The primary purpose of establishing normal values of mJSW in a healthy population of males and females is to provide age-specific references to which osteoarthritic values can be compared. In addition, it is important to determine if mJSW values appear to decrease with age in a healthy population or if, indeed, this is characteristic of only those affected by knee OA. We also investigated the correlation between mJSW and medial tibial cartilage morphometry in this healthy population. Results of this pilot study appear to suggest that mJSW values are not significantly different between younger and older individuals without radiographic evidence of OA as shown by mean (SD) values of 4.8 (0.7) mm and 5.7 (0.8) mm in females and males, respectively. Individuals included in these analyses were those with K-L grades of 0 and 1 as was the case in a recent study by Conrozier et al. . While it may be argued that a K-L grade of 1 may correspond to early OA, the definition of this categorization states the doubtful presence of osteophyte without regard for joint space narrowing. This would support the notion that cartilage thickness measurements would not be affected by the inclusion of those with K-L grade 1. To verify this, additional analyses including only those with K-L grades of 0 (grade 1 excluded) were performed and results did not differ significantly from those which included both K-L grades 0 and 1.
Results from this study suggest that there is no identifiable decade of life when one might expect joint space width to narrow. When considered as a continuous variable, age was not found to be significantly related to mJSW in either males or females again supporting the notion that joint space narrowing may not simply be a consequence of aging. To confirm this, however, we recognize that a longitudinal study collecting data over decades would be required and therefore was not feasible at this point in time.
Despite the fact that there are few studies which are longitudinal in nature, there are a few cross-sectional studies which have investigated the relationship between JSW measurements and age with methodologies slightly different than the ones used in the present study. For example, a study of healthy young adults 16–22 years of age reported mean (SD) medial mJSW values of 4.74 (0.94) and 5.65 (0.93) in females and males respectively, results much like those of similar aged participants in this study . Also like our study, JSW values in 125 healthy individuals between 40 and 75 years of age were not found to decrease with increasing decade of life with mean values ranging from 4.6 – 5.0 mm in females and 5.0 – 5.5 mm in males . In both of these studies, males were generally found to have larger JSWs compared to females, results which are also consistent with those reported here [31, 34]. In contrast, studies conducted by Dacre et al. and Sargon et al. showed that JSW decreased with increasing age group, although both of these studies were cross sectional in nature and methodological differences existed including X-ray acquisition technique (weight bearing vs. non-weight bearing), joint space analysis (manual vs. automated, joint space area (mm2) vs. mJSW ) and the symptomatic nature of patients [32, 33]. While our results and those of other cross-sectional studies of healthy individuals suggest mJSW values remain constant, other results suggesting the opposite justify the need for large-scale, cross-sectional and longitudinal population-based data of healthy individuals acquired using the most reproducible techniques [22, 35].
Although there is a paucity of radiographic data from healthy individuals conducted over time, a study conducted by Conrozier et al. examined longitudinal changes in mJSW in individuals reporting chronic knee pain (>3 months) but lacking radiographic evidence of knee OA (K-L grade ≤ 1), as was considered in the present study. These authors reported a mean (SD) annual rate of joint space narrowing of 0.05 (0.22) mm . However, the symptomatic nature of the participants may be indicative of cartilage lesions that may not be radiographically detectable, as reported by Ding et al. in patients with K-L grade 1, thereby questioning the status of this sample as a "healthy" population . In addition, this study did not report whether this change was statistically significant from baseline to one-year follow-up. In fact, such small changes in joint space width are often within the range of reproducibility error of measurement .
Other studies of medial JSW values in healthy individuals have reported average values for the entire populations under investigation but have not analyzed these measurements as they varied with age or sex [32, 37]. Dacre et al., for instance, reported the mean medial JSW acquired from non-weight-bearing radiographs to be 5.73 (0.15) mm in females and 7.03 (0.12) mm in males, results which are 17% and 19% larger than those of the present study, respectively . However, joint space width values acquired from non weight-bearing X-rays may be larger than those acquired from weight-bearing ones, suggesting that these results may, indeed, be consistent with those of the present study .
Given that mJSW is a surrogate measure of cartilage thickness, one would hypothesize that these variables would be correlated with one another. However, it is widely understood that joint space width measurements reflect only a thickness measure at one specified location in the joint and may include tissues such as menisci and synovial fluid, findings that are supported by previously published studies [39–44]. Our results revealed that the variance in medial tibial cartilage thickness, normalized to bone area, can explain less than half of the variation in medial mJSW. It must be noted here, however, that medial femoral cartilage thickness was not analyzed in this study population. This variable would certainly also account for some of the variation in joint space, although cartilage thickness in one plate is not highly correlated with cartilage in another plate . Analyses previously conducted in an osteoarthritic population where both medial femoral and tibial cartilage were examined suggested the variation in cartilage thickness accounted for 54% of mJSW .
The issue of whether sex and age are significantly related to cartilage volume and thickness has been the subject of many studies [47–51]. Mixed results have been reported with respect to gender differences in cartilage volume and thickness, although our results revealed that males have significantly larger mean tibial cartilage volume and thickness just as was the case with mJSW. Similarly, studies by Faber et al. and Cicuttini et al. reported significantly larger mean cartilage volume values in healthy males compared to healthy females [49, 51–53]. While Cicuttini et al. also reported significantly larger medial tibial cartilage thickness values in males of the same population, although thickness was not assessed directly but calculated as volume per unit area, Faber et al. did not find such significant differences in thickness between genders [49, 51, 53]. For instance, in our study, men had 18% more medial tibial cartilage thickness compared to women while Faber demonstrated that men had 13.3% thicker cartilage than women, although this difference was not statistically significant. Discrepancies in results from these studies likely exist because of differences in sample populations (i.e. age, definition of "healthy") and the relatively small sample size.
In the current study, in contrast to results investigating age as a categorical variable, increasing age (as a continuous variable) was found to be associated with less cartilage volume normalized to total bone area (β = -0.41) and thickness (β = -0.37) in females after adjusting for BMI. Such age-related differences in medial tibial cartilage volume and thickness were not observed in males. While this may be related to the relatively small number of healthy males over 50 years of age in our study sample, it is also possible that inconsistencies between males and females may be related to hormonal changes which occur during menopause of which there are no comparable changes that occur in men. This is similar to the BMD findings in osteoporosis [54, 55]. Other cross-sectional studies which have investigated the relationship between age and cartilage volume and thickness have shown inconsistent results with one reporting a significant decrease in medial cartilage thickness, but not volume, with age , while another reports no significant changes in tibial cartilage thickness with age . However, one should be cautious about the interpretation of these results since these data are cross-sectional in nature and do not reflect changes in a single person over time but comparisons between different individuals.
Three studies have reported longitudinal changes in cartilage volume for healthy individuals and have shown that cartilage volume does, indeed, decrease with aging [54, 57, 58]. In healthy males (N = 28, mean age 52 years), the mean annual reduction in tibial cartilage volume was found to be 2.8% (95% CI = 0.2% to 5.5%) . In healthy postmenopausal females, the average annual decrease in total tibial cartilage volume was similar at 2.4% (3.2%) . What is notable in these two studies is the mean age of subjects being investigated was over 50 years. To this point, there is only one study investigating longitudinal changes in a population including younger adults. Ding et al. demonstrated a significant association between age and loss of cartilage volume by approximately 1.5 – 4.2% per annum in individuals between the ages of 26 and 60 years, with a higher rate of loss in females as compared to males . However, despite these seemingly age-related declines, it is still plausible that these values lie within what may be considered to be a "normal" or "healthy" range.
It is important to recognize that there are a number of methodological limitations to this study including the small sample sizes, particularly in some age groups, the cross-sectional nature of the data and the lack of medial femoral cartilage analyses. Despite these limitations, results suggest that mJSW values do not decrease with increasing age group in males or females between the ages of 20 and 69 years. This information may be helpful in defining radiographic joint space width references for comparison with those suspected of having knee OA. These results also suggest that there is no defined decade at which point joint space width decreases. Cartilage volume and thickness did not decrease with increasing age in males as was the case with mJSW. However, the observation that cartilage volume and thickness decreased with ageing in females may support the role of estrogen in cartilage physiology, although the exact mechanism remains unknown. It is also possible that tissue other than medial tibial cartilage may play a more significant role in joint space narrowing than in males, although this has not yet been shown.