Rheumatoid arthritis (RA) is an autoimmune disease that is chronic and generalised, based on the inflammation and hyperplasia of the synovial membrane. It naturally progresses towards structural disruption, which in turn leads to musculoskeletal deformity, disability and a reduction in life expectancy, all of which are correlated with the level of inflammation.1

In the language of clinical nutrition, the expression “body composition” is associated with the evaluation of different components within the body, and it may refer to different levels such as the anatomical, molecular, cellular and tissue levels, or the total body.2,3 The determination of body composition at tissue level by using anthropometric techniques, such as the body mass index (BMI), skin fold measurements, waist or hip circumference and bio-electrical impedance analysis (BEI), makes it possible to gauge nutritional status as well as the percentage of body fat, energy reserve and the risk of diseases associated with alterations of these parameters.4,5

An increasing amount of scientific evidence shows that alterations in bodily composition characterised by an increase in the mass of fat, as well as heightening the probability of cardiovascular, metabolic and neoplastic diseases, may also play a role in modulating certain inflammatory diseases.6 In states of hyperadiposity, the adipocytes display hypertrophy and a status of activation characterised by the liberation of increased amounts of proinflammatory soluble mediators (adipocynes), such as leptin and resistin, giving rise to the proliferation and differentiation of mononuclear phagocytes system cells and the activation of the natural killer cells. Moreover, activated adipocytes in states of hyperadiposity have been shown to be a source of other proinflammatory cytokines such as chemerin, the retinol-4 transporter protein, lipocalin and, importantly for RA, TNFα, IL-6 and IL-12.7–9 Given that IL-1 as well as TNF are fundamental for the creation of chronic inflammation and locally destructive synovial proliferation, which are the aetiopathogenic basis of RA, study of the modulation of the inflammatory activity of RA by the level of body fat is a potentially interesting aspect for the study and clinical treatment of patients with RA. Although this subject has been covered by several previous studies, their results in this respect are contradictory.10,11 This may be explained by design weakness, the different evaluation methods used—for body composition as well as the level of the inflammatory activity of RA—or low statistical power. These methodological aspects may be detected in the majority of the reports on study of this subject.12

The above arguments clearly show that there is a relationship between alterations in bodily composition characterised by an increase in the amount of fat, such as being overweight or obese, and the level of inflammatory activity in patients with RA. This relationship has yet to be explained scientifically, so the aim of this study was to evaluate the degree of association between the physical composition and clinical status of RA patients. We used several anthropometric methodologies for this purpose, characterising body type and evaluating the clinical status of RA by measuring the degree of inflammatory activity, the level of physical functioning and the type of treatment required.

A total of 123 participants were selected (121 women and 2 men). The clinical and demographic characteristics of the participants are described in Table 1, and they all correspond to patients with RA treated in a tertiary hospital. The ESR-DAS 28 score could only be evaluated in 85 patients, as due to logistical reasons in 38 patients it was impossible to obtain the ESR values. As it was possible to measure the other individual parts of the DAS 28 in these 38 subjects (self-evaluation of their clinical state and number of painful and inflamed joints), the sample size used in the comparisons involving these individual components of the DAS 28 score amounted to 123. The results describing all of the anthropometric measurements are shown in Table 2.

Body composition type classification according to BMI scores are shown in Table 3. According to the cut-off points of the BMI-WHO, 76.4% of the patients were classified as having hyperadiposity: 30.9% of them were classified as overweight and 45.5% as obese. When the BMI classification of body composition was re-evaluated using Stavropoulos-Kalinoglou cut-off points, the prevalence of patients classified as having a normal BMI fell from 22.0% to 8.1%, giving rise to a redistribution of subjects categorised as overweight. Nevertheless, the net prevalence of overweight subjects did not change, because a high proportion of subject previously classified as overweight according to WHO were reclassified as obese using the Stavropoulos-Kalinoglou cut-off points. The net result of this reclassification was a marked increase in the prevalence of obesity (58.5%).

A tendency towards higher scores for the individual components of the DAS 28 score—such as self-evaluation by the patients and their number of inflamed and painful joints—as well as higher HAQ-Di scores and a higher proportion of patients categorised as obese who required biological therapies—based on WHO criteria or those proposed by Stavropoulos-Kalinoglou et al.—in comparison with subjects who are overweight or categorised as normal or below normal weight. However, none of these comparisons attained statistical significance (data not shown).

Based on the hypothesis that this lack of statistical significance could be due to a type ii error arising from the small size of the subsamples in subjects classified as normal or below normal weight, a post hoc analysis was performed which distributed the categories of body composition in a dichotomy: one category was composed of subjects classified as below normal and normal weight, while the other grouped the subjects classified as overweight and obese. Using this approach, we firstly found no statistically significant association between the categories of level of inflammatory activity evaluated by DAS 28 score and the type of body composition. This was the case for the WHO recommended BMI cut-off points as well as for those used by Stavropoulos-Kalinoglou et al. (Table 4).

On the other hand, Table 5 shows a dichotomised comparison between type of body composition according to BMI values (below normal/normal weight and overweight/obesity) and the values of the clinical variables studied. This time the subcomponents of the overall DAS 28 scores were compared as well as the overall scores. Although no difference was found when using the cut-off points recommended by the WHO, a notable finding is that when the cut-off points recommended by Stravropoulos-Kalinouglu et al. were used, the overweight/obese patients were found to show statistically significant higher scores for inflamed joints as well as a clear tendency towards a higher number of painful joints (at the limit for statistical significance), in comparison with patients classified as being of below normal or normal weight.

Table 6 shows the correlations between the different anthropometric measurements, the degree of inflammatory activity, the level of physical function and the required treatment. Notably, although the overall DAS 28 score shows no significant correlation with any of the methods studied for evaluating body composition, a significant positive correlation was found between the number of inflamed joints and 6 of the 11 anthropometric methods studied for body composition: BMI, hip and arm circumference, body fat measured by BEI and skin folds and triceps skin fold. The number of painful joints only showed a positive correlation with triceps skin fold and the percentage of body fat measured by plicometry. The daily dose of PDN required was negatively correlated with each anthropometric method, although it only attained statistical significance with average arm circumference. Finally, the HAQ-Di score correlated negatively with plicometric evaluations, although this was only statistically significant for the triceps fold.

The results of this work show that using the BMI cut-off points proposed by Stavropoulos-Kalinoglou et al., RA patients classified as overweight or obese have a significantly higher number of inflamed joints. They also show a clear tendency towards having more painful joints than is the case for patients classified as being of normal or below normal weight. On the other hand, a significant positive correlation was found between the number of inflamed joints and the majority of the anthropometric methods used to evaluate body fat volume. Both results support the hypothesis that, in patients with RA, being overweight or obese is associated with an increase in inflammatory joint activity.

The effect of body composition type on patients’ clinical state and more specifically the degree of inflammatory activity in RA has been examined in several previous studies,10–12,17–19 although their conclusions are contradictory. Some reports found a significant association between increased inflammatory activity and the presence of adiposity. Nevertheless, other reports found no proven association between the percentage of fat in body mass and the level of inflammatory activity in RA.

Some factors may explain the above-mentioned discrepancies. Firstly, the sample size and statistical power of the reports in question seem to be the most common cause of the inconsistencies. In some reports based on large samples a significant correlation was found, while those with smaller samples found no association between the degree of inflammatory activity in RA and type of body composition. Moreover, although body composition was evaluated using the BMI in all of the papers that were reviewed, those which classified body composition used the WHO cut-off points. Rheumatoid cachexia is highly frequent in patients with RA, and it is characterised by inflammation or physical inactivity with loss of muscle mass and replacement of this by fatty tissue, with relative conservation of total body weight. Based on this concept Stavropoulos-Kalinoglou et al. offered evidence that using BMI cut-off points for the general population to classify body composition in subjects with RA may be an erroneous indicator of body fat volume in these patients, as it may under-estimate it. These authors suggest that to use BMI as an indicator of body composition in patients with RA, BMI scores over 23 should be classified as overweight, while obesity corresponds to scores over 28. The use of this classification for body composition leads to a net increase in the number of patients with RA within the limits of obesity.

The majority of studies that have covered the relationship between degree of inflammation in RA and body composition use the DAS 2820 index to measure the inflammatory status of RA. DAS 28 is a viable, trustworthy and valid instrument for evaluating the level of inflammation in patients with RA. It is universally accepted as a valuable clinical and research tool for this purpose. However, the results of our study give rise to certain concerns regarding the validity of the DAS 28 index when evaluating the inflammatory status of subjects with RA in the presence of excessive fatty tissue. This observation is supported by the finding that, regardless of the classification criteria used (WHO cut-off points or the ones used by Stavropoulos-Kalinoglou), in our patients the overall DAS 28 score and ESR values were not associated at all with type of body composition. These results differ widely from the statistical association we found between other individual components of the DAS 28, such as the number of inflamed joints and being overweight or obese according to the Stavropoulos-Kalinoglou et al. cut-off points. Due to all of the above points, it is encouraging to state that the acute phase response in patients with RA is reduced in some way when the percentage of body fat exceeds certain limits. This opens up a line of research open on the effects of fatty tissue and the acute response in patients with RA. If this is proven, calculating the DAS 28 should be adjusted depending on body composition type.

We consider that our study has certain methodological advantages in comparison with previous reports on research into the relationship between level of inflammatory activity and type of body composition in RA: we used 2 classification methods for body composition (WHO and Stavropoulos-Kalinoglou); we also include the individual components of DAS 28 as well as its overall score in the statistical analysis. We also examine the correlation between levels of inflammatory activity and body composition scores by using 6 other anthropometric methods apart from the BMI.

On the other hand, we admit that the transversal nature of our study design is a major limitation that hinders proving any causal relationship between the associations found. The difficulty of recording the DAS 28 score in all of the patients included also reduced the statistical power of our study. Together with these considerations, the fact that our study was composed of a majority of female subjects is an analytical distortion. The fact is that the majority of patients who visit the Rheumatology Department and who agreed to take part in the study are women. Lastly, we admit that the statistical significances reported here on the association between the number of inflamed or painful joints and body composition evaluated by the BMI were obtained by post hoc analysis. This constitutes another significant restriction on internal validity for the results shown in this study.

We find that being overweight or obese according to the criteria of Stavropoulos-Kalinoglou et al. in patients with RA is associated with a higher number of inflamed joints. It is also significantly correlated with the number of inflamed joints and the score of the majority of anthropometric body composition measurements, such as BMI, arm and waist circumference, triceps fold and the percentage of body fat evaluated by BEI and plicometry. As a whole, these results support the hypothesis that the level of inflammatory activity is modulated by the volume of body fat in patients with RA. Finally, we believe our results support the need for clinical trials to evaluate the effect of body composition correction or optimisation on the clinical condition of patients with RA. If this intervention is proven to have an effect, monitoring nutritional status may come to occupy a central role in the basic recommendations for the management of RA.

The resources used to carry out this study originate in the institution where it took place.

The authors have no conflict of interests to declare.