The Interplay between Adipose Tissue Properties and Levels of NT-proBNP in People with HIV

Objective We aimed to assess the association between low N-terminal pro-brain natriuretic peptide (NT-proBNP) and body mass index (BMI), adipose tissue distribution, adiponectin, and HIV-specific risk factors among people with HIV (PWH). Methods We included 811 PWH with measurement of height, weight and waist circumference, blood samples analyzed for NT-proBNP, and visceral-(VAT) and subcutaneous (SAT) adipose tissue areas measured from CT-scans. Low concentrations of NT-proBNP were defined as concentrations below the limit of quantification (5.9 pmol/L). Associations were explored with multivariable logistic regression analyses adjusted for relevant confounders. Results We identified 471 (58%) individuals with low concentrations of NT-proBNP. Increasing BMI was associated with higher odds of low NT-proBNP (adjusted OR (aOR) 1.06 (95% CI: 1.01–1.11) per 1 kg/m2). Central obesity and large areas of VAT were associated with higher odds of low NT-proBNP (aOR 1.66 (1.16–2.36) and aOR 1.69 (1.09–2.62), respectively). Higher adiponectin was associated with lower odds of low NT-proBNP (aOR 0.86 (0.79–0.95) per 10% increase). No associations were found between low NT-proBNP and HIV-specific risk factors. Conclusions In PWH, low NT-proBNP is associated with an adverse adipose tissue profile with high BMI, central obesity, accumulation of VAT, and low adiponectin.


Introduction
Cardiometabolic diseases are a main cause of non-AIDSrelated morbidity and mortality in people with HIV (PWH) [1], and the risk of cardiovascular disease (CVD), including heart failure, is signifcantly higher among PWH than in the background population despite efective treatment with combination antiretroviral therapy (cART) [2].Natriuretic peptides are secreted from cardiomyocytes in response to increases in intracardiac pressure and elicit their primary role as cardioprotective hormones through constriction of the eferent renal arterioles, inhibition of cardiac fbrosis, and suppression of the renin-angiotensin-aldosterone system (RAAS) [3].Elevated levels of B-type natriuretic peptides (BNP) and the biologically inactive N-terminal prohormone of BNP (NT-proBNP) are associated with heart failure and mortality in both PWH and the general population [4][5][6][7], and current guidelines from both the European Society of Cardiology and the American Heart Association recommend measurement of BNP/NT-proBNP for both diagnosis and prognosis of heart failure [8,9].
Interestingly, while the risk of heart failure and CVD is strongly associated with increasing body mass index (BMI), levels of natriuretic peptides are lower in obese individuals [10][11][12].Tis paradox is known as "the natriuretic handicap" [13] and might in itself be associated with worse cardiovascular outcomes as insufcient levels of natriuretic peptides cause water retention, potentially promoting heart failure [3], and have been linked with poor cardiometabolic health [14].
Abdominal obesity and adipose tissue dysfunction are common among PWH [15,16], and chronic infammation is a hallmark of both untreated and treated HIV infection [17].Previous studies have found higher concentrations of NT-proBNP in PWH compared with uninfected controls [7,18,19].However, knowledge regarding the interplay between natriuretic peptides and HIV-associated changes to adipose tissue properties is sparse and limited to a single study reporting lower concentrations of BNP in PWH than controls when stratifying for BMI [20].
We aimed to investigate the association between concentrations of NT-proBNP and adipose tissue quantity, distribution, and function in a large cohort of well-treated PWH.Furthermore, we aimed to investigate potential HIVrelated risk factors for low NT-proBNP.We hypothesized that low concentrations of plasma NT-proBNP were associated with (1) higher BMI, (2) central obesity, large VAT, and large subcutaneous adipose tissue (SAT) areas, and (3) lower concentrations of adiponectin.

Study Population. Te Copenhagen Comorbidity in
HIV infection (COCOMO) study is a prospective, observational study aiming to assess the burden and pathogenesis of comorbidity in PWH.Te study includes 1,099 PWH enrolled between March 2015 and December 2016 [21].All participants provided oral and written informed consent, and the study was approved by the Ethics Committee of the Capital Region of Denmark (H-8-2014-004) and the Danish Data Protection Agency .Te present analyses included all PWH with both available biobank samples for measurement of NT-proBNP and measures of adipose tissue properties (Figure 1).

Clinical Assessments.
All participants underwent a physical examination with measurements of height, weight, waist circumference, and calculation of BMI performed by the trained medical staf [21].Information on demographics, smoking, and use of medication was collected from questionnaires.Data regarding HIV infection, including cART, CD4 + cell counts, CD4 + nadir, date of HIV-diagnosis, and hepatitis C virus (HCV)-antibodies, were obtained from review of all participants' medical records.

Biochemical Analyses.
NT-proBNP was measured in plasma collected at baseline and stored at −80 °C [22].Analyses were performed by sandwich electrochemiluminescenceimmunoassay (ECLIA) at the Department of Clinical Biochemistry, Rigshospitalet.Analyses for LDL, glucose, creatinine, and adiponectin were performed at the Department of Clinical Biochemistry, Herlev Hospital [21].Te estimated glomerular fltration rate (eGFR) was calculated based on serum creatinine using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation [23].

Defnitions.
Low concentrations of NT-proBNP were defned as plasma concentrations below the limit of quantifcation (5.9 pmol/L).BMI was divided into categories in accordance with the World Health Organization's classifcation (underweight: BMI <18.5 kg/m 2 , normal weight: 18.5-24.9kg/m 2 , overweight: 25.0-29.9kg/m 2 , and obese: BMI ≥30 kg/m 2 ) [24].Central obesity was defned as a waist circumference >80 cm for women and >94 cm for men in accordance with the International Diabetes Federation [25].Large VAT and SAT areas were ascribed PWH with areas in the upper quartile (≥141.1 cm 2 and ≥184.8 cm 2 , respectively).Hypertension was defned as antihypertensive treatment and/or having ≥140 mmHg systolic and/or ≥90 mmHg diastolic blood pressure [26].Dyslipidemia was defned as LDL ≥4.16 mmol/L (160 mg/ dL) and/or lipid-lowering treatment [27].Treatment with an antidiabetic drug for diabetes and/or nonfasting venous plasma glucose ≥11.1 mmol/L and/or HbA1c ≥6.5% was used to defne diabetes.Exposure to lipotoxic cART was defned as ever exposed to either stavudine, zidovudine, and/or didanosine.Measurements of VAT and SAT from single-slice CT-scans at the level of the 4 th lumbar vertebra were performed by trained personnel using commercially available CT software (Fat Measurement, Aquilion ONE; Canon, Japan).VAT area was defned as adipose tissue delineated by the intraabdominal muscular compartments, while SAT was defned as adipose tissue superfcial to the abdominal muscles as done previously (Supplemental Figure 1) [16].
Left atrial volume was determined using axial views.Te endocardial borders were manually traced on 15-20 tomographic slices, depending on the size of the atrial chamber.Te left atrial appendage was included in the measurement, while the pulmonary veins were meticulously excluded from the assessment of the atrial cavity [28].LA volume index (LAVi) was calculated by dividing LA volume with body surface area (BSA), which was calculated from the Du Bois formula as follows: weight(kg) 0.425 × height(cm) 0.725   × 0.007184. (1)

Statistical Analyses. Comparisons of baseline charac-
teristics between PWH with and without low NT-proBNP were performed with unpaired t-tests for normally distributed continuous variables or the Mann-Whitney U-test for non-normally distributed continuous variables, and χ 2 test for categorical variables.Normally distributed continuous variables were reported with means and standard deviations (SD), while non-normally distributed variables were reported with medians and interquartile ranges (IQR).Categorical data were reported with frequency counts and percentage of subjects within each category.NT-proBNP was regarded as a dichotomous variable (low concentrations: yes/no).Logistic regression models were used when exploring the association of diferent adipose tissue properties with low NT-proBNP and odds ratios (OR) were reported with 95% confdence intervals (CI).In addition to a base model adjusted for age and sex, a multiple regression model was applied with additional adjustment for possible confounders including smoking, diabetes, hypertension, eGFR, and HCV-antibodies.Tese variables were chosen a priori based upon previously reported associations [8][9][10]18].Models with VAT, SAT, or adiponectin as exposure were also adjusted for BMI.A sensitivity analysis was performed with additional adjustment for LAVi.Furthermore, multiple linear regression models were applied as a sensitivity analysis to utilize the full range of NT-proBNP concentrations.NT-proBNP was log-transformed, and linear models were adjusted for the same potential confounders as the logistic models.
Multiple logistic regression models adjusted for age, sex, smoking, diabetes, hypertension, eGFR, HCV-antibodies, and BMI were used to explore possible associations between NT-proBNP and HIV-specifc variables including time living with HIV, current CD4 + count, nadir CD4 + count (<200/µL: yes/no), and exposure to lipotoxic cART (yes/no) tested one at a time.
A two-sided p value <0.05 was considered statistically signifcant.All analyses were performed using R version 4.1.2(R Foundation for Statistical Computing, Vienna, Austria).

Baseline Characteristics.
We included 811 PWH from the COCOMO study with measurement of NT-proBNP, abdominal CT scans, and anthropometric measurements available (Figure 1).Of these, 672 (83%) also had a measurement of plasma adiponectin.We identifed 471 (58%) with low concentrations of NT-proBNP.PWH with low NT-proBNP were younger and more frequently male than PWH without low NT-proBNP (Table 1).Overall, PWH with low NT-proBNP were less likely to have dyslipidemia, hypertension, and diabetes but had a higher mean BMI.PWH with low NT-proBNP were diagnosed with HIV more recently, mode of transmission was more frequently men who have sex with men, and they were more likely to be virally suppressed than PWH without low NT-proBNP.

Body Mass Index.
Concentrations of NT-proBNP decreased with increasing BMI category, and overweight or obese PWH had lower NT-proBNP than those who had normal BMI (Figure 2).Increasing BMI was associated with 6% higher odds of low NT-proBNP per 1 kg/m 2 increase in BMI, both when adjusting for sex and age (OR 1.06 [95% CI: 1.02-1.10],p � 0.004) and after further adjustment for smoking, diabetes, hypertension, eGFR, and HCV antibodies (aOR 1.06 [1.01-1.11],p � 0.009) (Table 2).Furthermore, these fndings were consistent after additional adjustment for LAVi (Table 2) and when using linear regression models (Supplemental Table 1).
Journal of Obesity 3.4.Adipose Tissue Function.Higher concentrations of adiponectin were associated with lower odds of low NT-proBNP in both the base models (OR 0.87 [0.80-0.96]per 10% increase in adiponectin, p � 0.004) and after additional adjustments (aOR 0.86 [0.79-0.95]per 10% increase in adiponectin, p � 0.003) (Table 2) and similar after further adjustment for LAVi (Table 2).Consistently, higher adiponectin was associated with higher NT-proBNP in linear regression models (Supplemental Table 1).
No associations were found between low concentrations of NT-proBNP and time since HIV diagnosis, current CD4 + count, nadir CD4 + count, or previous exposure to lipotoxic cART (data not shown).

Discussion
We investigated the association between natriuretic peptides and adipose tissue properties in a large cohort of well-treated PWH and found that low concentrations of NT-proBNP were associated with (1) higher BMI, (2) central obesity and large VAT areas, and (3) lower concentrations of adiponectin.We found no association between low NT-proBNP and large SAT areas or any of the HIV-specifc variables tested.
Natriuretic peptides act upon multiple target organs and serve as regulators of cardiometabolic homeostasis by stimulating lipolysis, increasing GFR, and inhibiting both  Tus, genetically lower concentrations of natriuretic peptides have been linked with an adverse cardiometabolic profle with increased prevalence of metabolic syndrome, hypertension, and myocardial infarction [14].Furthermore, patients with heart failure but concentrations of natriuretic peptides lower-than-expected may be at an increased risk of delayed diagnosis [29].
In the present study, we found that increasing BMI was associated with lower concentrations of natriuretic peptides which is in line with previous studies from the general population [10][11][12].A previous study (N = 30) has also linked BMI and natriuretic peptides in PWH [20].When stratifed by both HIV-status and BMI, authors reported the lowest concentrations of BNP among obese individuals with HIV which was ascribed to the accumulation of dysfunctional fat seen among PWH.Notably, the study was conducted under strict physiological conditions during RAASactivation and only included 20 highly selected PWH.Although we did not include a control group in our study, we provide evidence for the frst time of an inverse relationship between obesity and NT-proBNP in a large and unselected cohort of PWH and verify that the association between BMI and low natriuretic peptides in PWH is consistent after adjustment for potential confounders.
Several hypotheses have been proposed to explain this "natriuretic handicap" and potential mechanisms include decreased synthesis and release of natriuretic peptides from cardiomyocytes in obese individuals either due to downregulation of the natriuretic peptide system [30], hyperinsulinemia [31], or accumulation of epi-and myocardial fat leading to myocardial damage [32].Hyperinsulinemia may especially contribute to natriuretic defciency in PWH, as previous studies from the COCOMO cohort found that PWH are at increased risk of diabetes and that HIV-specifc risk factors including time with HIV and prior immunodefciency are associated with higher insulin resistance [33,34].A previous study in PWH also found an inverse correlation between BNP and HOMA-IR, but BNP remained lower in PWH than controls after adjusting for HOMA-IR suggesting that lower BNP in PWH cannot be attributed to hyperinsulinemia alone [20].
Although BMI is a useful indicator of overall adiposity, it is a crude measure of obesity and does not account for body composition or adipose tissue distribution.Previous research indicates that adipose tissue in diferent compartments diferentially infuence the association with natriuretic peptides, and that VAT is more strongly associated with lower concentrations of natriuretic peptides than SAT [35][36][37].PWH are at increased risk of abdominal obesity [15] as well as redistribution of fat from the subcutaneous to the visceral compartment [38] and might therefore be at particular high risk of low NT-proBNP.Interestingly, we found specifcally central obesity and VAT, but not SAT, to be associated with low concentrations of NT-proBNP.
Te association between VAT and low concentrations of natriuretic peptides may be explained by increased clearance of BNP at the adipose tissue level due to an increased gene expression of NPR-C in VAT compared with SAT [39].However, since NT-proBNP is not eliminated by the NPR-C receptor [40], the association between VAT and low NT-proBNP cannot only be explained by this mechanism.Interestingly, one study found that the association between VAT and NT-proBNP was attenuated after additional adjustment for insulin resistance [35], suggesting that hyperinsulinemia may also facilitate natriuretic defciency in individuals with excess VAT.
Adiponectin is an adipokine that reduces infammation and improves insulin sensitivity, and low concentrations are associated with increased risk of CVD [41].Higher concentrations of adiponectin have previously been associated with higher natriuretic peptides in both healthy men [42] and patients with heart failure [43], and concentrations of adiponectin have been demonstrated to increase after intravenous administration of natriuretic peptides in patients with heart failure [44].VAT might be the common link between the two hormones, as VAT is associated with both lower concentrations of natriuretic peptides and produces TNF-α that reduces adiponectin expression [45].Tus, VAT accumulation in PWH may lead to lower concentrations of both NT-proBNP and adiponectin -especially among those previously exposed to lipotoxic cART, as lipotoxic cART may have long lasting detrimental efects upon levels of adiponectin [16].In the present study, increase in concentrations of adiponectin was associated with a decreased risk of low NT-proBNP.However, we did not observe an association between exposure to lipotoxic cART and low concentrations of NT-proBNP.Major strengths of the study include the large sample size, a well-characterized study population, and CTmeasured VAT and SAT areas.However, as this was a cross-sectional, observational study, we cannot make conclusions on causality or statements about the temporal relationship between obesity and concentrations of NT-proBNP.We also cannot assess if the efect of obesity on NT-proBNP levels is more pronounced in PWH compared with the general population as we did not include a control group for this study.Another limitation was that 58% of PWH had NT-proBNP values below the limit of quantifcation, which limited our ability to reliably distinguish low concentrations of NT-proBNP from very low concentrations.In addition, our population consisted mainly of Caucasian males, which disallowed sex-stratifed analyses, and more than 98% were on stable cART.Terefore, our fndings may not be generalizable to other HIV-populations.Moreover, the defnition of lipotoxic cART was narrow.However, we limited the defnition to treatments with welldocumented efects on adipose tissue physiology as done previously [16].Finally, even when adjusting analyses for age, sex, and several comorbidities, additional unmeasured confounders may be present.However, similar fndings in sensitivity analyses with additional adjustment for LAVi were reassuring.

Conclusion
Te present study found that in PWH, low NT-proBNP is associated with an adverse adipose tissue profle with high BMI, central obesity, accumulation of VAT, and low adiponectin.levels of natriuretic peptides help regulate cardiovascular homeostasis by protecting against volume overload, hypertension, and structural heart disease, and our fndings add another potential link between metabolic dysfunction and CVD in PWH.
Further studies comparing natriuretic peptides and adipose tissue properties in both PWH and controls are warranted to investigate the clinical impact of natriuretic peptide defciency among PWH.In addition, a lower cutpoint for BNP/NT-proBNP may be relevant when diagnosing acute heart failure in obese PWH who may be at particular high risk of both natriuretic peptide defciency and heart failure and may present with lower-than-expected levels of natriuretic peptides.

Data Availability
According to Danish legislation and approvals from the ethical committee and data protection agency, the data underlying this article cannot be shared publicly.Te data can be accessed at our institution upon reasonable request to the corresponding author.

Additional Points
Study Important Questions.What is already known about this subject?Despite being associated with increased risk for cardiovascular disease, obesity is associated with low concentrations of natriuretic peptides.Abdominal obesity and adipose tissue dysfunction are prevalent among people with HIV (PWH), but knowledge about the association between natriuretic peptides and obesity is missing in this setting.What are the new fndings in your manuscript?Low NT-proBNP is associated with an adverse adipose tissue profle with high BMI, central obesity, accumulation of VAT, and low adiponectin among PWH.How might your results change the direction of research or the focus of clinical practice?A lower cut-point for BNP/NT-proBNP may be relevant when diagnosing acute heart failure in obese PWH who may be at particular high risk of both natriuretic peptide defciency and heart failure and may present with lower-than-expected levels of natriuretic peptides.

Disclosure
Te funder had no part in the design of the study, the collection, analysis, or interpretation of data or publication.

Conflicts of Interest
A.D.K. has received a grant from the Danish Heart Foundation and the European Commission.T.B. reports grants from Novo Nordisk Foundation, Lundbeck Foundation, Simonsen Foundation, GSK, Pfzer, Gilead, Kai Hansen Foundation, and Erik and Susanna Olesen's Charitable Fund 6 Journal of Obesity and personal fees from GSK, Pfzer,

4
Journal of Obesity RAAS and cardiac fbrosis [3].While high concentrations of BNP/NT-proBNP are used in diagnosis and prognosis of heart failure, low concentrations are also associated with a clinically important increased risk of adverse outcomes.

Table 1 :
Baseline characteristics for PWH with and without low concentrations of NT-proBNP a .
NT-proBNP, N-terminal prohormone of B-type natriuretic peptide; IQR, interquartile range; BMI, body mass index; SD, standard deviation; HIV, human immunodefciency virus; VAT, visceral adipose tissue; SAT, subcutaneous adipose tissue; MSM, men who have sex with men; RNA, ribonucleic acid; cART, combination antiretroviral therapy; HCV, hepatitis C virus. a Dyslipidemia was defned as LDL ≥4.16 mmol/L (160 mg/dL) and/or lipid-lowering treatment.Hypertension was defned as antihypertensive treatment and/or having ≥140 mmHg systolic and/or ≥90 mmHg diastolic blood pressure.Treatment with an antidiabetic drug for diabetes and/or nonfasting venous plasma glucose ≥11.1 mmol/L and/or HbA1c ≥6.5% was used to defne diabetes.Exposure to lipotoxic cART was defned as ever exposed to either stavudine, zidovudine, and/or didanosine.If numbers and percentages do not add up, it is due to missing values.

Table 2 :
Associations of adipose tissue measurements with low concentrations of NT-proBNP., body mass index; VAT, visceral adipose tissue; SAT, subcutaneous adipose tissue; LAVi, left atrial volume index.Values shown are odds ratios (95% CI) for having low concentrations of NT-proBNP adjusted for age and sex (base model) and after further adjustment for diabetes, hypertension, eGFR, HCV-antibodies (adjusted model) and LAVi.Additionally, models exploring associations between VAT, SAT, and adiponectin and low NT-proBNP were adjusted for BMI.Large VAT and SAT areas were ascribed PWH with areas in the upper quartile (≥141.1 cm 2 and ≥184.8 cm 2 , respectively).684 PWH were included in the adjusted model exploring associations between low NT-proBNP and BMI, central obesity as well as large VAT and SAT areas (538 when further adjusting for LAVi), while 593 were included in the adjusted model exploring associations between low NT-proBNP and adiponectin (467 when further adjusting for LAVi). BMI