Efficacy and Safety of a Highly Bioavailable Curcumin Formulation in Modulating Outcomes of Mild Knee Osteoarthritis: Multi-Centric, Randomized, Double-Blind, Placebo-Controlled Study

in Modulating Outcomes of Mild Knee Osteoarthritis: Multi-Centric, Randomized, Double-Blind, Placebo-Controlled Study


Background
Osteoarthritis (OA) is a degenerative disease of joints often associated with underlying inflammation and progressive cartilage degradation. Osteoarthritis begins with cartilage thinning, joint space narrowing, inflammatory reactions [1][2][3][4][5][6][7][8], loss of joint architecture, and deformation of the knee [9]. Osteoarthritis primarily affects the aging population worldwide with higher incidences in females and leading cause of morbidity and healthcare cost across the globe [10][11][12]. Current OA management includes pharmacological interventions through analgesics, non-steroidal anti-inflammatory drugs (NSAIDs), etc. While these therapies provide temporary relief from pain and improve physical function, they do not address underlying conditions and are frequently associated with several adverse events over long-term use in many individuals [13][14][15][16][17][18].
Several plant-based multi-functional anti-inflammatory products have been developed in the past which are being used by patients suffering from knee osteoarthritis with limited success [19]. Curcumin, a polyphenolic compound from turmeric (Curcuma longa) has been extensively used to treat knee OA [20][21][22][23][24]. Curcumin has been used for centuries in traditional Chinese and Ayurvedic medicine for its anti-inflammatory properties [25].
Meta-analysis of clinical studies has demonstrated the protective role of curcumin in OA with decreased VAS, and WOMAC scores, reduced inflammatory factors such as ILs, TNF-α, MMPs, and increased antioxidant activity as compared to placebo [26]. Curcumin inhibits the synthesis of inflammatory mediators such as interleukin (IL)-1β, tumor necrosis factor (TNF)-α, IL-6, IL-8, prostaglandin E2 (PGE2), cyclooxygenase-2 (COX-2) [27 -29], and IL-1β-induced extracellular matrix degradation [30]. The antiinflammatory function of curcumin is primarily mediated through the inhibition of activator protein 1 (AP-1) [31] and nuclear factor kappa B (NF-kB) activation [31][32][33][34]. Further, curcumin suppresses the expression of a number of matrix metalloproteinases (MMPs), which play critical role in the breakdown of cartilage extracellular matrix in OA [7,28,[31][32][33][35][36][37][38][39][40]. Several human clinical studies have demonstrated efficacy of curcumin leading to reduced pain and improved physical function, and quality of life in OA patients with the efficacy of curcumin shown to be similar to that of ibuprofen [41][42]. Curcumin has been found to be chondroprotective possibly due to their anti-inflammatory, antioxidant and anti-catabolic activity that helps in mitigating OA disease pathogenesis and symptoms [43,44]. Further, curcumin also functions as an antioxidant by scavenging reactive oxygen and nitrogen species that play an important role in the genesis of OA [20].
Despite therapeutic potential, poor bioavailability of curcumin due to low absorption, rapid metabolism, and elimination in the body limits its clinical use [10]. Curcumin is highly hydrophobic and hence insoluble in water [11], unstable in alkaline physiological pH of the small intestine [12,45] resulting in very low plasma curcuminoids levels even after consuming large doses. Numerous formulation approaches have been developed to improve the bioavailability of curcumin, such as the use of piperine that interferes with glucuronidation, development of formulations based on liposomes, nanoparticles, and phospholipid complex [46]. Previously a newly formulated curcumin formulation named as Curcuwin Ultra+ (CU+) was reported with 40% faster intestinal absorption and 144 times higher plasma curcuminoids concentration in human subjects as compared to 95% turmeric extract [47]. Further, CU+ supplementation restored joint architecture and reduced swelling of joint in monosodium iodoacetate (MIA)-induced knee OA in rats with a significant reduction in the levels of inflammatory mediators such as TNF-α, IL-1β, IL-6, COMP, and CRP, and expressions of MMP-3, 5-LOX, COX-2 and NFkB in synovial tissue [48]. The current study is designed to demonstrate the improved efficacy of CU+ due to enhanced bioavailability and faster absorption in human subjects suffering from mild knee OA on parameters of joint discomfort, cartilage degradation, inflammation, joint mobility, flexibility, knee muscle strength, quality of life, and use of recuse medication for pain.

Study Design and Procedures
This was a prospective, randomized, double-blind, parallel, placebo-controlled, clinical interventional study of two doses of CU+ in subjects with mild knee osteoarthritis. The study was initiated after obtaining written approval from respective institutional ethics committees of each site. The study was carried out as per the requirements of the Indian Council of Medical Research (ICMR) ethical guidelines, International Council for Harmonization (ICH) 'Guidance on Good Clinical Practice' (E6R2), and 'Declaration of Helsinki'. The study was registered with the Clinical Trials Registry of India (CTRI/2021/10/037187).
Voluntary informed consent was obtained from every participant before enrolment in the study. Subjects were randomly assigned in a 1:1:1 ratio to receive CU+ 250 mg (total of 50 mg curcuminoids) capsule or CU+ 500 mg (total of 100 mg curcuminoids) capsule or placebo. Curcuwin Ultra+ ® , manufactured by OmniActive Health Technologies Ltd., Mumbai, India contains not less than 20% total curcuminoids and composed of turmeric extract encapsulated in matrix of cellulose polymer, solubilizer, acidifier and antioxidant. The placebo capsules were filled with microcrystalline cellulose in scarlet red colored hard gelatin capsules by OmniActive Health Technologies to look like CU+ capsules and keep the study double-blinded.
The study visits were scheduled as -Screening/baseline visit, Placebo run-in visit, Randomization visit, follow up visits at Day 5, 28, 56 and End of study visit at Day 84. Demographic information such as gender, age, body weight, height, BMI, medical history, concomitant medication history, and physical activities, presence of allergic problems and drug reactions, and oral contraceptives were obtained during the screening visit. After the screening visit, the eligible subjects were provided with placebo capsules for the run-in period and instructed to consume one capsule every morning after breakfast for 7 days. Once the run-in period was completed, the eligible subjects were randomized by study investigators in a double-blinded fashion to receive one of the study products as per the randomization schedule. The randomization schedule was generated using block randomization by a non-study assigned, independent expert ensuring the treatment balance by using SAS® Version 9.4. One bottle of study product containing 35 capsules was dispensed to every subject at Day 0, Day 28 and Day 56 as per the randomization schedule. Subjects were instructed to consume one study capsule every morning after breakfast at the same time every day for 84 days (12 weeks). Subject compliance was calculated by performing study product reconciliation at each follow up visit.
Acetaminophen was prescribed as rescue medication for the subjects who could not withstand pain, and the details of use were recorded.

Inclusion/Exclusion Criteria
Subjects who met all inclusion and none of the exclusion criteria were enrolled in the study after signing a written informed consent. Inclusion criteria were as follows: male and female subjects aged between 35 and 75 years (both limits inclusive), clinical diagnosis of unilateral or bilateral knee osteoarthritis (OA) for at least greater than 3 months, as presented by pain in the knee at least for last 3 months, knee joint pain rated ≤ 44 on a 100-point Pain-VAS scale, knee OA grade II as per Kellgren-Lawrence classification, subjects involved in regular physical activity such as walking, climbing stairs, recording forms, eating, exercise behaviors, etc, subject who could walk for at least 6 minutes at a moderate-to-brisk pace on a treadmill, subjects willing to avoid NSAIDs and other anti-inflammatory medications during the study period and subjects who were 100% compliant with an investigational product during the Run-In Period.
The following criteria were used to exclude subjects in the study: Subjects with moderate to severe knee OA in Grade III or IV as per Kellgren and Lawrence Scale, previous history of knee OA for more than 3 years, previous history of any knee joint replacement surgery or bilateral hip joint replacements, positive urine pregnancy test at screening visit, subjects unwilling to refrain from the use of NSAIDs or herbal/ nutraceutical supplements for joint health or local analgesics during the study duration, habit of smoking or consuming tobacco products, any other chronic disease that could jeopardize the outcomes of the study.

Efficacy and Safety Parameters
Primary efficacy endpoint included Western Ontario and McMaster Universities Osteoarthritis (WOMAC) index measured for one knee joint at baseline visit, day 5, 28, 56 and 84. Secondary efficacy endpoints included Visual Analogue Scale for pain (VAS -pain), Knee Extension and Flexion Muscle Strength, Range of Motion (ROM) and 6 Minutes' Walk Test (6MWT) measured at baseline visit, day 5, 28, 56 and 84. Further, serum hsCRP and IL-1β were evaluated at baseline visit, day 5, and 84 whereas serum CTX-II, COMP, and MMP-3 were evaluated at baseline visit and day 84. Rescue medication usage was evaluated throughout the study using subject diary.
As part of safety assessments, adverse events were collected throughout the study using subject diary. Physical examination and vital signs measurement was conducted at all study visits. Laboratory assessments, and urine pregnancy tests for females of childbearing age were performed at screening visit and end of study visit.

Statistical Analysis
A sample size of 135 subjects (45 subjects for CU+ 250 mg, 45 subjects for CU+ 500 mg and 45 subjects for the Placebo group) was based on the assumption that there will be a mean difference in total WOMAC index of 10% between the CU+ formulations and placebo from baseline to end of treatment. With an expected inter-subject standard deviation of 30%, around 37 evaluable subjects per group would be required to estimate the reduction of total WOMAC index score at a 5% level of significance at the end of the study at 80% power. Based on the parallel design and considering possible dropouts, a sample size of 45 subjects was enrolled in each treatment group.
Baseline characteristics were presented separately for the study groups. Continuous variables were summarized using descriptive statistics which included number of observations, mean, median, minimum, maximum, and standard error. Categorical variables were summarized using frequencies and percentages. All statistical analyses were performed using SAS®, version 9.4. Subjects with missing data were excluded only from analyses for which data were not available.
The differences in the analysis variables were tested by using paired t-test and one-way ANOVA (Analysis of Variance) model using PROC ANOVA in SAS®. Kruskal-Wallis test was executed for the ordinal datasets. For the dichotomous variable Z-test was performed using SAS®. The criterion for the significant test by treatment was set at a p-value<0.05.

Efficacy Endpoints Evaluation
Licensed WOMAC OA Index version VA 3.1 English for India was used as per WOMAC® User Guide XII for maximum of 5 occasions for each subject. WOMAC questionnaire contained 24 items grouped into three sub-categories: pain, stiffness, and physical function (total score range 0-2400). The total WOMAC score was calculated by adding the subscale scores (5 -pain; 2stiffness; 17 -physical function). Visual analogue scale for pain was measured using 100 mm pain scale with a calibrated ruler to maintain the uniformity of measurement throughout the visits for all subjects. The VAS scale for pain was calculated for both the right and left legs at each visit for all the subjects. MicroFET2 TM Wireless Hand-Held Dynamometer (Hoggan Scientific, Salt Lake City, UT, USA) was used for force evaluation (strength) test during the extension and flexion of both dominant and non-dominant legs for each subject at each visit as per the product's user manual.
Three readings were recorded, and the average was calculated for each leg at each visit.
For the measurement of knee extension and flexion angles, a calibrated goniometer was used. The subjects were allowed to sit in an upright position on a table edge with their backs straight and knee position defined at 90⁰, and the axis of a goniometer was placed at the intersection of the thigh and shank at the knee joint. The subjects were asked to flex or extend their legs to the fullest ability without changing the position. Three readings were obtained, and the average was calculated for each leg at each visit.
For 6MWT, the participants were asked to walk on treadmill for 6 minutes continuously while making sure that at least one foot was bearing weight at all the times and the distance was measured in meters for each subject at each visit.
For biomarker analysis, serum was separated from blood samples and stored below -20⁰C until analysis. KRISHGEN BioSystems (Mumbai, MH, India) GENLISA TM ELISA kits were used for IL-1β, MMP-3, CTX-II, and COMP analysis and Calbiotech (El Cajon, CA, United States) ELISA kits and reagents were used for hsCRP. Samples were analysed at baseline and Day 84 while hsCRP and IL-1β was additionally analysed on Day 5.
The subjects were only prescribed acetaminophen as a rescue medication if the symptoms were unbearable for ethical reasons. The subjects were allowed to discontinue the consumption of rescue medication as per their willingness and if they sensed improvement in the symptoms. The rescue medication consumption was evaluated throughout the study.

Safety Analysis
Safety analyses included haematology, biochemistry, and urinalysis, the incidence of adverse events, physical examination, and vital sign measurements. Physical examinations and vital parameter assessments were performed at all the visits while blood sampling for laboratory tests were performed at baseline and end of the treatment (day 84), allowing evaluation versus baseline. Adverse events details were obtained throughout study duration using subject diary.
Descriptive statistics included number of subjects, mean, standard deviation, median, minimum, and maximum for continuous safety variables and frequency, and percentage for categorical safety variables such as adverse events were summarized by treatment.

Results
One hundred fifty-eight subjects were screened, out of which 23 subjects did not fulfil the eligibility criteria and were considered as screen failures (Figure 1). Out of the 135 randomized subjects, 45 subjects were allocated to the CU+ 250 mg, 45 subjects were There was no statistical significance among the three groups for age, and BMI at the baseline.

Efficacy
The efficacy analysis was performed for 134 subjects who completed the study as 1 subject from the placebo group withdrew consent due to personal reason.

Total WOMAC Score
The within-group analysis indicated that subjects in both CU+ 250 mg and 500 mg groups showed a significant decrease in total WOMAC score on Days 5, 28, 56, and 84 (p<0.05), while subjects in placebo group showed a significant decrease only on Day 5 (p<0.05) as compared to baseline.
According to the between-group analysis, subjects in the CU+ 250 mg and 500 mg groups showed significant decrease in their total WOMAC score on Days 5, 28, 56, and 84 compared to placebo (p<0.05), and CU+ 500 mg significantly outperformed CU+ 250 mg on Days 56 and 84 (p<0.05) (Supplementary Table  S2a; Supplementary Figure 1; Figure 2a).

WOMAC Pain Score
When compared to baseline, WOMAC pain scores for subjects in the CU+ 250 mg and 500 mg groups significantly decreased on Days 5, 28, 56, and 84 (p<0.05). In contrast, WOMAC pain scores for subjects in the placebo group did not significantly decrease at any visit.
In comparison to the placebo group, subjects in the CU+ 250 mg and 500 mg groups demonstrated a statistically significant decrease in WOMAC pain score on Days 28, 56, and 84 (p<0.05). In addition, CU+ 500 mg performed significantly better than CU+ 250 mg on Days 56 and 84 (p<0.05) (Supplementary Table S2b; Supplementary Figure 1; Figure 2b).

WOMAC Stiffness Score
According to the within-group analysis, subjects in the CU+ 250 mg and 500 mg groups demonstrated a significant decrease in WOMAC stiffness score on Days 5, 28, 56, and 84 (p<0.05), while those in the placebo group demonstrated a significant decrease only on Day 5 (p<0.05) as compared to baseline.
In comparison to the placebo group, the CU+ 500 mg group experienced a significant decrease in WOMAC stiffness score as early as Day 5 (p<0.05). Furthermore, both CU+ 250 mg and 500 mg groups significantly decreased (p<0.05) WOMAC stiffness score on Days 28, 56, and 84. Additionally, CU+ 500 mg demonstrated a significantly greater reduction in WOMAC stiffness score on Day 84 compared to CU+ 250 mg (p<0.05) (Supplementary Table S2c; Supplementary Figure 1; Figure 2c).

WOMAC Physical Function Score
The within-group analysis indicated that subjects in CU+ 250 mg and 500 mg groups showed a significant decrease in WOMAC physical function score on Days 5, 28, 56, and 84 (p<0.05), while subjects in placebo group showed a significant decrease only on Day 5 (p<0.05) as compared to baseline.
In comparison to the placebo group, the CU+ 250 mg and 500 mg groups showed significant decreases in WOMAC physical function score on Days 28, 56, and 84 (p<0.05). Additionally, this significant difference was seen as early as Day 5 (p<0.05) in the CU+ 500 mg group. Further, CU+ 500 mg demonstrated a significantly greater reduction in WOMAC physical function score on Day 84 compared to CU+ 250 mg (p<0.05) (Supplementary Table S2d; Supplementary Figure 1; Figure 2d).

Visual Analogue Scale for Pain
When compared to baseline, pain scores for subjects in all the study groups (CU+ 250 mg, CU+ 500 mg and placebo) significantly decreased on Days 5, 28, 56, and 84 (p<0.05).
In comparison to placebo, CU+ 500 mg group showed a significant decrease in pain score as early as Day 5 (p<0.05), whilst the CU+ 250 mg and 500 mg groups showed significant reductions in pain score on Days 28, 56, and 84 (p<0.05). Additionally, CU+ 500 mg demonstrated significantly better reduction in pain than CU+ 250 mg on Day 84 (p<0.05) for the right leg and Days 5, 56, and 84 (p<0.05) for the left leg (Supplementary Table S2e

Knee Extension (Dominant Leg)
The within-group analysis revealed that subjects in the CU+ 250 mg and 500 mg groups demonstrated significant increase in knee extension muscle strength for the dominant leg on Days 5, 28, 56, and 84 (p<0.05), while subjects in the placebo group demonstrated significant increases on Days 5 and 56 (p<0.05) compared to baseline.
Compared to placebo, the CU+ 500 mg group demonstrated a significant increase in knee extension muscle strength for the dominant leg on Day 5 (p<0.05). In addition, both 250 mg and 500 mg CU+ groups showed a significant increase in knee extension muscle strength on Days 28, 56, and 84 (p<0.05). Further, CU+ 500 mg outperformed CU+ 250 mg by a statistically significant margin on Days 5, 28, 56, and 84 (p<0.05) (Supplementary Table  S2g; Figure 4a).

Knee Extension (Non-Dominant Leg)
The within-group analysis revealed that subjects in all the study groups (CU+ 250 mg, Cu+ 500 mg and placebo) demonstrated significant increase in knee extension muscle strength for the nondominant leg on Days 5, 28, 56, and 84 (p<0.05) compared to baseline.
Both CU+ 250 mg and 500 mg groups showed significant increase in knee extension muscle strength for the non-dominant leg on Days 28, 56, and 84 (p<0.05) as compared to the placebo group where CU+ 500 mg significantly outperformed CU+ 250 mg on Days 56 and 84 (p<0.05) (Supplementary Table S2h; Figure  4b).

Knee Flexion (Dominant Leg and Non-Dominant Leg)
The within-group analysis indicated that subjects in all  Table S2i and S2j; Figure  4c and 4d).

Knee Extension (Dominant Leg)
When compared to baseline, CU+ 250 mg did not show any significant decrease in knee extension ROM for the dominant leg at any visit whereas CU+ 500 mg showed significant decrease on Day 5 (p<0.05). In contrast, placebo group showed a significant increase in knee extension from baseline on Day 5 (p<0.05).
The CU+ 500 mg group, but not the CU+ 250 mg group, showed a significant reduction in knee extension ROM for the dominant leg on Days 5 and 28 (p<0.05) compared to the placebo group. Further, neither CU+ groups differed significantly from the placebo on Days 56 or 84. (Supplementary Table S2k; Figure 4e).

Knee Extension (Non-Dominant Leg)
No significant differences in knee extension ROM for the non-dominant leg were seen in CU+ 250 mg, CU+ 500 mg and placebo groups from baseline or between CU+ groups compared to placebo during the study period (Supplementary Table S2l; Figure 4f).

Knee Flexion (Dominant Leg)
According to the within-group analysis, CU+ 250 mg The CU+ 500 mg group showed significant increase in knee flexion ROM for the dominant leg on Days 56 and 84 (p<0.05) as compared to placebo whereas CU+ 250 mg group did not show any significant difference throughout the study period. Moreover, CU+ 500 mg group performed significantly better than CU+ 250 mg group on Days 5, 28, and 84 (p<0.05) (Supplementary Table  S2m; Figure 4g).

Knee Flexion (Non-Dominant Leg)
According to the within-group analysis, CU+ 250 mg showed significant increase in knee flexion ROM for the nondominant leg on Days 56 and 84 (p<0.05), CU+ 500 mg on Days 28, 56 and 84 (p<0.05) and placebo group on Day 28 (p<0.05) compared to baseline.
On day 84, both the CU+ 250 mg and 500 mg groups demonstrated a significant increase (p<0.05) in knee flexion ROM for the non-dominant leg, however only the CU+500 mg group significantly increased the knee flexion ROM on Days 28 and 56 (p<0.05) compared to the placebo group. Furthermore, CU+ 500 mg demonstrated a substantially greater increase in knee flexion ROM than CU+ 250 mg on Days 56 and 84 (p<0.05) (Supplementary Table S2n; Figure 4h).

Distance covered over a time of 6 minutes' walk on the treadmill
When compared to baseline, CU+ 250 mg and 500 mg groups showed significant increase in walking distance on Days 5, 28, 56 and 84 (p<0.05). Further, placebo group showed a significant increase in walking distance from baseline on Days 5 and 28 (p<0.05).

Serum high-sensitivity C-reactive protein (hsCRP)
Within-group analysis revealed that subjects in the CU+ 250 mg and 500 mg groups showed significant decrease in hsCRP levels on Days 5 and 84 (p<0.05) compared to baseline, while subjects in the placebo group did not show any significant differences versus baseline.
Compared to placebo, the CU+ 500 mg group had a significant reduction in hsCRP levels on Day 84 (p<0.05) and a non-significant reduction on Day 5 (p=0.0552). The CU+ 250 mg group did not demonstrate any significant reduction in hsCRP levels compared to the placebo group. There was no statistically significant difference between the CU+ groups (Supplementary Table S2p; Figure 5a).

Serum Interleukin-1β (IL-1β)
Within-group analysis indicated that subjects in the CU+ 500 mg group showed significant decrease in IL-1β levels on Days 5 and 84 (p<0.05) compared to baseline, while subjects in the CU+ 250 mg and placebo groups did not show any significant differences versus baseline.
Compared to placebo, the CU+ 500 mg group showed a significant reduction in IL-1β levels on Day 84 (p<0.05) and a non-significant reduction on Day 5 (p= 0.0654). The CU+ 250 mg group did not demonstrate any significant reduction in IL-1β levels compared to the placebo group. No significant difference was seen between the CU+ groups (Supplementary Table S2q; Figure 5b).

Serum Carboxy-Terminal Telopeptides of Type II Collagen (CTX-II)
Within-group analysis revealed that subjects in the CU+ 250 mg and 500 mg groups showed significant decrease in CTX-II levels on Day 84 (p<0.05) compared to baseline, while subjects in the placebo group did not show any significant difference versus baseline.
Both the doses of CU+ showed significant decrease in CTX-II on Day 84 (p<0.05) as compared to placebo. No significant difference was seen between the CU+ groups (Supplementary Table S2r; Figure 5c).

Serum Cartilage Oligomeric Matrix Protein (COMP)
When compared to baseline, all the study groups showed significant increase in COMP levels on Day 84 (p<0.05).
No significant differences were seen between CU+ groups and placebo on Day 84 (Supplementary Table S2s; Figure 5d).

Serum Matrix Metalloproteinase-3 (MMP-3)
When compared to baseline, both the CU+ 250 mg and 500 mg groups showed significant reduction in MMP-3 levels on Day 84 (p<0.05), while no significant difference was observed in the placebo group.
When compared to placebo on Day 84, serum MMP-3 levels were significantly reduced in the CU+ 500 mg group (p<0.05), but not in the CU+ 250 mg group. No significant differences were seen between the CU+ groups (Supplementary Table S2t; Figure 5e).

Safety
The safety analysis set consisted of all subjects who were randomized for the trial and consumed at least one dose of study product. As far as safety is concerned, the composition of CU+ was found to be safe and well-tolerated. A total of 3 subjects reported adverse events, out of which 02 subjects were in the CU+ 500 mg group and 01 subject was in the placebo group. One adverse event (loss of appetite) was possibly related to the CU+ 500 mg group and the remaining 2 adverse events (mouth ulceration and headache) were unrelated to study products.
No severe adverse events were observed, and when occurred were of mild intensity and resolved completely. There were no notable changes in laboratory parameters, vital signs, physical examination, and systemic examination.
Current treatment approaches of OA use primarily analgesics and NSAIDs that provide only temporary relief. Curcumin is one of the most popular herbal supplements with widely established biological properties including anti-inflammatory and antioxidant activity and traditionally used for management of OA [20][21][22][23][24]. To overcome the challenges of poor oral bioavailability of curcumin, we developed an improved curcumin formulation that demonstrated 40% faster intestinal absorption and achieved 144 times higher plasma curcuminoids levels in human subjects as compared to 95% turmeric extract. Here we demonstrate that our improved formulation is associated with enhanced efficacy against mild knee OA at both 250 and 500 mg dose with 500 mg performing better than 250 mg dose as measured through widely used OA efficacy parameters.
Widely established WOMAC parameters [49] were used as primary endpoint of efficacy followed by VAS score for determination of relief from pain during CU+ supplementation in subjects with mild OA. We observed that CU+ doses at 250 and 500 mg showed significant (p<0.05) decrease in all WOMAC parameters as well as VAS pain score on day 28, 56 and 84 as compared to placebo. We believe that WOMAC parameters and VAS pain were significantly improved due to the anti-inflammatory effect of CU+ as was demonstrated by reduced levels (p<0.05) of serum inflammatory biomarkers (hsCRP and IL-1β) on Day 84. The anti-inflammatory effect of curcumin has been demonstrated previously in human studies with decreased serum concentrations of inflammatory mediators such as TNF-α, IL-1β, IL-6, CRP, MMP-3 in response to curcumin supplementation in subjects with knee OA [48,50,51].
Muscle weakness is common in subjects with knee OA and strongly related to the patient-reported outcomes of pain, activity limitations and falls. Muscle strength is the ability to exert a physical force which decreases with age as well as subjects with OA probably due to reduced physical activity [52]. The assessment of muscle strength is extensively used for patient management in OA [53]. Previous studies demonstrate that curcumin not only boosts endurance and muscle strength but also reduced muscle damage possibly due to its antioxidant and anti-inflammatory activity [54]. In the current study we observed significant improvement of knee muscle strength as measured through knee extension and flexion after supplementation with CU+ which further validates previous results where a formulated curcumin at 1000 mg dose improved muscle performance with reduced muscle soreness after the exercise [53]. The improved muscle performance was probably also responsible for improved walking distance in subjects supplemented with CU+ as demonstrated in 6-Minute Walk Test, a sub-maximal exercise test used to assess aerobic capacity and endurance of subjects [52,55,56]. Joint flexibility is critically affected in OA due to pain and stiffness [57,58] and regularly used to assess treatment outcome [57][58][59]. The reduced pain and stiffness of joints observed in subjects supplemented with CU+ probably due to reduced inflammation may have also helped to improve range of motion in knee flexion for both legs found in our study.
Surrogate molecular biomarkers such as serum levels of cytokines, MMPs, CRP not only play an important role in physiology of OA but also help in patient diagnosis and clinical management [20,[59][60][61][62][63]. We observed that CU+ lowered levels of serum hsCRP and IL-1β and cartilage damage biomarker CTX-II, and MMP-3, a proteolytic enzyme responsible for the breakdown of cartilage.
Throughout the study we observed that CU+ supplementation is associated with significant (p<0.05) reduction in the use of rescue medication by subjects at both doses of CU+ as compared to placebo. Further CU+ at both doses was well tolerated and safe throughout the study period with no serious adverse events reported.
Overall, we have observed significantly improved efficacy measures in knee OA subjects supplemented with CU+ which is in line with previous clinical trials where improvement in pain, stiffness, and muscle function was observed after 12 weeks of supplementation with curcumin [26,64,65]. The improved benefits observed in our study with lower doses might be due to the superior bioavailability and faster absorption of curcuminoids in CU+ formulation as compared to regular curcumin extracts [47,66].
Current study is limited by recruitment of study subjects suffering from only mild knee OA and supplementation for 84 days. Further studies are required to understand the scope of supplementation of CU+ in subjects suffering from moderate to severe form of OA and supplementation period beyond 84 days.

Conclusions
The results of this study demonstrate that CU+ supplementation not only reduces discomfort, but also improves joint mobility, flexibility, and knee muscle strength, resulting in an overall improvement in the quality of life for participants with mild OA. Our results indicate that CU+ reduced serum inflammatory markers, markers of cartilage degradation that are associated with pathophysiology of OA during the study period. Throughout the study CU+ was safe, well-tolerated and reduced the requirement of rescue medication in subjects suffering from mild OA.

Acknowledgments
We thank the participants of the study.

Authors' contributions
SKD, NJ, AS & SB contributed to the study design. Study conduct, subject recruitment and data collection were performed at respective study centres by NJ, AS & SB. Study was monitored by SKD in a blinded fashion. Statistical analysis and study report were prepared by SKD. The first draft of the manuscript was written by SKD, NJ, AS & SB and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding
The study was supported by OmniActive Health Technologies Limited (Mumbai, India).

Availability of Data
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request. All data generated or analyzed during this study are included in this published article and its supplementary information files.