Distribution of Finger Nodes and Their Association With Underlying Radiographic Features of Osteoarthritis

Arthritis Care & Research Vol. 64, No. 4, April 2012, pp 533 538 DOI 10.1002/acr.21586 2012, American College of Rheumatology ORIGINAL ARTICLE Distribution of Finger Nodes and Their Association With Underlying Radiographic Features of Osteoarthritis FRANCES REES, 1 SALLY DOHERTY, 1 MICHELLE HUI, 1 ROSE MACIEWICZ, 2 KENNETH MUIR, 3 WEIYA ZHANG, 1 AND MICHAEL DOHERTY 1 Objective. To determine the distribution of clinically palpable hand interphalangeal (IP) nodes at each finger and thumb joint in a population with nodes, the influence of left or right hand dominance and sex on the development of nodes, and the association between nodes and underlying radiographic features of osteoarthritis (OA). Methods. We performed a cross-sectional analysis of participants in the Genetics of Osteoarthritis and Lifestyle (GOAL) study who had >1 Heberden s nodes or Bouchard s nodes on clinical examination. Frequencies (%) of nodes were described for each IP joint in the hand. Associations between nodes and underlying radiographic OA were shown with odds ratios (ORs) and 95% confidence intervals. A logistic regression model was used to adjust for the following confounding factors: age, sex, body mass index, left or right hand dominance, hand trauma, occupation with heavy manual activity, and participation in sports. Results. Of the 3,170 GOAL participants, 1,939 had >1 nodes (mean age 68 years, 54% women). The distal IP joints of the index finger were the most frequently affected, followed by the thumb IP joint. Nodes were more common in dominant hands and women. There was a significant association between nodes and underlying radiographic OA (OR range 2.26 21.23). This association was stronger for joint space narrowing than for osteophytes. A dose-response relationship was found between clinical severity of Heberden s nodes and underlying radiographic change. Conclusion. Our study supports the positive association between nodes and radiographic OA, especially narrowing, and the influence of sex and left or right hand dominance on development of nodes. In this age group, presence of nodes may be taken as an indication of underlying small joint OA. INTRODUCTION Hand interphalangeal (IP) joint osteoarthritis (OA) is a common condition that can cause pain and functional impairment (1). Hand OA can be diagnosed radiographically by presence of joint space narrowing (JSN), osteophytes, subchondral sclerosis, and subchondral cysts, or clinically by symptoms and physical signs (2). Clinically, AstraZeneca, Macclesfield, UK provided financial support to establish the original Genetics of Osteoarthritis and Lifestyle study. 1 Frances Rees, BMBS, Sally Doherty, RGN, Michelle Hui, MBChB, Weiya Zhang, PhD, Michael Doherty, MD: University of Nottingham, Nottingham, UK; 2 Rose Maciewicz, PhD: AstraZeneca, Mölndal, Sweden; 3 Kenneth Muir, PhD: University of Warwick, Coventry, UK. Dr. Maciewicz owns stock or stock options in AstraZeneca. Address correspondence to Frances Rees, BMBS, Academic Rheumatology, Clinical Sciences Building, City Hospital, Nottingham NG5 1PB, UK. E-mail: frees@doctors. org.uk. Submitted for publication August 25, 2011; accepted in revised form December 9, 2011. hand OA is often accompanied by the presence of Heberden s nodes and Bouchard s nodes (3). Multiple nodes are regarded widely as a marker for generalized OA (4). Few studies have investigated the association between hand nodes and underlying structure changes (5 8). In the studies that have investigated this, the measures used to assess the association were inconsistent, which led to a different clinical interpretation of the association. For example, Caspi et al (5) used a composite radiographic score (JSN, osteophytes, cysts, sclerosis, and subluxation) and reported a good correlation with clinical score (nodes and lateral deviation). However, Cicuttini et al (6) reported poor agreement between Heberden s nodes and radiographic distal IP (DIP) joint osteophytes, although after examining the statistics there was a small but statistically significant relationship. Thaper et al (7) reported a positive association with both JSN and osteophytes and suggest they can be used as a clinical marker. Hart et al (8) examined Kellgren/Lawrence (K/L) scores and reported a high specificity but low sensitivity association and concluded that they were not able to replace radiographic definition with clinical nodes. 533

534 Rees et al Significance & Innovations There was a strong positive association between clinical nodes and radiographic osteoarthritis (OA). Therefore, in this age group, the presence of nodes would appear to be a good surrogate marker for underlying radiographic hand OA, a finding that may be of relevance in epidemiologic and genetic studies of OA where hand radiographs are unavailable. Nodes were more common in the dominant hands and women. A dose-response relationship was found between clinical severity of Heberden s nodes and underlying radiographic change (osteophyte and joint space narrowing). Nodes were most common at the index distal interphalangeal and thumb interphalangeal joints, which are 2 adjacent joints used in fine precision and pincer grip. The overall distribution of nodes suggests that biomechanical factors are important in their development. Hand OA is observed to occur more frequently in women and with increasing age. Other reported risk factors for hand OA include positive family history, obesity, occupation, and prior injury (9). There is conflicting evidence as to whether dominance of the left or right hand affects development of hand OA (5,10,11). The aims of this study were to determine the distribution of palpable nodes at each joint in a population with nodes, the influence of left or right hand dominance and sex on the development of nodes, and whether palpable nodes are associated with underlying radiographic features of OA. The latter issue is of relevance particularly in epidemiologic and genetic studies of OA where hand radiographs are not always available. PATIENTS AND METHODS Participants. Participants were taken from the Nottingham Genetics of Osteoarthritis and Lifestyle (GOAL) study, which has been described previously (12). In summary, the GOAL study was a case control study where the participants were recruited from a hospital-based population in Nottingham, UK from 2002 2006. Cases were patients with symptoms and radiographic changes of OA at the hip or knee sufficiently clinically severe enough to warrant hospital referral for consideration of joint replacement (n 2,049). Controls were people who had undergone intravenous urography at the same Nottingham hospitals and who had no symptoms, signs, or radiographic evidence of hip or knee OA (n 1,123). One patient subsequently withdrew her consent and 1 patient had withdrawn because it became apparent that she had secondary rather than primary hip OA. Ethical approval was obtained from the Nottingham Research Ethics Committee. Clinical assessment. Evidence of palpable nodes was determined by 5 specifically trained metrologists. Over the DIP and proximal IP (PIP) joints of all fingers, Heberden s nodes (DIP joints) were graded as the following: no nodes, radial node, ulnar node, both radial and ulnar nodes, or a single dorsal transverse bar. Bouchard s nodes (PIP joints) were graded as either present or absent. The nodes affecting the thumb IP and metacarpophalangeal (MCP) joints were also documented. This method has been previously validated (7,13). Intraobserver reproducibility was determined by the metrologists reassessment every 2 3 months. Age, sex, left or right hand dominance, previous hand trauma, occupation, and activity history were assessed by the GOAL study questionnaire. Body mass index (BMI) was measured at the interview (12). Radiographic assessment. Hand radiographs were dorsal palmar centered on the middle MCP joint. JSN and osteophytes were graded 0 3 using the Osteoarthritis Research Society International atlas (14) by a single observer blinded to the clinical scores. K/L scores 0 4 were also applied (15). Reproducibility has previously been validated (7). Eligibility for the present study. Patients were eligible to be included in this analysis if they had 1 nodes, irrespective of joint site, in either hand. This was regardless of case or control status. Statistical analysis. Distribution of nodes was measured using frequency of nodes at each finger joint. Unlike the prevalence measure that is the percentage of nodes in the total population (with or without nodes), this is a proportional measure that describes the node distribution in people with nodes. The association analysis was person based and undertaken separately for each joint; therefore, there was no need to adjust for within-participant effect. Associations between nodes and radiographic OA were shown with odds ratio (OR) and 95% confidence interval (95% CI). A logistic regression model was used to adjust for confounding factors, including age, sex, BMI, left or right hand dominance, previous hand trauma, occupations with heavy manual activity, and regular activities undertaken for 7 hours per week that may predispose to hand injury (e.g., rugby, cricket). Associations between left or right hand dominance and nodes were determined by comparing the OR (95% CI) of any node occurring on the right hand with right hand dominance and the OR (95% CI) of any node occurring on the left hand with left hand dominance. SPSS, version 14 was used for all analyses. Statistical significance was taken at P values less than 0.05 (2-sided). The reference group used to calculate the OR was the participants in the GOAL study who did not have a palpable node at the joint under investigation. RESULTS Population characteristics. From a total of 3,170 participants in the GOAL study database, 1,939 had 1 palpable

Finger Nodes in OA 535 Table 1. Population characteristics* Total population (n 1,939) Sex, no. (%) Women 1,047 (54) Men 892 (46) Age, mean SD (range) years 68.1 7.1 (45 86) BMI, mean SD kg/m 2 29.2 5.1 Dominant hand, no. (%) Right 1,758 (90.7) Left 145 (7.5) Both 36 (1.9) * BMI body mass index. nodes. Of these 1,939 participants, 1,400 participants (72.2%) were from the original case group and 539 participants (27.8%) were from the control group. The population characteristics are shown in Table 1. Clinical findings. The frequency of a node at each joint site in these 1,939 patients is shown in Figure 1. The index finger DIP joint was most commonly affected, followed by the thumb IP joint. On every finger, the DIP joints were more frequently affected than the PIP joints. This was also true for the thumb. The ring finger was least affected by nodes. The most frequently affected PIP joint was the middle finger. The right hand was more frequently affected than the left hand at all sites. When stratified by left or right hand dominance, right-handed people had more nodes on the right hand (OR 2.9, 95% CI 1.8 4.4; P 0.001), whereas left-handed people had more nodes on the left hand (OR 3.2, 95% CI 1.7 6.0; P 0.001). When stratified by sex, all nodes were more frequent in women than men, apart from the thumb MCP joints (Figure 2). With respect to grading of Heberden s nodes (Table 2), the most frequent finding was both, meaning a palpable node on both the radial and ulnar side of the joint. At the index and little fingers ulnar nodes were more frequent than radial nodes. The opposite was found at the middle and ring fingers. Association with underlying radiographic changes. The association between palpable nodes and underlying radiographic change (K/L grade 2) is shown in Figure 3. ORs and 95% CIs were adjusted for age, sex, BMI, left or right hand dominance, previous hand trauma, occupation with heavy manual activity, and regular activity in sports. At every joint there was a significant association between palpable node and underlying radiographic change (P 0.001). The association remained significant for individual radiographic features of osteophytes and JSN (P 0.001). The ORs were greater for JSN than osteophytes at every finger IP joint, except for the thumb. A dose-response relationship was observed for the association between Heberden s node severity and underlying radiographic OA change. Figure 4 shows that the OR of osteophytes increased gradually with severity of palpable node at the index finger DIP joint of the right hand. The same results were observed for all other DIP joints and for JSN. DISCUSSION This study demonstrates that nodes most frequently affect the index finger DIP joint, with the thumb IP joint the second most affected site. At every finger, DIP joint nodes were more frequent than PIP joint nodes and the ring finger was relatively spared from both nodes. The thumb IP joints were more frequently affected than the thumb MCP joints. Left or right hand dominance and female sex were significantly associated with the presence of nodes. Our findings demonstrate a significant association between palpable node and underlying radiographic OA, the association being stronger for JSN than for osteophytes. A dose-response relationship was found between clinical severity of Heberden s node and underlying radiographic change. Therefore, in this age group the presence of nodes would appear Figure 1. The distribution of nodes in people with any interphalangeal nodes (n 1,939).

536 Rees et al Figure 2. The frequency of nodes stratified by sex. to be a good surrogate marker for underlying radiographic hand OA, a finding that may be of relevance in epidemiologic and genetic studies of OA where hand radiographs are unavailable. Although a number of studies have examined the frequency distribution of symptoms (1,16) and radiographic changes (10,17,18) of OA at each joint in the hand, no studies have previously focused on the frequency distribution of nodes. The asymmetry related to left or right hand dominance suggests that differing hand functions and biomechanical stress may contribute to node development. Two adjacent distal joints, the index finger DIP joint and the thumb IP joint, were the 2 most affected sites. Both of these joints are required for fine precision pinch (index finger tip to thumb tip) and pincer grip (side to side opposition of the distal index finger to the distal thumb), 2 gripping actions involved in many daily activities. It may be that frequent repetitive gripping between the ends of the index finger and thumb particularly predisposes these 2 joints to microtrauma and subsequent OA. Such a biomechanical explanation has been suggested to explain the difference in prevalence and distribution of radiographic hand OA between dentists (an occupation with physically demanding hand usage) and teachers (an occupation with less physically demanding hand usage) (10) and the different patterns of hand OA observed in 3 groups of female workers who undertook repetitive but contrasting manual activities in an industrial setting (19). The middle finger is Table 2. Frequencies of grade of palpable node at each distal interphalangeal joint (n 1,854)* Index Middle Ring Little Right Left Right Left Right Left Right Left Radial 7.0 5.0 7.3 6.3 5.1 3.6 4.9 4.1 Ulnar 17.2 15.2 6.1 4.1 1.6 1.5 8.1 7.0 Both 21.7 17.5 10.9 9.0 5.8 5.3 12.1 10.8 Bar 18.2 11.3 7.7 6.0 2.9 2.4 10.6 7.9 * Values are the percentage. At the index and little fingers ulnar nodes were more frequent than radial nodes. The opposite was found at the middle and ring fingers.

Finger Nodes in OA 537 Figure 3. The odds ratio (95% confidence interval) of underlying radiographic changes (Kellgren/Lawrence grade 2) associated with any nodes (adjusted for age, sex, body mass index, dominant hand, hand trauma, occupation with heavy manual activity, and participation in sports). used to a lesser extent in precision grips so it might be less exposed than the index finger to repetitive microtrauma, although being the longest finger, it may be more exposed to external trauma. We hypothesize that the sparing of the ring finger IP joints may be due to its noninvolvement in most precision grips and its position between the middle and little fingers, which might confer protection from external trauma. No previous studies examining this were found. MCP joints, and even IP joints of the thumb, are often not considered to be a site for nodes. Developmentally, the thumb IP joint is a DIP joint, the thumb MCP joint is in fact a PIP joint, and the first carpometacarpal joint is the true MCP joint for the thumb; this whole ray of joints moved proximally in our past evolution in parallel with improved precision gripping and manual dexterity (20). In this study, the thumb MCP joints were affected by firm posterolateral and dorsal swelling with a frequency similar to that of the finger PIP nodes. The thumb MCP joints had the lowest OR when associated with underlying radiographic change, but this was still significant. The lower OR in the thumbs may be related to the fact that the hand radiographs were dorsal palmar centered and consequently the thumb was placed at almost 90 degrees to the other joints, which may have affected the ability to equally assess radiographic OA. This may also explain the stronger association with JSN than osteophytes in the fingers, since JSN in any plane will be captured on an anteroposterior radiograph; however, osteophytes that project posteriorly or anteriorly will not be identified. Female sex has previously been found to be associated Figure 4. The association between grade of node and underlying osteophytes at the right index distal interphalangeal joint. OR odds ratio; 95% CI 95% confidence interval. with hand OA (21). The mechanism for this is unclear, but may be due more to genetic or hormonal influences rather than environmental factors. Few studies have examined the association between radiographic and clinical OA. Our findings are in agreement with Caspi et al (5), who found a positive correlation between composite radiographic and clinical OA scores, and in addition found a correlation with right or left hand dominance. Caspi et al examined an elderly Jewish cohort (mean age 78.8 years) admitted to secondary care for a nonrheumatic reason. Although they found women had higher composite OA scores than men, this was not statistically significant. Cicuttini et al (6) found a small but statistically significant correlation between Heberden s nodes and DIP osteophytes ( correlation coefficient 0.36 [95% CI 0.33 0.39]). This was a female twin study in a general (non-nodal) population with a lower mean age (56.4 years) than in our study. Cicuttini et al did not examine JSN. One previous study from our unit by Thaper et al (7) also found a positive relationship between nodes and underlying radiographic change in a nodal OA population. In that study, however, osteophytes had a stronger association than JSN, the opposite finding of that of the present study. However, this previous study differed from the present study in that joints were not analyzed individually due to the limited sample size (n 489), but the ORs for all Heberden s nodes and Bouchard s nodes were pooled and were smaller. This difference in results may have been due in part to our larger study population, with a more equal sex balance (82.6% women in Thaper et al compared with 54% women in our study). Mean age was similar at 65.8 years. The difference may also be because of the sampling variation or different population at risk. Thaper et al examined siblings of people with OA, whereas the present study examined unrelated people with any hand nodes. Whether hand nodes related more to JSN or osteophytes requires further investigation, ideally a prospective study of people at risk of hand OA with a sample randomly selected from the population. There are a number of limitations to this study. First, the GOAL study was established as a case control study to examine genetic and lifestyle differences between patients with large joint OA and controls. We then examined a

538 Rees et al subset of this population with clinical evidence of hand OA. The GOAL study population is biased toward individuals with severe OA, so our findings cannot be extrapolated to estimate point prevalence of OA at each joint in the general population. These results would need to be confirmed with a further study. Second, this study only looked at one point in time. A prospective study would be required to examine the temporal relationship between development of clinical nodes and radiographic change. Finally, it is important to emphasize that although the ORs for the association between clinical and radiographic changes were highly significant, the presence of a node does not necessarily guarantee radiographic change will be present. Likewise, the absence of a node does not exclude the presence of radiographic OA. A prospective study using sensitive imaging to assess the development of nodes and radiographic changes over time is required. In conclusion, our study adds further evidence to the strong positive association between nodes and radiographic OA, especially JSN, and that sex and left or right hand dominance are significantly associated with palpable nodes. However, a prospective study is required to confirm the temporal association between nodes and radiographic change. AUTHOR CONTRIBUTIONS All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Rees had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study conception and design. Rees, Hui, Muir, Zhang, Michael Doherty. Acquisition of data. Sally Doherty, Zhang, Michael Doherty. Analysis and interpretation of data. Rees, Maciewicz, Muir, Zhang, Michael Doherty. ROLE OF THE STUDY SPONSOR AstraZeneca, Macclesfield, UK provided financial support to establish the original Genetics of Osteoarthritis and Lifestyle study as a collaborative project with the University of Nottingham, but had no involvement in the study design, data analysis, or writing of this manuscript. Publication of this study was not contingent on approval from AstraZeneca. REFERENCES 1. Zhang Y, Niu J, Kelly-Hayes M, Chaisson CE, Aliabadi P, Felson DT. Prevalence of symptomatic hand osteoarthritis and its impact on functional status among the elderly: the Framingham Study. Am J Epidemiol 2002;156:1021 7. 2. Zhang W, Doherty M, Leeb BF, Alekseeva L, Arden NK, Bijlsma JW, et al. EULAR evidence-based recommendations for the diagnosis of hand osteoarthritis: report of a task force of ESCISIT. Ann Rheum Dis 2009;68:8 17. 3. Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D, Brandt K, et al. The American College of Rheumatology criteria for the classification and reporting of osteoarthritis of the hand. Arthritis Rheum 1990;33:1601 10. 4. Kellgren J, Moore R. Generalized osteoarthritis and Heberden s nodes. Br Med J 1952;1:181 7. 5. Caspi D, Flusser G, Farber I, Ribak J, Leibovitz A, Habot B, et al. Clinical, radiologic, demographic, and occupational aspects of hand osteoarthritis in the elderly. Semin Arthritis Rheum 2001;30:321 31. 6. Cicuttini FM, Baker J, Hart DJ, Spector TD. Relation between Heberden s nodes and distal interphalangeal joint osteophytes and their role as markers of generalised disease. Ann Rheum Dis 1998;57:246 8. 7. Thaper A, Zhang W, Wright G, Doherty M. Relationship between Heberden s nodes and underlying radiographic changes of osteoarthritis. Ann Rheum Dis 2005;64:1214 6. 8. Hart D, Spector T, Egger P, Coggon D, Cooper C. Defining osteoarthritis of the hand for epidemiological studies: the Chingford Study. Ann Rheum Dis 1994;53:220 3. 9. Kalichman L, Hernandez-Molina G. Hand osteoarthritis: an epidemiological perspective. Semin Arthritis Rheum 2010;39: 465 76. 10. Solovieva S, Vehmas T, Riihimaki H, Luoma K, Leino-Arjas P. Hand use and patterns of joint involvement in osteoarthritis: a comparison of female dentists and teachers. Rheumatology (Oxford) 2005;44:521 8. 11. Acheson RM, Chan YK, Clemett AR. New Haven survey of joint diseases. XII. Distribution and symptoms of osteoarthrosis in the hands with reference to handedness. Ann Rheum Dis 1970;29:275 86. 12. Zhang W, Robertson J, Doherty S, Liu JJ, Maciewicz RA, Muir KR, et al. Index to ring finger length ratio and the risk of osteoarthritis. Arthritis Rheum 2008;58:137 44. 13. O Reilly S, Johnson S, Doherty S, Muir K, Doherty M. Screening for hand osteoarthritis (OA) using a postal survey. Osteoarthritis Cartilage 1999;7:461 5. 14. Altman RD, Hochberg M, Murphy WA Jr, Wolfe F, Lequesne M. Atlas of individual radiographic features in osteoarthritis. Osteoarthritis Cartilage 1995;3 Suppl A:3 70. 15. Kellgren J, Lawrence J. Radiological assessment of osteo-arthrosis. Ann Rheum Dis 1957;16:494 502. 16. Niu J, Zhang Y, LaValley M, Chaisson CE, Aliabadi P, Felson DT. Symmetry and clustering of symptomatic hand osteoarthritis in elderly men and women: the Framingham Study. Rheumatology (Oxford) 2003;42:343 8. 17. Wilder FV, Barrett JP, Farina EJ. Joint-specific prevalence of osteoarthritis of the hand. Osteoarthritis Cartilage 2006;14: 953 7. 18. Buckland-Wright JC, Macfarlane DG, Lynch JA. Osteophytes in the osteoarthritic hand: their incidence, size, distribution, and progression. Ann Rheum Dis 1991;50:627 30. 19. Hadler NM, Gillings DB, Imbus HR, Levitin PM, Makuc D, Utsinger PD, et al. Hand structure and function in an industrial setting. Arthritis Rheum 1978;21:210 20. 20. Flatt AE. Our thumbs. Proc (Bayl Univ Med Cent) 2002;15: 380 7. 21. Srikanth VK, Fryer JL, Zhai G, Winzenberg TM, Hosmer D, Jones G. A meta-analysis of sex differences prevalence, incidence and severity of osteoarthritis. Osteoarthritis Cartilage 2005;13:769 81.