Anti-TNFα therapy transiently improves high density lipoprotein cholesterol levels and microvascular endothelial function in patients with rheumatoid arthritis: a Pilot Study
- First Online:
Rheumatoid arthritis (RA) is associated with increased morbidity and mortality from cardiovascular disease (CVD). This can be only partially attributed to traditional CVD risk factors such as dyslipidaemia and their downstream effects on endothelial function. The most common lipid abnormality in RA is reduced levels of high-density lipoprotein (HDL) cholesterol, probably due to active inflammation. In this longitudinal study we hypothesised that anti-tumor necrosis factor-α (anti-TNFα) therapy in patients with active RA improves HDL cholesterol, microvascular and macrovascular endothelial function.
Twenty-three RA patients starting on anti-TNFα treatment were assessed for HDL cholesterol level, and endothelial-dependent and -independent function of microvessels and macrovessels at baseline, 2-weeks and 3 months of treatment.
Disease activity (CRP, fibrinogen, DAS28) significantly decreased during the follow-up period. There was an increase in HDL cholesterol levels at 2 weeks (p < 0.05) which was paralleled by a significant increase in microvascular endothelial-dependent function (p < 0.05). However, both parameters returned towards baseline at 12 weeks.
Anti-TNFα therapy in RA patients appears to be accompanied by transient but significant improvements in HDL cholesterol levels, which coexists with an improvement in microvascular endothelial-dependent function.
KeywordsDyslipidaemia Anti-TNFα Endothelial function Rheumatoid arthritis Cardiovascular disease
Anti-tumor necrosis factor alpha
Body mass index
Co-efficient of variation
Disease activity score in 28 joints
Erythtrocyte sedimentation rate
Glyceryl-trinitrate mediated dilatation
Health Assessment Questionnaire
High density lipoprotein cholesterol
International physical activity questionnaire
Laser Doppler imaging
Low density lipoprotein cholesterol
Rheumatoid arthritis (RA) is a chronic inflammatory disease, characterised by erosive symmetrical polyarthritis. It is the most common inflammatory arthritis, affecting approximately 0.8% of the adult population. RA is also associated with increased morbidity and mortality from cardiovascular disease (CVD) , which appears to be of equal frequency and severity as in those with diabetes mellitus of similar duration . This could be partially attributed to an increased prevalence of traditional CVD risk factors such as hypertension, dyslipidaemia and obesity . In particular, the prevalence of dyslipidaemia in RA is high , it is evident in both early  and advanced RA , and may even associate with some RA susceptibility genes  and precede the development of RA .
The most common lipid abnormality in RA is reduced levels of high-density lipoprotein (HDL) cholesterol [5, 9]. HDL prevents oxidation of low density lipoprotein (LDL) cholesterol and is involved in cellular efflux of accumulated LDL particles in the vasculature, which in turn prevents foam cell accumulation and subsequent endothelial dysfunction (ED) . Inflammation can reduce the anti-atherogenic and anti-oxidant properties of HDL by altering the function of key enzymes important to the metabolism of HDL. For example, pro-inflammatory cytokines such as tumor necrosis factor alpha (TNFα) can reduce hepatic lipase concentration . Hepatic lipase alters the size of HDL particles to allow greater uptake of cholesterol from other cells. Changes to the enzymatic content of HDL (i.e. reduced paraoxonase) also results from inflammation, and transforms HDL into a pro-oxidant and pro-inflammatory molecule [12, 13]. Inhibition of TNFα through treatment of RA patients can increase HDL levels , as well as reducing disease activity and delaying or even reversing progressive joint damage . Importantly, Anti-TNFα therapy may also reduce cardiovascular mortality in RA , possibly as a result of a reduction in inflammation .
HDL also exerts specific effects in the vasculature including the stimulation of nitric oxide (NO) in endothelial cells, the reduction of TNFα-mediated superoxide release and repair of damaged blood vessels [18, 19, 20, 21]. A number of cross-sectional studies have shown that HDL particle size inversely associates with microvascular and macrovascular ED in a variety of populations including healthy individuals, diabetics [10, 22, 23] as well as RA patients [24, 25, 26]. Furthermore, a systematic review of the literature revealed that HDL levels are inversely associated with the risk for stroke and carotid artery atherosclerosis in various patient groups . Collectively, these findings suggest that HDL may be an important determinant of vascular health in any population, including RA.
The objective of the present longitudinal study was to evaluate the effect of anti-TNFα therapy on HDL cholesterol levels and to explore associations between changes in HDL cholesterol levels with contemporary changes in microvascular and macrovascular endothelial function in RA patients.
Twenty three consecutive patients with RA who were due to start anti-TNFα therapy on clinical indication (UK guidelines ) were recruited from the Rheumatology outpatient clinics of the Dudley Group NHS Foundation Trust, UK. All patients met the retrospective application of the 1987 revised RA criteria of the American College of Rheumatology . The exclusion criterion was: previously confirmed acute coronary syndrome or established CVD. The study had Research Ethics Committee and Research and Development approval from Birmingham, East, North and Solihull Research Ethics Committee and all participants gave their written informed consent according to the Declaration of Helsinki.
Anti-TNFα patients attended a thermoregulated (22 ± 0.9°C) vascular laboratory after a 12 hour overnight fast between 7:00 am and 11:00 am for 3 separate visits: prior to starting anti-TNFα (pre-treatment), 2 weeks and 3 months after initiation of treatment. Fifteen (65%) of the anti-TNFα patients were started on 40 mg of adalimumab taken fortnightly, six (26%) on 50 mg of etanercept taken weekly and two (9%) on infliximab with a dosage of 3 mg/kg according to standard dose administration protocol. There was no change in any of these medications or their doses during the follow up period. Assessments at 2 weeks and 3 months were performed before the patient received the next dose of their drug. Patients were asked to refrain from exercise 24 hours before the session and from smoking 12 hours before the session.
Metabolic and serological assessments
All patients underwent evaluation of their medical history and general demographics, examination of height, weight and body mass index (BMI). Blood tests were carried out in the biochemistry laboratories at Russells Hall Hospital, UK. Serum levels of total cholesterol (TC), high-density lipoprotein cholesterol (HDL), triglycerides (TG) and C-reactive protein (CRP) were measured using the Vitros 5.1chemistry system (http://www.orthoclinical.com). Levels of erythrocyte sedimentation rate (ESR) were determined using the Westergren method via a Starrsed Compact (Mechatronics BV, Netherlands). Fibrinogen was measured by photo-optical clot detection based on the Clauss quantitative fibrinogen method, using an automated coagulation analyser (ACL Futura Plus, Instrumentation Laboratory, UK). On each occasion disease activity was assessed with disease activity score in 28 joints (DAS28), along with functional ability using Anglicised version of the Stanford Health Assessment Questionnaire (HAQ), and physical activity levels using International Physical Activity Questionnaire (IPAQ) [30, 31, 32]. Concentration of the TNFα cytokine was measured from stored serum samples using FlowCytomix kits (BenderMed Systems GmbH, Vienna, Austria). Fluorescent beads were coated with antigen-specific antibodies for the cytokine and incubated with the sample. The assay follows the same principle as a sandwich immunoassay. Flow-cytometry (Bender MedSystems GMBH, Vienna, Austria) was used to differentiate bead populations by their size and fluorescent signature.
Microvascular endothelial function
Endothelial function of the microvasculature was assessed non-invasively by a single observer (AS), using Laser Doppler Imaging (LDI) (Moor LDI 2 SIM, Moor Instruments Ltd, Devon, UK) with iontophoresis of 1% acetylcholine (ACh, endothelium-dependent) and 1% sodium-nitroprusside (SNP, endothelium-independent) (Sigma Chemical Co, Montvale, New Jersey, USA) in 0.5 ml of saline. The technique was performed according to previously established guidelines  and has been described in detail previously . Briefly, after a baseline scan, ten scans were recorded during iontophoresis of the vasoactive agents using a 30 μA current, followed by two scans during recovery. This technique has an intra-observer co-efficient of variation (CV) for ACh and SNP of 6.5% and 5.9% respectively in our laboratory.
Macrovascular endothelial function
Assessment of macrovascular endothelial-dependent function was performed using flow-mediated dilatation (FMD) with high-resolution ultrasonography of the brachial artery (Acuson Antares ultrasound system, Siemens PLC, Camberley, UK) according to previously established guidelines  and has been described in detail previously . Following ten minutes of rest, endothelium-independent responses were examined by administration of 500 microgram sublingual glyceryl-trinitrate (GTN) tablet (Alpharma, Barnstaple, UK) while the brachial artery was imaged continuously for five minutes. The intra-observer CV was 10.7% for FMD and 11.8% for GTN assessments respectively. For all vascular tests, endothelial function was expressed as the percentage increase in perfusion or diameter from baseline, and all analysis was carried out offline by AS who was blinded to the identity and time point of the patient.
Statistical analysis was performed using SPSS18 (SPSS Inc, Chicago, Illinois). Variables were tested for normality by the Kolmogorov-Smirnov test. Means and standard deviations (SD) were calculated for normally distributed continuous variables and proportions for categorical variables. Log transformation was performed for positively skewed variables as appropriate. The effect of anti-TNFα treatment on blood lipid parameters and vascular indices (continuous variables) was tested using repeated measures analyses of variance (ANOVA). Where appropriate, LSD post hoc analyses were conducted. The change in lipid and inflammatory parameters as well as endothelial function at 2 and 12 weeks was calculated by subtracting the pre-treatment baseline values from the values obtained at 2 and 12 weeks. Pearson correlations were used to examine whether changes in endothelial function related to changes in lipids or disease-related inflammation. Endothelial function and lipid parameters did not differ between the three different types of anti-TNFα treatment at any time point, therefore all treatments were analysed together.
Baseline characteristics of the RA patients.
54 ± 15
Sex female N (%)
165 ± 8
30 ± 6
RF Positive N (%)
Disease duration (years)
11 ± 11
CVD Risk Factors
Family history of CVD
Medium dose GC (≥7.5 mg) (%)
Anti-hypertensive treatment (%)
Treatment with anti-TNFα
Lipid profile, disease activity, dyslipidaemia risk factors, and vascular measurements across the three visits for patients starting anti-TNFα therapy.
4.7 ± 1.0
4.9 ± 1.1
4.8 ± 0.9
HDL cholesterol (mmol/l)
1.4 ± 0.3
1.5 ± 0.3†
1.4 ± 0.3
1.4 ± 0.6
1.5 ± 0.7
1.6 ± 0.9
10 (4 to 14)
3 (2.9 to 6)†
5 (2.9 to 10)†
5.1 ± 1.0
4.2 ± .75 †
4.3 ± .91†
4.22 ± 0.94
2.71 ± 1.38†
2.80 ± 1.30†
2.1 ± 0.5
1.3 ± 0.9 †
1.3 ± 0.9 †
TNFα cytokine (pg/ml)*
Dyslipidaemia Risk Factors
30.2 ± 6.1
30.3 ± 6.1
30.3 ± 6.2
IPAQ (met min/wk)
2850 ± 5451
2914 ± 3551
2503 ± 3163
Microvascular endothelial-dependent and endothelial-independent function
319 ± 217
437 ± 247 †
348 ± 209
252 ± 126
288 ± 150
261 ± 152
Macrovascular endothelial-dependent and endothelial-independent function
9.4 ± 6.9
12.0 ± 9.6
12.0 ± 8.1
22.6 ± 7.5
23.2 ± 7.0
23.8 ± 7.2
Of the vascular measurements, only microvascular endothelial-dependent function (ACh) exhibited a significant change (increased dilatation) after 2 weeks of anti-TNFα therapy compared to baseline measurements: 437 ± 247% vs. 319 ± 217% respectively, p = 0.001 (Table 2). None of the other vascular assessments demonstrated significant changes either at 2 weeks or 3 months of follow-up (Table 2). Pearson correlations did not reveal any significant associations between change in microvascular endothelial function and change in HDL cholesterol at 2 weeks (r (21) = -.03, p = .89). There were no significant correlations between changes in HDL cholesterol with changes in other vascular or inflammatory parameters.
The present study revealed a transient improvement in HDL cholesterol levels following 2 weeks of treatment with anti-TNFα theraphy which returned to baseline levels after 3 months in RA patients with active disease who had newly started anti-TNFα treatment. A concomitant increase in microvascular endothelial-dependent function but not macrovascular endothelial-dependent function was also observed in these patients.
To our knowledge the current study is the first to report the effects of anti-TNFα in RA patients who underwent simultaneous measurements of their lipid profile as well as microvascular and macrovascular endothelial function. Previous studies that measured lipid parameters and endothelial function in response to anti-TNFα treatment have examined endothelial function in the macrocirculation only  or have focused on drugs targeting other inflammatory pathways [37, 38], and report conflicting findings. Irace and colleagues  reported a transient increase in macrovascular endothelial-dependent function immediately after infusion of infliximab at 0, 2 and 6 weeks, however, HDL levels decreased. It was suggested that during inflammation, HDL particles bind with TNFα and buffer the cytokine from the circulation. Inhibition of TNFα with infliximab could reduce TNFα-HDL- complexes and consequently reduce circulatory levels of HDL cholesterol . In contrast, an increase in HDL cholesterol levels was reported in RA patients following short-term (2 weeks and 6 weeks) treatment with infliximab which was correlated with reductions in disease-related inflammation . Similarly, administration of the anti-CD20 agent, rituximab, led to a significant improvement in HDL cholesterol levels and macrovascular endothelial-dependent function after 16 weeks  and 24 weeks of treatment , both of which are time points associated with maximum efficacy for rituximab in reducing inflammation . Inflammation can reduce HDL levels  and its core components, such as the anti-oxidant enzyme paraoxonase [12, 13]. It is therefore possible that in the present study, the increase in HDL cholesterol reflected the acute reduction in global inflammation at 2 weeks, with stabilisation of inflammation at 3 months resulting in no further changes in the lipid profile .
The transient improvement in microvascular endothelial-dependent function mirrored the increase in HDL cholesterol at 2 weeks with the majority of other risk factors (including physical activity levels) remaining unchanged. HDL cholesterol can exert acute effects in the vasculature, including a reduction in TNFα mediated superoxide release, increased production of endothelial progenitor cells, and stimulation of NO in endothelial cells [18, 19, 20, 21] all of which increase endothelial dependent vasodilatation . Given that both HDL cholesterol levels as well as its anti-oxidant ability can increase with administration of anti-TNFα in RA [14, 43], it is likely that the improvement in microvascular endothelial-dependent function after 2 weeks of treatment with anti-TNFα was mediated by the increased HDL cholesterol levels. Nevertheless, no association was apparent between baseline HDL cholesterol and endothelial function or changes in HDL cholesterol levels and endothelial function following treatment. The absence of any associations in the present study may be due to the small sample size. To our knowledge, no other studies have reported associations with endothelial function and HDL cholesterol in RA. Further work exploring such associations in a larger sample size is required.
The finding that microvascular function, but not macrovascular function improved following treatment might reflect the heterogeneity of these vascular beds from each other . We have previously shown that microvascular and macrovascular endothelial function are independent from each other in patients with RA . The microvasculature makes up a much larger proportion than macrovessels and may therefore respond earlier to changes in CVD risk factors or other injurious stimuli . Indeed, it has been hypothesised that the inflammatory response as a result of hypercholesterolemia may be initiated by endothelial activation in the microvasculature . Consequently, it is possible that even small changes in blood lipids could have a greater effect on microvascular endothelial-dependent function. There is also some evidence that in other clinical conditions like diabetes, microvascular dysfunction develops independently of macrovascular dysfunction , and may even contribute to the development of macrovascular disease .
The majority of previous studies examining the effect of anti-TNFα on the lipid profile have included patients treated with infliximab [36, 48], and there is little evidence on the effects of other anti-TNFα agents. Different anti-TNFα agents have unique molecular structures, modes of action and half lives and could therefore differentially impact on inflammation and the lipid profile . In the current study only two patients were receiving infliximab infusions, with the majority of patients receiving adalimumab (n = 15) or etanercept (n = 6). Consequently, the effect of each agent on the lipid profile and endothelial function could not be characterised due to the small sample size of the study. Further studies which include a large sample of patients on all types of anti-TNFα agents are required to delineate their individual effects on CVD risk factors and endothelial function.
There is evidence that Inflammation might affect the size of lipid particles, as TNFα can increase the HDL particle size reducing its effectiveness in reverse cholesterol transport [11, 49]. This has profound implications on the assessment of blood lipids as the size of the HDL particle may be of equal (or greater) importance to HDL cholesterol levels in the circulation. Furthermore, a number of studies have demonstrated that the protein content of HDL particles is altered during chronic RA disease-related inflammation [50, 51]. In particular, there is an increase in acute phase proteins and proteins involved in the complement cascade which results in HDL becoming pro-inflammatory in RA . In addition, RA patients exhibiting pro-inflammatory HDL have greater inflammatory proteins within the HDL molecule when compared to patients exhibiting anti-inflammatory HDL and healthy controls, and pro-inflammatory HDL complexes associate with higher DAS28 and presence of erosive disease . These findings highlight the need for future studies which measure the effects of lipid sub-fractions as well as the protein content of HDL molecules on CVD risk in RA.
Even though the sample size was not large, it had sufficient power to detect differences in HDL and vascular parameters. Post hoc power analyses using GPower3  with significance set at .05 revealed that, given the sample size and effect size, the obtained power was .94 to detect differences in HDL cholesterol, .99 to detect differences in microvascular endothelial function, and .98 to detect differences in macrovascular endothelial function. A limitation of the study was the absence of a control group of RA patients not receiving anti-TNFα. It was decided that it would be unethical to withhold treatment to patients with active disease requiring anti-TNFα, and including patients receiving other disease modifying anti-rheumatic drugs would not be suitable, as relative to the anti-TNFα group, these patients would have much lower levels of baseline inflammation.
In conclusion, the present study revealed that treatment with anti-TNFα resulted in a transient improvement in HDL cholesterol levels and a concomitant improvement in microvascular but not in macrovascular endothelial function. This was possibly mediated by a favourable effect of HDL cholesterol on microvascular endothelial function. Further prospective studies utilising large sample sizes are required to confirm these findings.
The authors would like to thank Dr George Balanos for his help and assistance with the flow-mediated dilatation assessment. Dr A. Sandoo was supported by a PhD studentship of the University of Birmingham, and the Department of Rheumatology, Dudley Group of Hospitals NHS Foundation Trust by an arc infrastructure support grant (No 17682).
- 4.Toms TE, Panoulas VF, Douglas KM, Griffiths H, Sattar N, Smith JP: Statin use in rheumatoid arthritis in relation to actual cardiovascular risk: evidence for substantial undertreatment of lipid-associated cardiovascular risk?. Ann Rheum Dis. 2010, 69: 683-688. 10.1136/ard.2009.115717.CrossRefPubMedGoogle Scholar
- 12.Van Lenten BJ, Hama SY, de Beer FC, Stafforini DM, McIntyre TM, Prescott SM: Anti-inflammatory HDL becomes pro-inflammatory during the acute phase response. Loss of protective effect of HDL against LDL oxidation in aortic wall cell cocultures. J Clin Invest. 1995, 96: 2758-2767. 10.1172/JCI118345.CrossRefPubMedPubMedCentralGoogle Scholar
- 18.Sorrentino SA, Besler C, Rohrer L, Meyer M, Heinrich K, Bahlmann FH: Endothelial-vasoprotective effects of high-density lipoprotein are impaired in patients with type 2 diabetes mellitus but are improved after extended-release niacin therapy. Circulation. 2010, 121: 110-122. 10.1161/CIRCULATIONAHA.108.836346.CrossRefPubMedGoogle Scholar
- 21.Nicholls SJ, Dusting GJ, Cutri B, Bao S, Drummond GR, Rye KA: Reconstituted high-density lipoproteins inhibit the acute pro-oxidant and proinflammatory vascular changes induced by a periarterial collar in normocholesterolemic rabbits. Circulation. 2005, 111: 1543-1550. 10.1161/01.CIR.0000159351.95399.50.CrossRefPubMedGoogle Scholar
- 22.Ford MA, McConnell JP, Lavi S, Rihal CS, Prasad A, Sandhu GS: Coronary artery endothelial dysfunction is positively correlated with low density lipoprotein and inversely correlated with high density lipoprotein subclass particles measured by nuclear magnetic resonance spectroscopy. Atherosclerosis. 2009, 207: 111-115. 10.1016/j.atherosclerosis.2009.04.039.CrossRefPubMedPubMedCentralGoogle Scholar
- 25.Wallberg-Jonsson SW, Backman C, Johnson O, Karp K, Lundstrom E, Sundqvist KG: Increased prevalence of atherosclerosis in patients with medium term rheumatoid arthritis. J Rheumatol. 2001, 28: 2597-2602.Google Scholar
- 28.National Institute for Health and Clinical Excellence: Adalimumab, etanercept and infliximab for the treatment of rheumatoid arthritis. 2009,http://www.nice.org.uk/nicemedia/pdf/TA130QRGFINAL.pdf,Google Scholar
- 30.Prevoo ML, Van 't Hof MA, Kuper HH, Van Leeuwen MA, van de Putte LB, van Riel PL: Modified disease activity scores that include twenty-eight-joint counts. Development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum. 1995, 38: 44-48. 10.1002/art.1780380107.CrossRefPubMedGoogle Scholar
- 35.Corretti MC, Anderson TJ, Benjamin EJ, Celermajer D, Charbonneau F, Creager MA: Guidelines for the ultrasound assessment of endothelial-dependent flow-mediated vasodilation of the brachial artery: a report of the International Brachial Artery Reactivity Task Force. J Am Coll Cardiol. 2002, 39: 257-265.CrossRefPubMedGoogle Scholar
- 38.Hafstrom I, Rohani M, Deneberg S, Wornert M, Jogestrand T, Frostegard J: Effects of low-dose prednisolone on endothelial function, atherosclerosis, and traditional risk factors for atherosclerosis in patients with rheumatoid arthritis–a randomized study. J Rheumatol. 2007, 34: 1810-1816.PubMedGoogle Scholar
- 40.Gonzalez-Juanatey C, Llorca J, Vazquez-Rodriguez TR, az-Varela N, Garcia-Quiroga H, Gonzalez-Gay MA: Short-term improvement of endothelial function in rituximab-treated rheumatoid arthritis patients refractory to tumor necrosis factor alpha blocker therapy. Arthritis Rheum. 2008, 59: 1821-1824. 10.1002/art.24308.CrossRefPubMedGoogle Scholar
- 41.Cohen SB, Emery P, Greenwald MW, Dougados M, Furie RA, Genovese MC: Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks. Arthritis Rheum. 2006, 54: 2793-2806. 10.1002/art.22025.CrossRefPubMedGoogle Scholar
- 43.Popa C, van Tits LJH, Barrera P, Lemmers HLM, van den Hoogen FHJ, Van Riel PLCM: Anti-inflammatory therapy with tumour necrosis factor alpha inhibitors improves high-density lipoprotein cholesterol antioxidative capacity in rheumatoid arthritis patients. Ann Rheum Dis. 2009, 68: 868-872. 10.1136/ard.2008.092171.CrossRefPubMedGoogle Scholar
- 45.Sandoo A, Carroll D, Metsios GS, Kitas GD, Veldhuijzen van Zanten JJ: The association between microvascular and macrovascular endothelial function in patients with rheumatoid arthritis: a cross-sectional study. Arthritis Res Ther. 2011, 13: R99-10.1186/ar3374.CrossRefPubMedPubMedCentralGoogle Scholar
- 49.Lamarche B, Moorjani S, Cantin B, Dagenais GR, Lupien PJ, Despres JP: Associations of HDL2 and HDL3 subfractions with ischemic heart disease in men. Prospective results from the Quebec Cardiovascular Study. Arterioscler Thromb Vasc Biol. 1997, 17: 1098-1105. 10.1161/01.ATV.17.6.1098.CrossRefPubMedGoogle Scholar
- 50.Watanabe J, Charles-Schoeman C, Miao Y, Elashoff D, Lee YY, Katselis G: Proteomic profiling following immunoaffinity capture of HDL: Association of acute phase proteins and complement factors with pro-inflammatory HDL in rheumatoid arthritis. Arthritis Rheum. 2012, 64 (6): 1828-1837. 10.1002/art.34363.CrossRefPubMedPubMedCentralGoogle Scholar
- The pre-publication history for this paper can be accessed here:http://www.biomedcentral.com/1471-2474/13/127/prepub
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.