Introduction

Osteoarthritis (OA) is a highly heterogeneous disorder characterized by progressive cartilage loss, remodeling of adjacent bones and concomitant local low-grade inflammation. From a concept of mechanical wear and tear process, the idea of OA has evolved through the last few years to a whole organ disease affecting different joint structures and functions, being in interaction with the human body as a whole [1, 2]. OA affects people of all ages with different levels of physical activity and the knee is among the joints most commonly affected [3, 4].

To date, every single detail of the complex structural changes and the processes of articular remodeling in knee OA (KOA) are easily recognized thanks to the high-resolution imaging modalities and innovative biomarkers, respectively. The area with the most unsatisfactory progress seems to remain the non-surgical treatment of the disease. Although we already have a greater understanding of the pathogenetic mechanisms behind the scenes of OA development and progression, modern conservative management approaches serve mainly to relieve pain rather than to influence the biochemical environment of the joint and to impact the disease progression. In the last few years, a ray of hope is emerging on the research front as orthobiotics, stem cell-based approaches and gene therapy look promising. However, they may never be applicable or accessible for the majority of patients with KOA and their efficacy and safety is still in question. What if there are safe and potent conservative, non-pharmacological treatment options right before our eyes? They may often be overlooked by researchers and clinicians due to the temptation of investigating only innovations and presenting them to patients.

As has already been emphasized, OA is a multifactorial disease that affects the whole joint. Therefore, the expectations for the optimal treatment strategy are very high, since it has to impact a large number of tissues and functions. Modern management of KOA should be based on an individualized approach, including not only palliative pharmacological relief of pain but also identifying and managing risk factors contributing to the patient’s condition, resulting in improvement of the metabolic and biomechanical environment of the joint. In the search for the optimal treatment strategy, clinicians should aim at affecting where possible risk factors that could be potentially intervened.

The objectives of this narrative review were to identify, summarize, and cluster all the modifiable risk factors that influence the course of KOA, and discuss their susceptibility to intervention via personal, clinical and public strategy.

Search methodology

We adhered to the previously published recommendations for writing a narrative review [5]. PubMed and Scopus databases were queried for studies that reported risk factors of clinical and structural progression for KOA to January 2019. The initial search started with the terms “knee osteoarthritis”, “risk factors” and “improvement”. After the identification of potentially modifiable risk factors, the search was extended using each modifiable factor as a keyword. MeSH terms and relevant free-text terms were used, accordingly. The full-search strategy of biomedical literature in PubMed database is presented in Supplementary Appendix 1. References of retrieved studies and relevant reviews were hand-searched for a further supplement. The final search was carried out on 7 Feb 2019. A priority was given for studies in the last 5 years, which included more than 100 participants, had a control group and the duration of the trial was at least 3 months. In our quest for presenting a wider range of evidence-based knowledge, relevant data from observational studies, systematic reviews, and meta-analyses were also included. In accordance with the Osteoarthritis Research Society International (OARSI) recommendations, greater emphasis was put on publications related to symptomatic, physical function, and disability improvement; however, also discussing modifiable risk factors as potential structure-modifying interventions [6,7,8]. In this review, only exposures considered to be positively or negatively associated with clinical improvement or structural progression of KOA are addressed as potentially modifiable. Studies reporting associations of KOA with age, race, gender, family history of genetic disease, biomarkers and anthropometric indices (such as body height, knee height, and arm span) were not included as, to the best of our knowledge, these exposures are not a subject to an intervention. Trials of drug exposures were also excluded. Current recommendations of the leading OA societies were identified based on the prior knowledge of the authors.

Risk factor categories and current recommendations

The identified modifiable exposures were grouped into six main categories, based on their nature, as follows: (1) obesity and overweight, (2) comorbidity, (3) occupational factors, (4) physical activity, (5) biomechanical factors, (6) dietary exposures. In comparison, most of the current recommendations for non-pharmacological treatment mainly emphasize weight reduction, promotion of physical activity, and use of assistive devices for modifying the biomechanics of the knee joint, e.g., braces and insoles (Table 1) [9,10,11,12,13,14]. Less attention is paid on comorbidities, occupational factors, and diet exposures, although their alteration might produce a significant size effect on KOA progression.

Table 1 Modifiable risk factors addressed in the current non-pharmacological treatment recommendations of the leading organizations
Full size table

Overweight and obesity

The majority of patients with KOA are overweight or obese [15]. The risk of developing OA is twice higher in overweight individuals compared to those with normal body mass index (BMI) (< 25) [odds ratio (OR) 1.98]. Obesity (BMI ≥ 30) further increases this risk (OR 2.66) [16]. The high body mass promotes development and progression of KOA via two possible mechanisms: mechanical stress beyond the physiological capabilities of the weight-bearing knee joint, on the one hand, and altered metabolic and humoral profile, resulting in elevated adipocytokine levels and associated pro-inflammatory response, on the other [17] (Fig. 1).

Fig. 1
figure 1

The vicious circle of the complex cause-and-effect relationship between knee osteoarthritis and obesity

Full size image

There is strong evidence that weight loss is an effective treatment modality for KOA, resulting in pain alleviation, improvement of physical function, mobility, and quality of life [18,19,20,21]. These benefits are independent of structural damage severity [22]. In terms of structure-modifying capability of weight loss, the literature is divergent, with some investigators reporting that intensive diet-induced weight loss was not beneficial for slowing down structural damage progression in comparison with exercise-alone program in patients with KOA who did not significantly reduce their weight [19, 23] with a lack of decrease in markers of cartilage breakdown [24]. In contrast to these findings, an observational trial found out that 7% weight loss was associated with reduced medial femoral cartilage thickness loss [25]. Weight reduction is already included as a core treatment in the management of patients with KOA in most of the current guidelines and recommendations [9,10,11,12,13,14]. Nevertheless, under real-world conditions weight loss is recommended by physicians in a suboptimal number of overweight or obese patients [26].

Dose- and time-dependent effect of weight reduction

For each kilogram of body weight lost, the knee experiences a fourfold reduction in load during daily physical activity [27]. In a recent study investigating the dose–response relationship between weight loss and symptomatic improvement, the group with the largest amount of weight loss (> 10%) showed the greatest improvement in clinical scores, independently of age, sex, baseline weight, and baseline physical function. On average, minimal clinical improvement was achieved when 7.7% of body weight was reduced [28]. It seems that the more drastic the weight loss is, the better the outcome may be: 25% less pain and better function experienced by patients who reduced more than 20% weight compared to patients with more than 10% of weight reduction [29].

Diet, exercise or a combination of both?

The way the weight loss is achieved may be of particular interest for investigators and clinicians, influencing the preservation of muscle mass, inflammatory response and long-term maintenance of body weight. Although fasting was found out efficient in a small uncontrolled German study where pain, mobility, and joint function were improved significantly after a significant reduction in weight [30], there may be more favorable weight reduction programs in the long term. Theoretically, based on the anti-inflammatory effect of Mediterranean-like diets rich in fibers, polyphenols, and omega-3 fatty acids [31], healthy diet-based weight loss programs may provide additional benefits to patients suffering from low-grade inflammatory conditions [32] such as KOA. On the other hand, this may be true for patients performing habitual physical activity for weight reduction [33]. A recent meta-analysis showed that diet-induced weight loss alone or combined with exercise in overweight or obese people results in improvement of physical function, but moderate pain relief is achievable only with a combination of diet and physical activity, and not with diet-only strategies [20]. In terms of weight loss effect on structural progression, Henriksen et al. observed an increased number of bone marrow lesions in KOA patients allocated to the physical activity group compared to the diet-only and no-attention groups, with no significant difference in cartilage loss, synovitis or effusion [34]. A diet creating an energy-intake deficit of 800–1000 kcal/day from a ration 15–20% from protein, less than 30% from fat, and 45–60% from carbohydrates showed a significant reduction in serum biomarkers for KOA [19, 35]. The importance of these findings still remains unclear. Overall, patient preferences appear to be detrimental for the prescription of weight loss regimen. Real-world data show that physical training programs alone or in combination may have slightly worse adherence compared to diet-only programs [36]; obese patients with KOA may not be induced to exercise as instructed [37].

Bariatric surgery should be considered in refractory cases of morbidly obese patients (BMI ≥ 40) [12]. Better symptomatic relief after the surgery is expected in younger patients and those without prior knee injury or other OA involvement [38].

Weight maintenance

The list of randomized controlled trials for the effect of weight maintenance on KOA symptoms after achieved weight reduction is short, consisting, to the best of our knowledge, of a single trial. According to Christensen et al., benefits of weight reduction remain if weight maintenance is promoted with physical exercise or diet over a 12-month period, irrespective of the maintenance program. Nevertheless, patients who underwent only dietary interventions regained less weight than did those allocated to the exercise-based program [37]. Furthermore, an increased number of bone marrow lesions were observed in the exercise-based group, although no greater cartilage loss has been reported [34]. A structured formula diet strategy in combination with dietary counseling was successful for weight and symptoms maintenance over the long term in KOA patients [39].

Comorbidity

Diabetes

Both in healthy and OA joint, metabolism plays a crucial role in the physiological turnover and remodeling of synovial joint tissues, including articular cartilage [40]. In the setting of persistent hyperglycemia, the formation of oxidants is promoted, resulting in increased matrix catabolism of the osteoarthritic cartilage [41]. Cartilage was shown to be softer in diabetics [42]; thus, more prone to damage. Although there is still little evidence to conclude with certainty that impaired glucose metabolism increased the risk of KOA, independent of obesity and advanced age [43], a recent study of a relatively large cohort of KOA patients from the Osteoarthritis Initiative found that diabetics exhibited a significantly greater increase in cartilage and meniscus lesions when compared to diabetes-free controls over a period of 4 years. The more severe the course of diabetes was, the greater the structural progression was observed [44]. Whether stringent glycemic control, in addition to cardioprotective, is also chondroprotective is yet to be determined.

Depression

Since every fifth patient with OA experiences depressive symptoms, a clearly higher prevalence of patients with OA suffered from depressive disorders compared to age-matched non-OA individuals [45]. Greater disability, mortality rate, and medical costs are associated with OA patients with concomitant depression [46, 47]. Furthermore, the persistence of depressive symptoms is significantly associated with worsening of knee pain in patients with KOA, although their cause-and-effect relationship is still in question [48]. Depression also does not significantly affect changes in radiographic disease severity over time [49]. Nevertheless, since depression is a modifiable exposure, its management in the routine clinical practice may extensively contribute to symptom alleviation in KOA patients. Decreasing the levels of depression is also beneficial for patients undergoing total knee arthroplasty [50] and should be a part of the complex disease management in the broader context.

Cardiovascular diseases

Both KOA and cardiovascular diseases are strongly associated with sedentary behavior and obesity. A dual association thus was previously suggested with comorbidity and KOA accelerating the progression of each other, resulting in a non-causal correlation [16].

Occupational factors

One in every seven cases of KOA is attributable to work [51]. Occupation is one of the strongest modifiable exposures for developing and subsequent progression of KOA. It exerts its effect via two possible mechanisms: first, force exertion, repetitive physical movements or demanding posture at work may result in non-physiological stress on the knee joint structures and increase the patellofemoral joint reaction, and secondly, specific occupational activities (mainly among professional athletes) may lead to higher incidence of knee and ACL injury. The latter is being confirmed in a study among current and retired professional footballers who are nearly twice as likely to suffer from KOA and the risk increases with a knee injury or surgery accumulation [52]. However, a history of knee trauma is hardly modifiable, as opposed to occupational activities that may be intervened via vocational rehabilitation. Professional activities that increase the risk for KOA are kneeling, squatting, lifting heavy weight and climbing, with OR ranging from 1.55 to 1.7 [53].

Vocational rehabilitation

Although to the best of our knowledge there are no cohort studies to support the effect of vocational rehabilitation on symptom alleviation, there is a wide range of interventions that could theoretically help relieve pain and improve physical function in patients with KOA and slowing down the progression of the disease. This includes alteration of occupational behavior, change of work tasks or working hours, use of assistive technology, workplace modification [12] and neuromuscular education.

Occupational activities requiring deep flexion such as kneeling and squatting have a dose-dependent effect on KOA risk contributing for 26% risk elevation per 5000 h increase of kneeling or squatting [53]; thus, the most simple and effective measure is to reduce the working hours of activities demanding for kneeling or squatting. Use of assistive technology may also help, since the duration of kneeling in the sand–cement-bound screed floor layers is shorter using electrical screed leveling machines (particularly manually moved) [54]. Incorporating proper kneepad and posture in a work environment could reduce the risk of knee injuries and favorably redistribute the load forces unloading the knee joint to a varying degree [55]. Education for lumbopelvic-thigh muscle co-contraction while squatting may improve functional performance and reduce patellofemoral joint crepitus, showed a single case report [56]. The use of a low assemble seat on wheels may also help reducing the time spent in a forced working posture by alternating kneeled or squatted work with seated work [57]. Occupational disease surveillance should be of utmost importance in the prevention of work-related KOA.

Physical activity

Supervised individualized exercise therapy is clinically and cost-effective; thus, it should be integrated as a central part in the treatment of every patient with KOA [58]. After adjusting for BMI, better physical performance was associated with a higher percentage of lean mass and a lower percentage of fat [59]. Patients should aim at being physically active at least 180 min per week for optimal pain reduction and functional improvement [60]. Nevertheless, up to 70% of patients with KOA do not achieve the recommended levels of physical activity, especially females; therefore, new strategies should be developed to encourage KOA patients’ engagement in exercise [61]. To optimize patient outcome, a tailored regimen should be prescribed using proper dosage, progression, and behavioral modulation techniques, accounting for individual differences and comorbid conditions [62].

Prejudice

Underutilization of physical activity interventions due to fear of disease acceleration is a common problem. Nevertheless, the notion of putting aside preconceived ideas will allow the clinician to trust high-quality evidence-based data that moderate exercise is safe and improves pain and knee function if trauma is avoided [63]. Loading the knee joint with exercise seems not to damage the articular cartilage [64]. Notwithstanding, patients with advanced radiographic disease (Kellgren–Lawrence grade 4) should be approached with caution since greater daily physical activity is associated with symptom aggravation [65]. For patients with BMI over 35, psychological support and lifestyle changes promote their participation in physical activities, but orthopedic consultations should preferably not be delayed [66].

Types of exercise

The remarkable heterogeneity of study protocols leads to poor reproducibility of the data, making a direct comparison of various exercise programs difficult [67]. A distinct and significant difference between aquatic and land-based exercise in terms of symptomatic improvement has not been established so far [68]. Nevertheless, aquatic activities with their high relative load and optimal speed were hypothesized to be superior for addressing power deficits while avoiding pain in elderly adults [69]. While thigh-muscle strengthening may alleviate pain and ameliorate function alone, adding electrical stimulation seems to improve the outcomes [70]. Patients with KOA may benefit additionally from a combination of resistance hip exercises with quadriceps strengthening [71].

A training strategy involving the use of cuffs for venous blood flow restriction during exercise may induce quadriceps muscle hypertrophy and improve strength in patients with weakness and atrophy related to knee pathology while appearing safe when properly performed [72]. Tai Chi and other physical exercise have a dose-dependent effect on the improvement of pain and physical function [73]. Backward walking with conventional physiotherapy treatment is effective in treating gait impairment [74]. А combination of diet- and exercise-induced weight loss is thought to be an optimal treatment for reducing pain and improving physical function in overweight adults with knee OA [75].

Motivation

Engaging and maintaining a physically active lifestyle is a challenge for people with KOA. Factors associated with a good outcome include having positive exercise experience, attitude and beliefs, proper education, adjusting and prioritizing physical activity, and presence of professional and social support. Some of the common barriers in the implementation of physical activity in daily life are the presence of pain and physical limitations, negative experience, beliefs and information, a resigned attitude, and lack of motivation and professional support. There is a complex interplay of individual, psychological and socio-environmental factors for the successful activity intervention; therefore, a personalized approach should be considered for every patient [76].

Exercise and physical activity promotion according to American College of Sports Medicine guidelines is effective for aerobic and resistance training in patients with OA [69, 77]. Healthcare providers should actively encourage patients to engage in muscle-strengthening and aerobic exercise as there is strong evidence that it can delay the onset and improve outcomes [78]. Sedentary behavior is strongly influenced by environmental and social factors [79]. Adherence, competence, and motivation should be assessed to identify patients in need of social support and exercise behavior modulation [80].

Biomechanical factors

In recent years, strong evidence has emerged that mechanical forces play an important role in the predisposition to complaints and the development of structural damage. Biomechanics of the knee joint may be modified by intervening intrinsic factors like muscle strength and lower extremity axis or using helping aids.

Thigh muscle strength

Acting as a shock-absorber and patellar stabilizer, quadriceps femoris muscle plays a major role in the knee biomechanics and its strengthening should also be an integral part of the treatment regimen, where possible, in patients with KOA [12]. The atrophy of quadriceps femoris is not only a risk factor but also a consequence due to the inactivity and disuse of the “diseased” limb [81]. The latter may explain the fact that concurrent loss in muscle strength is associated with symptomatic KOA progression [82], probably causing a vicious circle of cause and effect. Increasing the muscle mass of the quadriceps femoris improves pain and function in the KOA [83, 84], stabilizing the patella and the knee joint during gait [85]. A study investigating the effect of strength training on KOA progression shows that it reduces the progression of structural changes [86]. Strengthening exercises improve pain and physical function with no apparent superiority between different types of activities, at least in the short term [70].

Lower-extremity alignment

Varus and valgus malalignment increase the medial and lateral stress distribution of the tibiofemoral joint and are associated with pain aggravation and cartilage loss progression of the medial and lateral compartment, respectively [87, 88]. Varus deformity in combination with medial meniscal extrusion may be the reference to perform an advanced intervention to slow down KOA progression [89]. Shifting the axis with the greatest load away from the injured cartilage to the stronger compartment of the knee by slightly over-correcting to valgus may reduce pain, delay or completely postpone total knee arthroplasty [90, 91].

Orthoses and other assistive devices

Orthoses and assistive devices exert their curative effects mainly by affecting the pathomechanics of the osteoarthritic joint. Although often underestimated in clinical practice, they may be no less effective than the routinely used pharmacologic therapeutic modalities [84].

There are currently a large number of assistive devices including shoe insoles, braces, splints, canes, crutches, walkers that may potentially modify mechanical forces acting at the knee joint. Although some of the current recommendations for the treatment of osteoarthritis advocate the use of specific insoles [13, 14], resulting in a reduction of the knee adduction moment, a recent meta-analysis found no benefit in terms of pain and joint function over placebo (flat) insole [92]. Medial unloading (valgus) knee brace and medial knee taping may provide short-term improvement in pain when compared with sham [93, 94]. Assistive devices such as canes, crutches and walkers may minimize disability when individually tailored to patient’s height, weight, and joint alignment.

Dietary exposures

Information on the efficacy of dietary exposures in KOA patients is rather controversial documented in related trials with highly heterogeneous results. Moreover, most of the studies have a short-term follow-up. High-quality data mostly come from industry-sponsored trials and should be considered in light of the potential positive funding outcome bias. Overall analysis trials showed that dietary supplements provide moderate and clinically meaningful treatment effects on pain and function in patients in the short term, although the quality of evidence is low [95].

Plant-derived products

Fruits

Results from two longitudinal studies invariably indicated that greater fiber consumption was related to a lower risk of symptomatic KOA, while the relation to the incidence of radiographic KOA remained unclear [96]. Nevertheless, the observational nature of both studies raised concerns about residual confounding. In a small randomized controlled trial, Schell et al. observed that dietary strawberries had a significant analgesic effect in obese KOA patients with mild-to-moderate knee pain. 12-Week intake of dietary strawberries results in a reduction of total pain and disability. The authors have hypothesized both effects may be due to dietary polyphenols and other bioactive compounds found in strawberries [97]. Similarly, another recent study showed that pomegranate juice rich in antioxidants decreased significantly Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) total, stiffness, and physical function scores over a short-term exposure in a small randomized trial [98].

Herbal formulations

Supplementation with 500 mg of Momordica charantia (bitter melon) for 3 months resulted in reducing pain and improving symptoms among KOA patients. However, when compared to the placebo group, Momordica charantia differed significantly only in the analgesics use [99]. Similarly, a trend towards reduced analgesics consumption compared to placebo was observed with a special formulation composed of a mixture of rosehip (Rosa canina L.) puree/juice concentrate, nettle (Urtica dioica L.) leaf extract, devilʼs claw root extract, and vitamin D [100]. A low dose of the medicinal plant Artemisia annua (150 mg) showed an effect in a randomized control trial of 42 patients, improving pain and WOMAC total score from baseline to 12 weeks. Significant improvement of symptoms was not seen in the placebo group and in the group taking a higher dose of Artemisia annua. However, a direct comparison between Artemisia annua and placebo was not evaluated [101]. Consumption of spearmint tea with or without high rosmarinic-acid content for 16 weeks improved physical disability and stiffness, while the pain was alleviated only in the rosmarinic-acid group [102].

In a 12-week randomized double-blind, placebo-controlled, parallel-design trial, KOA symptoms in overweight or obese female patients taking 1000 mg garlic tablets were significantly improved for the study period. However, there was no difference between active and placebo groups [103].

Avocado/soybean unsaponifiables

Avocado/soybean unsaponifiables have been suggested to possess chondroprotective, anabolic, and anti-inflammatory characteristics. At the clinical level, they seem to reduce pain and stiffness while improving joint function [104, 105]. Although effective for symptomatic treatment more research seems warranted for structural progression inhibition and long-term effects [106].

Animal products

Fish and crustacean extracts

Placebo-controlled trials did not show any significant clinical effect of krill oil and oral salmon calcitonin over placebo [107]. Low- and high-dose fish oil intake improved significantly pain in a randomized non-placebo controlled trial [108]. Nevertheless, this improvement for pain was comparable to the one commonly observed with the ‘placebo effect’. Further studies are warranted before any conclusions could be drawn.

Collagen

Van Vijven et al. did not find sufficient evidence to recommend the generalized use of collagen hydrolysate in clinical practice [109]. A recent meta-analysis, however, showed a significant reduction in the stiffness subscore of the WOMAC index in patients with OA [110].

Glucosamine and chondroitin formulations

Chondroitin and glucosamine supplements are found to be safe, but results of their efficiency are inconsistent, varying among different formulations. Тheir place in the management of KOA and clinical effectiveness are a frequent topic of debate among researchers, resulting in high heterogeneity of recommendations of the leading organizations: from a first-line “pharmacological” therapy in European Society for Clinical and Economic Aspects of Osteoporosis and Osteoarthritis treatment algorithm [11] with potential structure-modifying effect according to the European League Against Rheumatism [111] to a conditional recommendation of the American College of Rheumatology that KOA patients “should not use the following: Chondroitin sulfate, Glucosamine” [10]. A highly cited meta-analysis by Wandel et al. showed no superiority of chondroitin sulfate and/or glucosamine (sulfate or hydrochloride) over placebo for reduction of pain and joint space narrowing [112]. Since its publication, however, numerous criticisms have been raised [11, 113] and in contrast to its results, favorable effects on symptoms were reported in other high-quality studies, especially for glucosamine sulfate in crystalline form [114, 115]. Conclusively, further research, including industry-independent trials, is warranted to assess the proper dosage, the most effective formulation, and the optimal duration of treatment and to identify specific subsets of patients in whom glucosamine and chondroitin, either alone or in combination, may improve prognosis.

Probiotics

Skimmed milk containing Lactobacillus casei Shirota (LcS) showed greater improvement in pain and function in 6 months than placebo. Serum levels of hs-CRP were found also significantly lower in patients consuming LcS. The authors of the study have hypothesized that probiotic consumption could serve as a novel therapeutic option in the clinical management of knee OA [116].

Minerals

Choline-stabilized orthosilicic acid (a bioavailable form of silicon) was tested in a multicenter, double-blind, placebo-controlled study with 211 patients with KOA for 12 weeks. In the total study population, no differences in clinical parameters were observed between the placebo and the active-treatment groups, but gender interaction was found, where men receiving choline-stabilized orthosilicic acid had a significantly greater improvement in WOMAC total score, stiffness and physical function as well as a lower increase in biomarker levels of cartilage degradation [117]. Lower magnesium intake is associated with worse pain and function in KOA patients [118]. Serum levels may also have an inverse association with radiographic OA, joint space narrowing and hsCRP levels [118, 119].

Vitamins

Vitamin D

A randomized controlled pilot trial of KOA patients with vitamin D insufficiency suggested there was a small but statistically significant clinical benefit of vitamin D on knee pain and knee function over placebo [120]. In addition, an observational study investigating the effect of vitamin D insufficiency on the clinical and structural outcome of the KOA found that the consistently sufficient group had significantly less loss of tibial cartilage volume, less increase in effusion-synovitis volume, and less loss of physical function compared with the consistently insufficient group [121]. However, in a multicenter randomized, double-blind, placebo-controlled vitamin D supplementation, compared with placebo, did not result in significant differences in change of MRI-measured tibial cartilage volume or knee pain over 2 years [122]. Another high-quality 3-year, double-blind, randomized, placebo-controlled trial comparing 800 IU cholecalciferol daily supplementation with placebo confirms the non-efficacy of vitamin D [123]. Therefore, it may be concluded that supplementing with vitamin D does not play a significant role in modifying the progression of KOA and the previously reported small positive effects may be due to residual confounding.

Vitamin K

An increased risk of radiographic knee osteoarthritis and cartilage loss over time was associated with subclinical deficiency of vitamin K [124]. Moreover, very low plasma vitamin K1 levels (< 0.2 nM) were related to cartilage and meniscus damage progression after 3 years of follow-up. Decreased plasma levels of dephospho-uncarboxylated matrix gla-protein, a marker for vitamin K deficiency, were associated with the presence of KOA symptoms but not with progression [125].

Conclusion

The present narrative synthesis of modifiable risk factors does not pretend to cover all aspects of the possible non-pharmacological interventions for KOA management, nor to create recommendations, as there are obviously enough of them. We have rather outlined the main categories of risk factors that could be intervened and pointed out their significant effect size in modifying the symptoms and the course of KOA. In fact, to the best of our knowledge, this is the first review to propose a detailed classification of the modifiable risk factors for KOA and to discuss most of them in terms of therapeutic interventions. In the era of age- and obesity-related diseases, the combined effects of local and systemic risk factors should be managed by combined measures—simple but effective individual interventions on health-altering exposures forming a comprehensive package of care. Theoretically, femoral muscle-strengthening activities, complemented with proper diet, weight loss, vocational rehabilitation, management of comorbidities (especially diabetes and depression), and biomechanical support may add up to the holistic treatment approach towards the patient with KOA. An individual risk factor modification program should be developed in accordance with patient preferences and habits, level of physical activity, workplace, medical history, and overall health condition. Unlike pharmacological treatment, non-pharmacological interventions could be easily combined without any concerns for drug interactions or cumulative side effects. Due to its great impact on a wide range of functions and tissues, risk factor modification seems to be getting really close to the optimal treatment of KOA, capable of improving the metabolic, biochemical, and biomechanical environment of the joint. We have already started thinking of osteoarthritis as a whole; now, it is time to treat it as a whole.