Patients with Systemic Lupus Erythematosus (SLE) have increased arterial stiffness, which leads to cardiovascular diseases (CVD) of arteriosclerotic origin, which are the main cause of mortality in these patients. Exercise is a modifiable factor that reduces cardiovascular mortality and associated risk factors in the general population. Preliminary studies suggest that exercise may improve endothelial function and lipid profile in patients with SLE. However, whether meeting the international physical activity guidelines from the American College of Sports Medicine (ACSM; i.e. ≥150 min / week of moderate to vigorous intensity physical activity) can improve arterial stiffness (subclinical atherosclerosis marker) and inflammation is unknown.

The primary aim of this study is to assess the effect of an exercise program based on meeting the ACSM physical activity guidelines on arterial stiffness and inflammation in patients with SLE.

The secondary aim is to assess the effect of an exercise program based on meeting the ACSM physical activity guidelines on endothelial function, oxidative stress, as well as other cardiometabolic risk factors, physical fitness, health-related quality of life, and other psychosocial outcomes.

Our hypothesis is that meeting the ACSM guidelines will improve arterial stiffness and inflammation in patients with SLE.

The study is a non-randomized clinical trial. To minimize selection bias, participants in the intervention and control groups will be matched by age, BMI, and disease activity (SLEDAI), which are important contributors to arterial stiffness.

2. Design and Protocol Registration This non-randomized controlled trial was registered at clinicaltrials.gov [NCT03107442] on 11 April 2017, before the enrolment of participants started (i.e., on 12 April), and no deviations occurred regarding the primary outcome and the secondary outcomes analyzed here.

3. Setting and Eligibility Criteria Participants were recruited from the Systemic Autoimmune Diseases Unit of the "Virgen de las Nieves" and the "San Cecilio" University Hospitals. Women with a diagnosis of SLE according to the ACR criteria [26], a follow-up of 12 months, clinical and treatment stability during the previous six months, and not performing regular exercise (defined as 60 min/week of structured exercise) were included. Exclusion criteria were to have been under biological treatment in the previous six months or to need a prednisone dose of >10 mg/day; a background of CVD in the previous year; to present contraindications to perform exercise; other associated rheumatic conditions; pregnancy; active acute or chronic infection; neoplasms; acute renal failure; cardiac or pulmonary involvement; body mass index (BMI) >35; or not being able to read, understand, and sign written informed consent.

All participants received detailed information about the study procedures, and signed written informed consent. The Research Ethics Committee of Granada approved the protocol on 11 November 2016 (reference No.: 10/2016).

4. Procedures A telephone screening was conducted. Potentially eligible participants were invited to a personal screening and, if included, day 1 of the baseline examination was performed. The baseline examination comprised two assessment days. On day 1, pulse wave velocity (PWV) was assessed. Thereafter, cardiorespiratory fitness testing was performed, and socio-demographic and clinical information was collected. On day 2 (i.e., between two and four days after day 1), 8-h fasting blood samples were collected between 8:00 a.m. and 10:00 a.m. This article follows the TREND statement for improving the reporting of non-randomized experiments of behavioral and public health interventions (downloadable at EQUATOR Network: https://goo.gl/ZSyLrj; Supplementary Table S1) [27]. The funding source had no role in the study.

5. Methods 5.1. Interventions 5.1. 1. Exercise Group To maximize transparency and replicability, the exercise program described in this manuscript follows the Consensus on Exercise Reporting Template (CERT; Supplementary Table S2) [28]. The patients assigned to exercise performed two 75-min sessions per week during a total of 12 weeks (i.e., 24 sessions) of moderate to vigorous intensity aerobic exercise on a treadmill (BH, Serie i.RC12 Dual, Vitoria-Gasteiz, Spain) from 24 April to 14 July 2017. The sessions took place in a quiet room of the "Virgen de las Nieves" Hospital, Granada (Spain).

All sessions were performed in groups of a maximum of five persons (depending on the patients' schedule preferences) and were supervised by both exercise professionals with a degree in Sports Sciences and residents from the Internal Medicine Department.

Attendance at the sessions was registered daily and patients were contacted upon any missing session to ask for the reason and motivate them to replace it on an alternative day of the same week. Adherence to exercise is reported as the median attendance frequency and the proportion of patients attending >75% (i.e., 18 sessions; the minimum pre-defined attendance to assess efficacy) and >90% of the sessions. All the sessions began with a warm-up comprising 3-4 min of activation on the treadmill at about 35-40% of the heart rate reserve (HRR) and 3-4 min of active stretching of major muscle groups, and ended with a cool down phase of static stretching of major muscle groups and relaxation. Exercise was individually prescribed to represent moderate-to-vigorous intensity, with training intensity ranging from 40% to 75% of each patient's HRR. The maximum heart rate (HRmax) was estimated with the formula by Tanaka et al. (HRmax = 208 x (0.7 x age)) [29]. The training (or target) heart rate (tHR) was calculated with the formula tHR = HRrest + (%HRR). Heart rate was continuously monitored during all sessions (Polar V800, Kempele, Finland). We used the session rating of perceived exertion (RPE) as a measure of subjective training load [30], and the feeling scale to assess positive affective responses experienced before and after each session [31].

The starting level was specific for each individual according to her previous exercise experience and physical fitness. During the first half of the program, only continuous exercise was performed so that the patients got used to the treadmill and felt confident at increasing intensities. Continuous sessions comprised several bouts of exertion at constant intensity, followed by a couple of minutes of recovery (i.e., rest) to drink water. During month 2, there were alternated continuous and interval sessions, and at month 3, the patients undertook interval training sessions, where there were periods of lower and periods of higher intensity efforts followed by some minutes of rest for hydration (Table 1). The progression in volume and/or intensity was patient-limited and was undertaken by increasing the treadmill speed (first) or inclination according to the symptoms and perceived exertion. There were no home-based or non-exercise components within this intervention. However, if a participant was eventually not able to attend a particular session, we provided her with a heart rate monitor and allowed recovery of that session out of the Hospital (a total of seven sessions were recovered in this fashion). Finally, the exercise intensity progressions had to be slightly modified from the initial plan. For instance, several patients perceived a 5% HRR intensity increase (i.e., from one week to another) as very heavy and difficult-to-follow. Consequently, there were weeks in which exercise intensity increased by 2.5% instead of 5% (Table 1).

5.1.2. Control Group After the baseline evaluation, the SLE patients assigned to the (usual care) control group received verbal information about a healthy lifestyle, including physical activity guidelines and basic nutritional information.

5.2. Outcome Measures Primary Outcome Measure: Arterial Stiffness Arterial stiffness was assessed in a sitting position by PWV [9], using the Mobil-O-Graph® 24 h pulse wave analysis monitor (IEM GmbH, Stolberg, Germany), whose operation is based on oscillometry recorded by a blood pressure cuff placed on the brachial artery. The coefficient of variation (CV) of Mobil-O-Graph for consecutive PWV analyses is 3.4% and its intraclass correlation coefficient is 0.98 [0.96-0.99] [32]. This device has been largely shown to be valid and reliable for measuring PWV and central blood pressure in different populations [33,34], meets the accuracy requirements of the British Hypertension Society (BHS) standard [35], and can be recommended for clinical use [36].

Secondary Outcome Measures Blood Samples and Biochemical Analyses Fasting blood specimens for biochemical and immunological tests were collected and routinely processed by the central laboratory of our hospital. Among other measurements, they included lipids, insulin (BioRad, Marne-la-Coquette, France), and a routine biochemical profile. The homeostatic model assessment for insulin resistance (HOMA-IR) was calculated (HOMA-IR = glucose (mmol/L) insulin ( U/L)/22.5).

Inflammatory Markers Serum high-sensitivity CRP was assessed by an immunoturbidimetric method using the ARCHITECT cSystems (MULTIGENT CRP Vario assay); the limit of quantitation was 0.2 mg/L and the upper limit for normal serum was 5 mg/L (coefficient of variation <6%). Interleukin 6 and TNF-

, as well as myeloperoxidase (MPO; as marker of oxidative stress), were measured in plasma. Serum was initially separated by centrifugation and stored at 70 C.

Bioserum concentrations of IL-6/TNF- (pg/mL) and MPO (ng/mL) were measured by an immunoradiometric assay using commercial kits (MILLIPLEX MAP Kit Human High Sensitivity T Cell Magnetic Bead Panel (HSTMAG-28SK) and Human Cardiovascular Disease Magnetic Bead Panel 2 (HCVD2MAG-67K)), Millipore) following the manufacturer's instructions. Quantitative data were obtained by using the Luminex-200 system (Luminex Corporation, Austin, TX, USA), and data analysis was performed on XPonent 3.1 software (Austin, TX, USA). The detections limits were 0.73 pg/mL for IL-6, 0.43 pg/mL for TNF-a, and 0.024 ng/mL for MPO.

Cardiorespiratory Fitness Cardiorespiratory fitness was assessed with the Bruce submaximal treadmill protocol [37]. The test comprised five increasing workload stages of 3 min each (stage 1: 2.7 km/h and 10% inclination; stage 2: 4 km/h and 12% inclination; stage 3: 5.5 km/h and 14% inclination; stage 4: 6.8 km/h and 16% inclination; stage 5: 8 km/h and 18% inclination). The test concluded when the participant achieved 85% of the individual's HRmax, as estimated with the formula by Tanaka et al. [29]. As validated SLE-specific formulas to estimate VO2max are not available, we used the total time to reach 85% HRmax as the outcome of interest.

Other Measurements All participants filled out a socio-demographic and clinical data questionnaire. Height (cm) was measured using a height gauge, weight (kg) with a bioimpedance device (InBody R20, Biospace, Seoul, Korea), and body mass index (BMI) was calculated (kg/m2). Blood pressure was measured with Mobil-O-Graph (IEM GmbH, Stolberg, Germany) [36]. Disease activity was assessed through the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI, range 0-105 where a higher score indicates higher degree of disease activity). Physical activity was self-reported at baseline and at week 12 with the International Physical Activity Questionnaire [38].

Sample Size The sample size was calculated for the primary outcome (i.e., PWV). Ashor et al. found an average effect of aerobic exercise on PWV of 0.63 m/s in adults aged 18 years [21]. A total of 52 patients (26 per group) were needed to detect an effect of 0.63 (SD 0.75) m/s, with a power of 85% and an error of 0.05. Anticipating a maximum loss to follow-up of 15%, we aimed at recruiting a total of 60 patients.

Treatment Allocation and Blinding Randomization was not feasible because more than half of the patients who regularly attend the Autoimmune Disease Units lived far from the Hospital and were not able to attend twice per week in case of being randomized to exercise. Therefore, participants from the city of Granada were included in the exercise group and participants living outside Granada were included in the control group.

To minimize potential selection bias, we aimed to match the groups by age ( 2 years), BMI ( 1 kg/m2), and SLEDAI ( 1 unit). The data analyzer was blinded to the patient allocation.

Inclusion Criteria:

• Diagnosis of Systemic Lupus Erythematosus (SLE) according to the American College of Rheumatology criteria (presenting at least 4 classification criteria).
• Follow-up of at least 12 months at our Unit.
• Clinical and treatment stability during the 6 months prior to the study.
• Not performing regular exercise (defined as < 60min/week of structured exercise)

Exclusion Criteria:

• Biological treatment in the previous 6 months or to need prednisone dosis >10 mg/day.
• Background of clinical cardiovascular disease in the last year.
• To present contraindications to perform exercise.
• Other associated rheumatic conditions.
• Pregnancy.
• Acute renal failure.
• Cardiac or pulmonary involvement.
• Body Mass Index > 35
• Not being able to read, understand and sign written informed consent.

Adverse event data were collected during 3 months.

Any untoward or unfavorable medical occurrence in a participant, including any abnormal sign (for example, abnormal physical exam or laboratory finding), symptom, or disease, temporally associated with the participant's participation in the research, whether or not considered related to the participant's participation in the research.

This study has several limitations. First, the final sample size was relatively small and was calculated to assess the between-group change in arterial stiffness as the primary trial outcome. Consequently, the power to detect significant between-group differences in the outcomes reported was checked.