAT volume (log)

SC

SLE

0.134 (0.17)

p value

0.47

0.44

ACCEPTED MANUSCRIPT

AC score (log),

AC score (log),

AC score (log),

(n=281)

SLE (n=140)

HC (n=141)

Adj R2 p-value

β (SE)

β (SE)

Adj R2 p-value

aPVAT density 0.423 (0.082)

0.090

<0.0001

0.445 (0.11)

0.10

β (SE)

Adj R2

RI PT

Models

<0.0001

0.335 (0.12)

p-value

0.006

0.05

+ Age

0.267 (0.076)

0.26

0.001

0.314 (0.10)

0.29

0.003

0.120 (0.11)

0.28

0.26

+ Age, CVD risk

0.165 (0.079)

0.30

0.037

0.240 (0.10)

0.36

+ Age, cInflamm

0.165 (0.079)

0.33

0.038

0.276 (0.11)

0.34

0.01

-0.004 (0.12)

0.049 (0.11)

0.17

0.65

0.169 (0.16)

0.14

0.30

-0.059 (0.14)

0.027 (0.099)

0.30

0.78

0.166 (0.15)

0.29

0.26

-0.109 (0.13)

AC C

EP

TE D

+ Age, aPVAT volume

0.024 (0.12)

SC

M AN U

+ aPVAT volume

0.02

0.84

0.29 0.97 0.31 0.68 0.17 0.41 0.30

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

ACCEPTED MANUSCRIPT

ACCEPTED MANUSCRIPT

AC C

EP

TE D

M AN U

SC

RI PT

Highlights Systemic lupus erythematosus (SLE) women have more aortic perivascular adipose tissue, which • is denser when compared to their HC counterparts. • Aortic perivascular adipose tissue (PVAT) density is associated with aortic calcification and is independent of traditional CVD risk factors and circulating markers of inflammation in women with SLE. • Higher density aorta PVAT may be indicative of adipose tissue fibrosis.