- Email: [email protected]
Incidence and progression of hand osteoarthritis in a large community-based cohort: the Johnston County Osteoarthritis Project
Journal Pre-proof Incidence and Progression of Hand Osteoarthritis in a Large Community-Based Cohort: The Johnston County Osteoarthritis Project E. Amanda Snyder, MD, Carolina Alvarez, MS, Yvonne M. Golightly, PT PhD, Jordan B. Renner, MD, Joanne M. Jordan, MD MPH, Amanda E. Nelson, MD MSCR PII:
S1063-4584(20)30094-7
DOI:
https://doi.org/10.1016/j.joca.2020.02.028
Reference:
YJOCA 4604
To appear in:
Osteoarthritis and Cartilage
Received Date: 24 September 2019 Revised Date:
23 December 2019
Accepted Date: 3 February 2020
Please cite this article as: Snyder EA, Alvarez C, Golightly YM, Renner JB, Jordan JM, Nelson AE, Incidence and Progression of Hand Osteoarthritis in a Large Community-Based Cohort: The Johnston County Osteoarthritis Project, Osteoarthritis and Cartilage, https://doi.org/10.1016/j.joca.2020.02.028. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.
1
Incidence and Progression of Hand Osteoarthritis in a Large Community-Based Cohort:
2
The Johnston County Osteoarthritis Project
3 4
E. Amanda Snyder MD1,2, Carolina Alvarez MS1, Yvonne M. Golightly PT PhD1,3, Jordan B.
5
Renner MD1,4, Joanne M. Jordan MD MPH1,2,3, Amanda E. Nelson MD MSCR1,2
6
1. Thurston Arthritis Research Center, University of North Carolina at Chapel Hill
7
2. Department of Medicine, University of North Carolina at Chapel Hill
8
3. Department of Epidemiology, Gillings School of Global Public Health, University of North
9
Carolina at Chapel Hill
10
4. Department of Radiology, University of North Carolina at Chapel Hill
11 12
Corresponding author: Amanda E. Nelson MD MSCR, University of North Carolina at Chapel
13
Hill, 3300 Doc J. Thurston Building, Campus Box 7280, Chapel Hill, NC 27599-7280
14 15
Emails - EAS: [email protected]; CA: [email protected]; YMG:
16
[email protected]; JBR: [email protected]; JMJ:
17
[email protected]; AEN: [email protected]
1
18
Objective: To describe the incidence and progression of radiographic and symptomatic hand
19
osteoarthritis (rHOA and sxHOA) in a large community-based cohort.
20
Design: Data were from the Johnston County OA Project (1999-2015, 12 ±1.2 years follow-up,
21
age 45+). Participants had bilateral hand radiographs each visit, read for Kellgren-Lawrence
22
grade (KLG) at 30 joints. We defined rHOA as KLG ≥2 in ≥1 joint. SxHOA was defined in a
23
hand/joint with rHOA and self-reported symptoms or tenderness on exam. Incidence was
24
assessed in those without, while progression was assessed in those with, baseline rHOA.
25
Proportions or medians are reported; differences by sex and race were assessed using models
26
appropriate for dichotomous or continuous definitions, additionally adjusted for age, education,
27
BMI, and weight change.
28
Results: Of 800 participants (68% women, 32% African American, mean age 60 years), 327
29
had baseline rHOA and were older, more often white and female, than those without rHOA
30
(n=473). The incidence of HOA was high, for rHOA (60%) and for sxHOA (13%). Women were
31
more likely than men to have incident HOA, particularly for DIP rOA (aOR 1.60 95% CI [1.03,
32
2.49]) and sxHOA (aOR 2.98 [1.50, 5.91]). Progressive HOA was more similar by sex, although
33
thumb base rOA progressed more frequently in women than in men (aOR 2.56 [1.44, 4.55]).
34
Particularly HOA incidence, but also progression, was more frequent among whites compared
35
with African Americans.
36
Conclusion: This study provides much needed information about the natural history of HOA, a
37
common and frequently debilitating condition, in the general population.
38 39
Keywords: Hand osteoarthritis, radiographs, epidemiology
40 41
Running title: Incidence and Progression of Hand OA
42 43 2
44
The presentation of osteoarthritis (OA) varies immensely – affecting adults of all ages and
45
involving different joints throughout the body.1 Although hand osteoarthritis (HOA) is
46
exceedingly common and will affect 4 in 10 people by age 852, this joint site is understudied in
47
comparison to other sites, like the knee. HOA is a heterogeneous disease that presents as
48
bony enlargement and deformities of various joints in the hand.3 Patients may experience pain,
49
stiffness and decreased mobility, which can result in decreased grip strength and functionality of
50
the hand.4 As such, HOA can lead to significant disability and increased healthcare utilization.5
51 52
To date, studies of HOA are limited – particularly with regard to incidence and progression of
53
disease in longitudinal cohorts. Review of the literature reveals that prevalence estimates of
54
HOA in adults vary depending on the population studied and definitions used to define the
55
disease. HOA can be defined radiographically and/or symptomatically, and such definitions can
56
affect frequency estimates.6 For example, the Framingham OA Study estimated the frequency
57
of symptomatic HOA as 13% in men and 26% in women > 70 years of age, but only includes
58
white individuals.7 As the US population is aging, these data suggest more people will be
59
affected by HOA in the future highlighting the need for better understanding of the natural
60
history of this common condition, including differences by sex and race.
61 62
The purpose of this study was to utilize recent longitudinal data to describe the incidence and
63
progression of radiographic and symptomatic HOA (rHOA and sxHOA, respectively) in a large
64
community-based cohort of African American and white men and women.
65 66
Participants and Methods
67
Data were obtained from adult participants (age ≥ 45 years) in the Johnston County
68
Osteoarthritis Project (JoCo OA), an ongoing community-based prospective observational
69
cohort study monitoring the occurrence and natural history of OA in Johnston County, North 3
70
Carolina.8 JoCo OA enrolled non-institutionalized white and African American adults age 45
71
and older who resided in 6 townships in the county in a population-based manner (without
72
regard for symptoms or diagnosis). All participants provide informed consent, answer
73
standardized questionnaires and undergo an examination with trained staff, in addition to
74
radiography of multiple joint sites. This analysis included all participants at the third JoCo OA
75
follow up (n=909, collected 2013-15), and included data from all prior time points with hand data
76
for these individuals (first follow-up: 1999-2003 or cohort enrichment: 2003-4, and second
77
[combined] follow up: 2006-10). After exclusion of 109 participants due to missing radiographs
78
over time (n=84) or missing any of the 30 specific joint scores over time (n=25), 800 participants
79
were included in the complete case analysis (Figure 1).
80
Sex, race, and age were from self-report at baseline. Educational attainment at baseline was
81
categorized as completing less than a high school education vs at least a high school education.
82
Body mass index (BMI) at baseline was computed from measured standing height (cm) and
83
weight (kg) during the research clinic visit. Weight gain from baseline (T1) to second follow-up
84
(T3) was defined as more than a 5% increase in BMI.
85
All participants had posteroanterior radiographs of the bilateral hands at each visit, read
86
together in known time order9, 10 by an experienced musculoskeletal radiologist (JBR) using the
87
Kellgren-Lawrence grading (KLG) system11 for all 30 joints (bilateral distal interphalangeal [DIP]
88
2-5; proximal interphalangeal [PIP] 2-5; metacarpophalangeal [MCP] 1-5; interphalangeal [IP]
89
and carpometacarpal [CMC] of the thumb). The radiologist has previously demonstrated high
90
reliability (kappa >0.7), detailed by joint in a prior paper.2 Erosive OA was not systematically
91
assessed in this study, but noted only as a comment and appeared to be relatively uncommon
92
in this sample (9 participants). Therefore, erosive OA was not analyzed separately, although
93
these participants were included in the analyses for other outcomes.
94
4
95
There are many ways to define rHOA.12 For these analyses, we considered several definitions.
96
First, rHOA overall was defined as a KLG of 2 or more in at least 1 of the 30 hand joints.
97
Additionally, rOA by joint group (DIP, PIP, thumb IP, MCP, or CMC, involving at least one joint
98
in the group) and rHOA affecting 2 or more joints was assessed. We also considered the
99
number of joints affected by KLG of 2 or more and the sum score of KLG (0-4 at each joint)
100
across all 30 joints (range 0-120). SxHOA was defined as the presence of symptoms defined in
101
one of two ways with rOA in the same hand or specific joint site. First, self-reported pain, aching
102
or stiffness in the entire hand was assessed using the answer to “On most days, do you have
103
pain, aching, or stiffness in your hands” at baseline, which was modified at follow up to: “On
104
most days of any one month in the last twelve months, did you have pain, aching, or stiffness in
105
your hands?” at follow up, both with indicators for right or left hand. Second, tenderness to
106
palpation at each of the 30 hand joints during the research exam was assessed (previously
107
shown to be reliable, with agreement >0.9 for two trained readers1). We additionally considered
108
the number of joints with sxHOA.
109 110
Two groups were defined: those with baseline rHOA (KLG of 2 or more in at least 1 of the 30
111
hand joints) and those without. Among those without baseline rHOA, we defined incident rHOA
112
as newly meeting the above criteria in a hand or joint previously without rOA. Among those with
113
baseline rHOA, progression was defined as an additional joint involved, either overall or in a
114
given joint group for both rHOA and sxHOA. We again considered the overall number of rHOA
115
joints and sxHOA joints as well as the total sum KLG for progression.
116 117
Incidence and progression were assessed over an average of 12 years involving a baseline
118
(either first follow up: 1999-2004 or cohort enrichment: 2003-4) and two combined follow-up
119
time-points (2006-2010 and 2013-2015). Descriptive statistics were shown as proportions for
120
dichotomous variables and medians (interquartile range) for continuous variables. 5
121
Most incidence and progression differences between sex and race were assessed by modeling
122
dichotomous definitions with logistic regression and producing adjusted odds ratios (aOR) and
123
95% confidence intervals (95% CI) describing associations. For continuous definitions
124
(consisting of skewed count distributions) either zero-inflated negative binomial, scaled negative
125
binomial or scaled regression models were used to produce adjusted incidence density ratios
126
(aIDR) or adjusted betas (β) and 95% CI describing associations. Models were adjusted for
127
age, sex, race, education, BMI, weight gain of 5% or more, and HOA baseline values (for
128
progression only).
129 130
Results
131
Of the 800 included participants (Figure 1), 68% were women and 32% were African American
132
(Table 1). The mean age of the sample was 59.6 (standard deviation ± 7.5) years, and the
133
mean body mass index (BMI) was 31.0 ± 6.4 kg/m2. Those with baseline rHOA (n=327, 41% of
134
the sample) tended to be older (63 ± 7 vs 57 ± 7 years) and were more often white (80% vs
135
61%) and women (72% vs 65%) than those without baseline rHOA (n=473, 59% of the sample).
136
Socioeconomic status as represented by education level did not differ between groups, nor did
137
the proportion experiencing weight gain over the course of the study (Table 1).
138 139 140
Incident HOA The frequencies of incident rHOA (≥1 or ≥2 joints), DIP rOA, PIP rOA, thumb IP rOA,
141
and thumb base rOA were slightly higher among women, while men were more likely to have
142
incident MCP rOA (Table 2). In adjusted models, women had statistically significantly greater
143
odds of DIP rOA (aOR 1.60, 95% CI [1.03, 2.49]) and had, on average, 1.33 times the number
144
of rHOA affected joints compared with men.
145 146
Incident sxHOA was significantly more likely in women (16% vs 7%), who also had more joints with sxHOA (but not rHOA) compared with men. Both differences were statistically 6
147
significant in adjusted models, where women had nearly 3 times the odds of developing sxHOA
148
and had 4.7 times more joints affected with sxHOA than men. The overall KLG sum score
149
across all joints was similar by sex.
150
All definitions of rHOA and sxHOA were more common in white compared with African
151
American participants, particularly for rHOA (≥1 joint), DIP OA, and thumb base OA (Table 2),
152
all of which were at least twice as likely among whites in adjusted models. SxHOA was more
153
than twice as frequent in white compared to African American participants (17% vs 6%), with an
154
adjusted odds ratio over 4. The number of joints affected by rHOA or sxHOA, and the overall
155
KLG sum, were also higher in white participants, although the adjusted IDR was only significant
156
for number of joints with rHOA (aIDR 2.00 [1.54, 2.61]).
157
Progression of HOA
158
Progressive rHOA was seen at similar frequencies between men and women, except for
159
MCP rOA, which more commonly progressed in men (58% vs 39%), and thumb base rOA,
160
which progressed more often in women (61% vs 41%, Table 3). Only thumb base OA was
161
statistically significant in adjusted models, where women had 2.5 times the odds of progression
162
compared with men. Progressive rHOA and sxHOA were slightly but not statistically significantly
163
more common in women compared with men. Women also had more joints involved with rHOA
164
and sxHOA over time, although this was not significant in adjusted models.
165
All measures of HOA progression were higher among whites compared with African
166
Americans, although most were modestly so (Table 3). Thumb base rOA progression was
167
significantly more frequent among white compared with African American participants (61% vs
168
34%), as was sxHOA progression (32% vs 25%, respectively); both remained significant in
169
adjusted models. Additionally, the number of joints affected and the change in number of joints
170
involved were slightly higher among whites compared with African Americans, although not
171
statistically significant after adjustment.
172 7
173
Discussion
174
HOA remains understudied in comparison to other common sites of OA, like the hip and knee.
175
While several studies have described the prevalence of hand OA in various populations, there
176
have been surprisingly few analyses of the incidence and progression of OA in longitudinal,
177
community-based cohorts. An understanding of the natural history of this common and often
178
debilitating condition is essential to informing further study, including the development of new
179
interventions. In this longitudinal, community-based cohort including African American and white
180
men and women, we found high frequencies of new disease over an average of 12 years,
181
particularly for rHOA (60%) but also for sxHOA (13%). For most joint sites, compared with men,
182
women were somewhat more likely to have incident HOA, although the opposite was true at the
183
MCP. The differences by race were more apparent, with white participants experiencing much
184
higher frequencies of incident HOA by most definitions. HOA progression was similarly
185
common, with rHOA progression occurring in the majority of the sample and sxHOA in about
186
two out of every five (41%). Differences by sex were similar for progression and for incidence,
187
while differences by race were somewhat less pronounced. Of note, more than 1/4 of the
188
sample, regardless of subgroup, had progressive sxHOA with at least one additional joint
189
involved over the follow-up period (28% of African Americans, 35% of men, 43% of whites and
190
44% of women), emphasizing the burden of this condition.
191 192
There are a few large cohort studies to which we can compare our findings. Haugen, et al13,
193
using data from the Framingham OA Study (FOA, collected 1992-5 and 2002-5), reported 9-
194
year cumulative incidence for similar definitions of HOA that were overall lower than those
195
reported here (e.g., rHOA approximately 34%, PIP 5%, although thumb base OA was
196
comparable at 17% in men and 21% in women). However, the pattern of sex differences was
197
similar between the two studies. For progression over 9 years, the overall frequencies were
198
fairly similar to the current report (e.g., rHOA progression in >90%), although DIP progression 8
199
was slightly lower among women in the JoCo OA than in the FOA (74% vs. 78%, respectively)
200
but higher in men, (72% in JoCo OA vs. 59% in FOA). Additionally, PIP and MCP progression
201
were more common in JoCo OA compared with FOA. Thumb base OA progression was more
202
frequent in FOA (<50% in JoCo OA, >60% in FOA), again with similar patterns by sex. Among
203
individuals with hand pain but without baseline rHOA in the Clinical Assessment Studies of the
204
Hand study14, the KLG sum (for 20 joints; no MCP) increased by 9 in women and 7 in men over
205
7 years, not dissimilar to the increase of 12 and 11 in the current analysis over 12 years
206
(although the latter sum was for 30 joints). In the Genetics, Arthrosis and Progression study15, a
207
family study of multiple joint OA, the pattern of progression was similar to the present study, with
208
HOA progression most frequent in the DIPs, then CMC, PIP/IP and MCPs. In a few smaller
209
studies in more specific populations, progression has been shown to be a common occurrence,
210
although often modest in severity16-18. Of course, these are very different populations and
211
utilized somewhat different methodologies so are not directly comparable. The fact that there
212
are not more studies, and that there are none in diverse populations (all of the above studies
213
were among whites), highlights the need for additional work in this area.
214 215
We have previously reported on the cross-sectional frequency of multiple joint OA phenotypes,
216
including HOA, by race and sex in the JoCo OA.19, 20 We found that, compared to whites, African
217
Americans were less likely to have phenotypes that included rHOA20 or sxHOA19, although they
218
were more likely to have phenotypes involving one or more large joints. Additionally, the lifetime
219
risk of sxHOA was lower among African Americans compared with whites in this cohort (29.2%
220
vs. 41.4%).2 A large cross-sectional study in young (age 42-53) African American and white
221
women nearly two decades ago found a similar prevalence of hand OA by race, although MCP
222
OA was more common among African American (12%) compared with white (3%) women;
223
African American women were also more likely to have both hand and knee OA (8% vs 2% of
224
white women)21. Unfortunately, we are not aware of any other community-based studies 9
225
including African Americans that are available for comparison, and the mechanisms behind
226
these differences are not known.
227 228
Limitations of this work include small numbers in some categories which precluded more
229
detailed modeling. As in most large cohorts, we have relatively infrequent assessments of
230
symptoms, and are not able to model detailed fluctuations in symptoms over time. We focused
231
on incidence and progression and did not separately analyze improvements in symptoms over
232
time. However, the strengths, including the large overall sample size, the unique biracial cohort
233
including both men and women, and the community-based nature of the cohort, make this an
234
important contribution to the HOA literature.
235 236
In this large, longitudinal, community-based cohort inclusive of African American and white men
237
and women, we found overall high frequencies of both incidence and progression of HOA over
238
an average of 12 years of follow up. Although common across the board, the frequencies of
239
incident and progressive HOA by a variety of definitions were generally higher in women
240
compared with men and among whites compared with African Americans; differences were
241
overall more marked for definitions of incidence than for progression. Additional studies are
242
needed to understand potential mechanisms for the identified differences (e.g., occupational,
243
genetic, nutritional, biomechanical, or other factors).
244 245
10
246
Acknowledgments: We would like to thank the staff and participants in the Johnston County
247
Osteoarthritis Project, without whom this work would not be possible.
248
Author contributions:
249
Conception and design: EAS, CA, YMG, AEN
250
Acquisition of data: JBR, JMJ
251
Analysis and interpretation of data: EAS, CA, YMG, JBR, JMJ, AEN
252
Drafting the article or revising it critically for important intellectual content: EAS, CA, YMG, JBR,
253
JMJ, AEN
254
Final approval of the version to be submitted: EAS, CA, YMG, JBR, JMJ, AEN
255
Dr. Nelson takes responsibility for the integrity of the work as a whole.
256
Role of the funding source:
257
This work is funded in part by: CDC U01 DP006266 and DP003206, NIH/NIAMS P60
258
AR049465 and AR064166. The funding agencies had no role in the design, analysis, or
259
interpretation of data, nor in the writing of the manuscript or the decision to submit it for
260
publication.
261
Competing interest statement:
262
None of the authors has any competing interests in relation to this work.
263 264 265 266 267 268
Figure Legend Figure 1. Flow chart of participant inclusion in the current analysis
11
269 270 271
References 1.
Gullo TR, Golightly YM, Cleveland RJ, Renner JB, Callahan LF, Jordan JM, et al.
272
Defining multiple joint osteoarthritis, its frequency and impact in a community-based
273
cohort. Semin Arthritis Rheum 2019;48:950-957. doi:10.1016/j.semarthrit.2018.10.001.
274
2.
Qin J, Barbour KE, Murphy LB, Nelson AE, Schwartz TA, Helmick CG, et al. Lifetime
275
Risk of Symptomatic Hand Osteoarthritis: The Johnston County Osteoarthritis Project.
276
Arthritis Rheumatol 2017;69:1204-1212. doi:10.1002/art.40097.
277
3.
Zhang W, Doherty M, Leeb BF, Alekseeva L, Arden NK, Bijlsma JW, et al. EULAR
278
evidence-based recommendations for the diagnosis of hand osteoarthritis: report of a
279
task force of ESCISIT. Ann Rheum Dis 2009;68:8-17. doi:10.1136/ard.2007.084772.
280
4.
Michon M, Maheu E, Berenbaum F. Assessing health-related quality of life in hand
281
osteoarthritis: a literature review. Ann Rheum Dis 2011;70:921-928.
282
doi:10.1136/ard.2010.131151.
283
5.
Oliveria SA, Felson DT, Reed JI, Cirillo PA, Walker AM. Incidence of symptomatic hand,
284
hip, and knee osteoarthritis among patients in a health maintenance organization.
285
Arthritis Rheum 1995;38:1134-1141.
286
6.
Pereira D, Peleteiro B, Araujo J, Branco J, Santos RA, Ramos E. The effect of
287
osteoarthritis definition on prevalence and incidence estimates: a systematic review.
288
Osteoarthritis Cartilage 2011;19:1270-1285. doi:10.1016/j.joca.2011.08.009.
289
7.
Zhang Y, Niu J, Kelly-Hayes M, Chaisson CE, Aliabadi P, Felson DT. Prevalence of
290
symptomatic hand osteoarthritis and its impact on functional status among the elderly:
291
The Framingham Study. Am J Epidemiol 2002;156:1021-1027.
292
8.
Jordan JM, Helmick CG, Renner JB, Luta G, Dragomir AD, Woodard J, et al. Prevalence
293
of knee symptoms and radiographic and symptomatic knee osteoarthritis in African
294
Americans and Caucasians: the Johnston County Osteoarthritis Project. J Rheumatol
295
2007;34:172-180. 12
296
9.
Felson DT, Nevitt MC. Blinding images to sequence in osteoarthritis: evidence from
297
other diseases. Osteoarthritis Cartilage 2009;17:281-283.
298
doi:10.1016/j.joca.2008.09.008.
299
10.
Gensburger D, Roux JP, Arlot M, Sornay-Rendu E, Ravaud P, Chapurlat R. Influence of
300
blinding sequence of radiographs on the reproducibility and sensitivity to change of joint
301
space width measurement in knee osteoarthritis. Arthritis Care Res (Hoboken)
302
2010;62:1699-1705. doi:10.1002/acr.20311.
303
11.
304 305
Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis 1957;16:494-502. doi:10.1136/ard.16.4.494.
12.
Marshall M, Dziedzic KS, van der Windt DA, Hay EM. A systematic search and narrative
306
review of radiographic definitions of hand osteoarthritis in population-based studies.
307
Osteoarthritis Cartilage 2008;16:219-226. doi:10.1016/j.joca.2007.06.004.
308
13.
Haugen IK, Englund M, Aliabadi P, Niu J, Clancy M, Kvien TK, et al. Prevalence,
309
incidence and progression of hand osteoarthritis in the general population: the
310
Framingham Osteoarthritis Study. Ann Rheum Dis 2011;70:1581-1586.
311
doi:10.1136/ard.2011.150078.
312
14.
Marshall M, Peat G, Nicholls E, Myers HL, Mamas MA, van der Windt DA. Metabolic risk
313
factors and the incidence and progression of radiographic hand osteoarthritis: a
314
population-based cohort study. Scand J Rheumatol 2019;48:52-63.
315
doi:10.1080/03009742.2018.1459831.
316
15.
Bijsterbosch J, Meulenbelt I, Watt I, Rosendaal FR, Huizinga TW, Kloppenburg M.
317
Clustering of hand osteoarthritis progression and its relationship to progression of
318
osteoarthritis at the knee. Ann Rheum Dis 2014;73:567-572. doi:10.1136/annrheumdis-
319
2012-202461.
13
320
16.
Paradowski PT, Lohmander LS, Englund M. Natural history of radiographic features of
321
hand osteoarthritis over 10 years. Osteoarthritis Cartilage 2010;18:917-922.
322
doi:10.1016/j.joca.2010.04.008.
323
17.
Balblanc JC, Mathieu P, Mathieu L, Tron AM, Conrozier T, Piperno M, et al. Progression
324
of digital osteoarthritis: a sequential scintigraphic and radiographic study. Osteoarthritis
325
Cartilage 1995;3:181-186. doi:10.1016/s1063-4584(05)80052-4.
326
18.
Harris PA, Hart DJ, Dacre JE, Huskisson EC, Spector TD. The progression of
327
radiological hand osteoarthritis over ten years: a clinical follow-up study. Osteoarthritis
328
Cartilage 1994;2:247-252. doi:10.1016/s1063-4584(05)80076-7.
329
19.
Nelson AE, Golightly YM, Renner JB, Schwartz TA, Kraus VB, Helmick CG, et al. Brief
330
report: differences in multijoint symptomatic osteoarthritis phenotypes by race and sex:
331
the Johnston County Osteoarthritis Project. Arthritis Rheum 2013;65:373-377.
332
doi:10.1002/art.37775.
333
20.
Nelson AE, Renner JB, Schwartz TA, Kraus VB, Helmick CG, Jordan JM. Differences in
334
multijoint radiographic osteoarthritis phenotypes among African Americans and
335
Caucasians: The Johnston County Osteoarthritis project. Arthritis Rheum 2011;63:3843-
336
3852. doi:10.1002/art.30610.
337
21.
Sowers M, Lachance L, Hochberg M, Jamadar D. Radiographically defined osteoarthritis
338
of the hand and knee in young and middle-aged African American and Caucasian
339
women. Osteoarthritis Cartilage 2000;8:69-77. doi:10.1053/joca.1999.0273.
340 341
14
Table 1. Descriptive characteristics of the sample overall and by baseline hand OA status Sample
Overall (n=800)
No baseline hand OA
Baseline hand OA
(n=473)
(n=327)
mean (SD) or %
mean (SD) or %
mean (SD) or %
Age, years
59.6 (7.5)
57.3 (6.6)
63.1 (7.4)
Female, %
68.0
65.3
71.9
African American, %
31.5
39.5
19.9
<12 years education, %
13.8
13.8
13.8
31.0 (6.4)
31.6 (6.7)
30.2 (5.9)
32.6
33.4
31.5
Characteristics
Baseline body mass index, kg/m2 >5% weight gain from baseline to follow up, %
15
Table 2. Proportions by hand OA incidence definition over ~ 12 (±1.2) year follow-up for those without baseline hand OA, by sex or race Hand OA incidence definition
Overall
Women
Men
(n=473)
(n=309)
(n=164)
Women vs Men*
White
African
(n=286)
American
White vs AA*
(n=187) % or median
effect (95% CI)
% or median
(interquartile
(interquartile
range: Q1-Q3)
range: Q1-Q3)
effect (95% CI)
59.4%
60.5%
57.3%
1.26 (0.84, 1.90)
65.0%
50.8%
2.26 (1.49, 3.43)
– 5 ) OA
32.1%
34.0%
28.7%
1.60 (1.03, 2.49)
40.2%
19.8%
3.16 (1.97, 5.07)
PIP (2nd – 5th) OA
20.1%
21.4%
17.7%
1.24 (0.82, 1.87)
22.7%
16.0%
1.17 (0.77, 1.77)
Thumb IP OA
24.5%
24.9%
23.8%
1.02 (0.64, 1.61)
23.1%
26.7%
0.87 (0.55, 1.37)
MCP (1 – 5 ) OA
13.5%
11.7%
17.1%
0.59 (0.34, 1.04)
13.6%
13.4%
1.08 (0.60, 1.95)
Thumb base OA
20.7%
22.3%
17.7%
1.59 (0.96, 2.64)
24.1%
15.5%
2.11 (1.24, 3.58)
Radiographic HOA (≥2 joint(s))
31.7%
33.0%
29.3%
1.18 (0.78, 1.79)
35.3%
26.2%
1.60 (1.05, 2.45)
Symptomatic HOA (≥1 joint(s))
12.9%
15.9%
7.3%
2.98 (1.50, 5.91)
17.1%
6.4%
4.29 (2.08, 8.86)
Number of radiographic HOA
1 (0-3)
1 (0-3)
1 (0-2)
1.33 (1.03, 1.72)
1
1 (0-3)
1 (0-2)
2.00 (1.54, 2.61)
1
0 (0-0)
0 (0-0)
0 (0-0)
4.74 (2.08, 10.8)
1
0 (0-0)
0 (0-0)
1.69 (0.41, 7.06)
1
Max=93
Max=9
Max=2
Max=9
Max=3
At T3:
At T3:
At T3:
At T3:
At T3:
Radiographic HOA (≥1 joint(s)) DIP (2
nd
st
th
th
joints Number of symptomatic HOA joints Total sum KL score (30 joints)
0.99 (0.95, 1.03)2
1.00 (0.96, 1.04)2
16
11 (7-19)
12 (8-19)
11 (7-18)
12 (8-20)
11 (7-16)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
7 (4-12)
7 (4-12)
8 (4-11)
8 (5-14)
6 (3-9)
Distal interphalangeal joint (DIP), Hand osteoarthritis (HOA), Interphalangeal joint (IP), Kellgren-Lawrence grade (KLG), Metacarpophalangeal joint (MCP), Osteoarthritis (OA), Proximal interphalangeal joint (PIP); ∆=change from T1 to T3 (baseline to follow up) Effects shown are adjusted for baseline: age, sex, race, education, BMI, and weight gain—additionally adjusted for HOA baseline values for total sum KL score. *Logit models for categorical definitions producing adjusted odds ratios (aOR) and 1 zero inflated negative binomial models for continuous definitions producing adjusted incidence density ratios (aIDR) and 2 scaled negative binomial model for continuous definitions producing adjusted incidence density ratios (aIDR)
17
Table 3. Proportion by hand OA progression definition over ~ 12 (±1.2) year follow-up for those with baseline hand OA, by sex or race Hand OA progression definition
Overall
Women
Men
(n=327)
(n=235)
(n=92)
Women vs Men*
White
African
(n=262)
American
White vs AA*
(n=65) % or median
effect (95% CI)
(interquartile
effect (95% CI)
(interquartile
range: Q1-Q3) Radiographic HOA
% or median
range: Q1-Q3)
94.2%
94.0%
94.6%
0.70 (0.22, 2.22)
95.4%
89.2%
1.15 (0.38, 3.45)
n=306
n=218
n=88
1.26 (0.67, 2.37)
n=242
n=64
1.39 (0.70, 2.76)
77.8%
79.4%
73.9%
81.0%
65.6%
n=318
n=228
n=90
n=254
n=64
63.5%
66.2%
56.7%
65.0%
57.8%
n=286
n=205
n=81
n=228
n=58
56.3%
54.2%
61.7%
57.9%
50.0%
MCP (1 – 5 ) OA
44.0%
38.7%
57.6%
0.61 (0.36, 1.05)
44.7%
41.5%
0.96 (0.52, 1.77)
Thumb base OA
n=274
n=193
n=81
2.56 (1.44, 4.55)
n=215
n=59
3.39 (1.74, 6.60)
54.7%
60.6%
40.7%
60.5%
33.9%
80.4%
81.3%
78.3%
82.4%
72.3%
(additional in ≥1 joint(s)) nd
th
DIP (2 – 5 ) OA
nd
th
PIP (2 – 5 ) OA
Thumb IP OA
st
th
Radiographic HOA
1.38 (0.80, 2.37)
0.80 (0.46, 1.37)
1.11 (0.59, 2.10)
0.94 (0.51, 1.76)
1.30 (0.71, 2.40)
1.08 (0.53, 2.18)
18
(additional ≥2 joint(s))
Symptomatic HOA
31.0%
31.9%
28.6%
1.25 (0.72, 2.16)
32.4%
25.0%
At T3:
At T3:
At T3:
1
At T3:
At T3:
9 (4-14)
9 (5-15)
8 (4-12)
10 (5-15)
6 (3-10)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
5 (2-8)
6 (3-9)
5 (2-7)
6 (3-8)
4 (2-8)
Number of symptomatic
At T3:
At T3:
At T3:
At T3:
At T3:
HOA joints
0 (0-1)
0 (0-2)
0 (0-1)
0 (0-1)
0 (0-0)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
0 (0-1)
0 (0-1)
0 (0-0)
0 (0-1)
0 (0-0)
At T3:
At T3:
At T3:
At T3:
At T3:
34 (21-45)
35 (22-47)
32.5 (19-42)
35 (23-47)
27 (17-37)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
16 (9-22)
16 (10-23)
14.5 (8.5-20.5)
17 (10-22)
13 (8-18)
1.96 (1.01, 3.83)
(additional ≥1 joint(s)) Number of radiographic HOA joints
Total sum KL score (30 joints)
1.17 (0.99, 1.39)
0.23 (-0.34, 0.81)
1.09 (0.97, 1.24)
2
1
1.12 (0.92, 1.36)
1
0.46 (-0.21, 1.12)
1.09 (0.94, 1.26)
2
1
Distal interphalangeal joint (DIP), Hand osteoarthritis (HOA), Interphalangeal joint (IP), Kellgren-Lawrence grade (KLG), Metacarpophalangeal joint (MCP), Osteoarthritis (OA), Proximal interphalangeal joint (PIP); ∆=change from T1 to T3 (baseline to follow up) Effects shown are adjusted for baseline: age, sex, race, education, BMI, weight gain, and HOA baseline values. *Logit models for categorical definitions producing adjusted odds ratios (aOR) and 1 scaled negative binomial model for continuous definitions producing adjusted incidence density ratios (aIDR) and 2 scaled regression models for the number of symptomatic HOA joints producing adjusted betas.
19
STROBE Statement—Checklist of items that should be included in reports of cohort studies
Title and abstract
Item No 1
Line or page # in ms Recommendation (a) Indicate the study’s design with a commonly used term in the
Line 1
title or the abstract (b) Provide in the abstract an informative and balanced summary of
Page 2
what was done and what was found Introduction Background/rationale
2
Explain the scientific background and rationale for the
Page 3
investigation being reported Objectives
3
State specific objectives, including any prespecified hypotheses
Methods Study design
4
Present key elements of study design early in the paper
Setting
5
Describe the setting, locations, and relevant dates, including
Lines 68-70 Page 3-5 Lines 73-82
periods of recruitment, exposure, follow-up, and data collection Participants
6
(a) Give the eligibility criteria, and the sources and methods of
Lines 73-82 and 109-
selection of participants. Describe methods of follow-up
115
(b) For matched studies, give matching criteria and number of
n/a
exposed and unexposed Variables
7
Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if
Lines 84-107
applicable Data sources/
8*
measurement
For each variable of interest, give sources of data and details of
Lines 84-115
methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group
Bias
9
Describe any efforts to address potential sources of bias
Study size
10
Explain how the study size was arrived at
Lines 78-82
Quantitative
11
Explain how quantitative variables were handled in the analyses. If
Lines 92-107
variables Statistical methods
n/a (descriptive)
applicable, describe which groupings were chosen and why 12
(a) Describe all statistical methods, including those used to control
Lines 109-115
for confounding (b) Describe any methods used to examine subgroups and
Lines 109-115
interactions (c) Explain how missing data were addressed (d) If applicable, explain how loss to follow-up was addressed (e) Describe any sensitivity analyses Results Participants
13*
(a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, examined for eligibility, confirmed
Line 82 Line 77-82 n/a Lines 77-82 and 118120
eligible, included in the study, completing follow-up, and analysed (b) Give reasons for non-participation at each stage (c) Consider use of a flow diagram
Lines 77-82 Unable due to journal constraints
Descriptive data
14*
(a) Give characteristics of study participants (eg demographic,
Lines 118-122
clinical, social) and information on exposures and potential confounders (b) Indicate number of participants with missing data for each variable of interest
Complete case analysis, line 82
1
(c) Summarise follow-up time (eg, average and total amount)
Lines 109-111 and
Report numbers of outcome events or summary measures over time
Lines 120-122 and
Tables 1&2 Outcome data
15*
Tables 1&2 Main results
16
(a) Give unadjusted estimates and, if applicable, confounder-
Tables 1 & 2
adjusted estimates and their precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and why they were included (b) Report category boundaries when continuous variables were
n/a
categorized (c) If relevant, consider translating estimates of relative risk into
n/a
absolute risk for a meaningful time period Other analyses
17
Report other analyses done—eg analyses of subgroups and interactions, and sensitivity analyses
Pages 5-6 and Tables 1&2
Discussion Key results
18
Summarise key results with reference to study objectives
Lines 154-166
Limitations
19
Discuss limitations of the study, taking into account sources of
Lines 197-200
potential bias or imprecision. Discuss both direction and magnitude of any potential bias Interpretation
20
Give a cautious overall interpretation of results considering
Lines 202-208
objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence Generalisability
21
Discuss the generalisability (external validity) of the study results
Other information Funding
22
Give the source of funding and the role of the funders for the
Lines 168-187 Lines 222-5
present study and, if applicable, for the original study on which the present article is based *Give information separately for exposed and unexposed groups. Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at http://www.strobe-statement.org.
2
Inclusions N=3085
Exclusions
JoCoOA Project T1/T1* clinic participants (baseline, 1999-2004) No hand X-ray films read (n=1559)*
N=1526
JoCoOA Project T1/T1* clinic participants (1999-2004) with hand X-rays Did not participate at T3 (n=1071) • Lost to follow-up (n=616) • Deceased (n=455)
N=908
JoCoOA Project T3 clinic participants (follow up, 2013-2015) Missing complete x-ray information (n=109) • 83 missing x-rays • 25 missing specific joint scores
N=800
Available KL grades for all 30 hand joints at baseline and follow up
*Based on priority, hand x-rays for JoCo OA participants without follow-up have not yet been read.
S1063-4584(20)30094-7
DOI:
https://doi.org/10.1016/j.joca.2020.02.028
Reference:
YJOCA 4604
To appear in:
Osteoarthritis and Cartilage
Received Date: 24 September 2019 Revised Date:
23 December 2019
Accepted Date: 3 February 2020
Please cite this article as: Snyder EA, Alvarez C, Golightly YM, Renner JB, Jordan JM, Nelson AE, Incidence and Progression of Hand Osteoarthritis in a Large Community-Based Cohort: The Johnston County Osteoarthritis Project, Osteoarthritis and Cartilage, https://doi.org/10.1016/j.joca.2020.02.028. This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of record. This version will undergo additional copyediting, typesetting and review before it is published in its final form, but we are providing this version to give early visibility of the article. Please note that, during the production process, errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. © 2020 Osteoarthritis Research Society International. Published by Elsevier Ltd. All rights reserved.
1
Incidence and Progression of Hand Osteoarthritis in a Large Community-Based Cohort:
2
The Johnston County Osteoarthritis Project
3 4
E. Amanda Snyder MD1,2, Carolina Alvarez MS1, Yvonne M. Golightly PT PhD1,3, Jordan B.
5
Renner MD1,4, Joanne M. Jordan MD MPH1,2,3, Amanda E. Nelson MD MSCR1,2
6
1. Thurston Arthritis Research Center, University of North Carolina at Chapel Hill
7
2. Department of Medicine, University of North Carolina at Chapel Hill
8
3. Department of Epidemiology, Gillings School of Global Public Health, University of North
9
Carolina at Chapel Hill
10
4. Department of Radiology, University of North Carolina at Chapel Hill
11 12
Corresponding author: Amanda E. Nelson MD MSCR, University of North Carolina at Chapel
13
Hill, 3300 Doc J. Thurston Building, Campus Box 7280, Chapel Hill, NC 27599-7280
14 15
Emails - EAS: [email protected]; CA: [email protected]; YMG:
16
[email protected]; JBR: [email protected]; JMJ:
17
[email protected]; AEN: [email protected]
1
18
Objective: To describe the incidence and progression of radiographic and symptomatic hand
19
osteoarthritis (rHOA and sxHOA) in a large community-based cohort.
20
Design: Data were from the Johnston County OA Project (1999-2015, 12 ±1.2 years follow-up,
21
age 45+). Participants had bilateral hand radiographs each visit, read for Kellgren-Lawrence
22
grade (KLG) at 30 joints. We defined rHOA as KLG ≥2 in ≥1 joint. SxHOA was defined in a
23
hand/joint with rHOA and self-reported symptoms or tenderness on exam. Incidence was
24
assessed in those without, while progression was assessed in those with, baseline rHOA.
25
Proportions or medians are reported; differences by sex and race were assessed using models
26
appropriate for dichotomous or continuous definitions, additionally adjusted for age, education,
27
BMI, and weight change.
28
Results: Of 800 participants (68% women, 32% African American, mean age 60 years), 327
29
had baseline rHOA and were older, more often white and female, than those without rHOA
30
(n=473). The incidence of HOA was high, for rHOA (60%) and for sxHOA (13%). Women were
31
more likely than men to have incident HOA, particularly for DIP rOA (aOR 1.60 95% CI [1.03,
32
2.49]) and sxHOA (aOR 2.98 [1.50, 5.91]). Progressive HOA was more similar by sex, although
33
thumb base rOA progressed more frequently in women than in men (aOR 2.56 [1.44, 4.55]).
34
Particularly HOA incidence, but also progression, was more frequent among whites compared
35
with African Americans.
36
Conclusion: This study provides much needed information about the natural history of HOA, a
37
common and frequently debilitating condition, in the general population.
38 39
Keywords: Hand osteoarthritis, radiographs, epidemiology
40 41
Running title: Incidence and Progression of Hand OA
42 43 2
44
The presentation of osteoarthritis (OA) varies immensely – affecting adults of all ages and
45
involving different joints throughout the body.1 Although hand osteoarthritis (HOA) is
46
exceedingly common and will affect 4 in 10 people by age 852, this joint site is understudied in
47
comparison to other sites, like the knee. HOA is a heterogeneous disease that presents as
48
bony enlargement and deformities of various joints in the hand.3 Patients may experience pain,
49
stiffness and decreased mobility, which can result in decreased grip strength and functionality of
50
the hand.4 As such, HOA can lead to significant disability and increased healthcare utilization.5
51 52
To date, studies of HOA are limited – particularly with regard to incidence and progression of
53
disease in longitudinal cohorts. Review of the literature reveals that prevalence estimates of
54
HOA in adults vary depending on the population studied and definitions used to define the
55
disease. HOA can be defined radiographically and/or symptomatically, and such definitions can
56
affect frequency estimates.6 For example, the Framingham OA Study estimated the frequency
57
of symptomatic HOA as 13% in men and 26% in women > 70 years of age, but only includes
58
white individuals.7 As the US population is aging, these data suggest more people will be
59
affected by HOA in the future highlighting the need for better understanding of the natural
60
history of this common condition, including differences by sex and race.
61 62
The purpose of this study was to utilize recent longitudinal data to describe the incidence and
63
progression of radiographic and symptomatic HOA (rHOA and sxHOA, respectively) in a large
64
community-based cohort of African American and white men and women.
65 66
Participants and Methods
67
Data were obtained from adult participants (age ≥ 45 years) in the Johnston County
68
Osteoarthritis Project (JoCo OA), an ongoing community-based prospective observational
69
cohort study monitoring the occurrence and natural history of OA in Johnston County, North 3
70
Carolina.8 JoCo OA enrolled non-institutionalized white and African American adults age 45
71
and older who resided in 6 townships in the county in a population-based manner (without
72
regard for symptoms or diagnosis). All participants provide informed consent, answer
73
standardized questionnaires and undergo an examination with trained staff, in addition to
74
radiography of multiple joint sites. This analysis included all participants at the third JoCo OA
75
follow up (n=909, collected 2013-15), and included data from all prior time points with hand data
76
for these individuals (first follow-up: 1999-2003 or cohort enrichment: 2003-4, and second
77
[combined] follow up: 2006-10). After exclusion of 109 participants due to missing radiographs
78
over time (n=84) or missing any of the 30 specific joint scores over time (n=25), 800 participants
79
were included in the complete case analysis (Figure 1).
80
Sex, race, and age were from self-report at baseline. Educational attainment at baseline was
81
categorized as completing less than a high school education vs at least a high school education.
82
Body mass index (BMI) at baseline was computed from measured standing height (cm) and
83
weight (kg) during the research clinic visit. Weight gain from baseline (T1) to second follow-up
84
(T3) was defined as more than a 5% increase in BMI.
85
All participants had posteroanterior radiographs of the bilateral hands at each visit, read
86
together in known time order9, 10 by an experienced musculoskeletal radiologist (JBR) using the
87
Kellgren-Lawrence grading (KLG) system11 for all 30 joints (bilateral distal interphalangeal [DIP]
88
2-5; proximal interphalangeal [PIP] 2-5; metacarpophalangeal [MCP] 1-5; interphalangeal [IP]
89
and carpometacarpal [CMC] of the thumb). The radiologist has previously demonstrated high
90
reliability (kappa >0.7), detailed by joint in a prior paper.2 Erosive OA was not systematically
91
assessed in this study, but noted only as a comment and appeared to be relatively uncommon
92
in this sample (9 participants). Therefore, erosive OA was not analyzed separately, although
93
these participants were included in the analyses for other outcomes.
94
4
95
There are many ways to define rHOA.12 For these analyses, we considered several definitions.
96
First, rHOA overall was defined as a KLG of 2 or more in at least 1 of the 30 hand joints.
97
Additionally, rOA by joint group (DIP, PIP, thumb IP, MCP, or CMC, involving at least one joint
98
in the group) and rHOA affecting 2 or more joints was assessed. We also considered the
99
number of joints affected by KLG of 2 or more and the sum score of KLG (0-4 at each joint)
100
across all 30 joints (range 0-120). SxHOA was defined as the presence of symptoms defined in
101
one of two ways with rOA in the same hand or specific joint site. First, self-reported pain, aching
102
or stiffness in the entire hand was assessed using the answer to “On most days, do you have
103
pain, aching, or stiffness in your hands” at baseline, which was modified at follow up to: “On
104
most days of any one month in the last twelve months, did you have pain, aching, or stiffness in
105
your hands?” at follow up, both with indicators for right or left hand. Second, tenderness to
106
palpation at each of the 30 hand joints during the research exam was assessed (previously
107
shown to be reliable, with agreement >0.9 for two trained readers1). We additionally considered
108
the number of joints with sxHOA.
109 110
Two groups were defined: those with baseline rHOA (KLG of 2 or more in at least 1 of the 30
111
hand joints) and those without. Among those without baseline rHOA, we defined incident rHOA
112
as newly meeting the above criteria in a hand or joint previously without rOA. Among those with
113
baseline rHOA, progression was defined as an additional joint involved, either overall or in a
114
given joint group for both rHOA and sxHOA. We again considered the overall number of rHOA
115
joints and sxHOA joints as well as the total sum KLG for progression.
116 117
Incidence and progression were assessed over an average of 12 years involving a baseline
118
(either first follow up: 1999-2004 or cohort enrichment: 2003-4) and two combined follow-up
119
time-points (2006-2010 and 2013-2015). Descriptive statistics were shown as proportions for
120
dichotomous variables and medians (interquartile range) for continuous variables. 5
121
Most incidence and progression differences between sex and race were assessed by modeling
122
dichotomous definitions with logistic regression and producing adjusted odds ratios (aOR) and
123
95% confidence intervals (95% CI) describing associations. For continuous definitions
124
(consisting of skewed count distributions) either zero-inflated negative binomial, scaled negative
125
binomial or scaled regression models were used to produce adjusted incidence density ratios
126
(aIDR) or adjusted betas (β) and 95% CI describing associations. Models were adjusted for
127
age, sex, race, education, BMI, weight gain of 5% or more, and HOA baseline values (for
128
progression only).
129 130
Results
131
Of the 800 included participants (Figure 1), 68% were women and 32% were African American
132
(Table 1). The mean age of the sample was 59.6 (standard deviation ± 7.5) years, and the
133
mean body mass index (BMI) was 31.0 ± 6.4 kg/m2. Those with baseline rHOA (n=327, 41% of
134
the sample) tended to be older (63 ± 7 vs 57 ± 7 years) and were more often white (80% vs
135
61%) and women (72% vs 65%) than those without baseline rHOA (n=473, 59% of the sample).
136
Socioeconomic status as represented by education level did not differ between groups, nor did
137
the proportion experiencing weight gain over the course of the study (Table 1).
138 139 140
Incident HOA The frequencies of incident rHOA (≥1 or ≥2 joints), DIP rOA, PIP rOA, thumb IP rOA,
141
and thumb base rOA were slightly higher among women, while men were more likely to have
142
incident MCP rOA (Table 2). In adjusted models, women had statistically significantly greater
143
odds of DIP rOA (aOR 1.60, 95% CI [1.03, 2.49]) and had, on average, 1.33 times the number
144
of rHOA affected joints compared with men.
145 146
Incident sxHOA was significantly more likely in women (16% vs 7%), who also had more joints with sxHOA (but not rHOA) compared with men. Both differences were statistically 6
147
significant in adjusted models, where women had nearly 3 times the odds of developing sxHOA
148
and had 4.7 times more joints affected with sxHOA than men. The overall KLG sum score
149
across all joints was similar by sex.
150
All definitions of rHOA and sxHOA were more common in white compared with African
151
American participants, particularly for rHOA (≥1 joint), DIP OA, and thumb base OA (Table 2),
152
all of which were at least twice as likely among whites in adjusted models. SxHOA was more
153
than twice as frequent in white compared to African American participants (17% vs 6%), with an
154
adjusted odds ratio over 4. The number of joints affected by rHOA or sxHOA, and the overall
155
KLG sum, were also higher in white participants, although the adjusted IDR was only significant
156
for number of joints with rHOA (aIDR 2.00 [1.54, 2.61]).
157
Progression of HOA
158
Progressive rHOA was seen at similar frequencies between men and women, except for
159
MCP rOA, which more commonly progressed in men (58% vs 39%), and thumb base rOA,
160
which progressed more often in women (61% vs 41%, Table 3). Only thumb base OA was
161
statistically significant in adjusted models, where women had 2.5 times the odds of progression
162
compared with men. Progressive rHOA and sxHOA were slightly but not statistically significantly
163
more common in women compared with men. Women also had more joints involved with rHOA
164
and sxHOA over time, although this was not significant in adjusted models.
165
All measures of HOA progression were higher among whites compared with African
166
Americans, although most were modestly so (Table 3). Thumb base rOA progression was
167
significantly more frequent among white compared with African American participants (61% vs
168
34%), as was sxHOA progression (32% vs 25%, respectively); both remained significant in
169
adjusted models. Additionally, the number of joints affected and the change in number of joints
170
involved were slightly higher among whites compared with African Americans, although not
171
statistically significant after adjustment.
172 7
173
Discussion
174
HOA remains understudied in comparison to other common sites of OA, like the hip and knee.
175
While several studies have described the prevalence of hand OA in various populations, there
176
have been surprisingly few analyses of the incidence and progression of OA in longitudinal,
177
community-based cohorts. An understanding of the natural history of this common and often
178
debilitating condition is essential to informing further study, including the development of new
179
interventions. In this longitudinal, community-based cohort including African American and white
180
men and women, we found high frequencies of new disease over an average of 12 years,
181
particularly for rHOA (60%) but also for sxHOA (13%). For most joint sites, compared with men,
182
women were somewhat more likely to have incident HOA, although the opposite was true at the
183
MCP. The differences by race were more apparent, with white participants experiencing much
184
higher frequencies of incident HOA by most definitions. HOA progression was similarly
185
common, with rHOA progression occurring in the majority of the sample and sxHOA in about
186
two out of every five (41%). Differences by sex were similar for progression and for incidence,
187
while differences by race were somewhat less pronounced. Of note, more than 1/4 of the
188
sample, regardless of subgroup, had progressive sxHOA with at least one additional joint
189
involved over the follow-up period (28% of African Americans, 35% of men, 43% of whites and
190
44% of women), emphasizing the burden of this condition.
191 192
There are a few large cohort studies to which we can compare our findings. Haugen, et al13,
193
using data from the Framingham OA Study (FOA, collected 1992-5 and 2002-5), reported 9-
194
year cumulative incidence for similar definitions of HOA that were overall lower than those
195
reported here (e.g., rHOA approximately 34%, PIP 5%, although thumb base OA was
196
comparable at 17% in men and 21% in women). However, the pattern of sex differences was
197
similar between the two studies. For progression over 9 years, the overall frequencies were
198
fairly similar to the current report (e.g., rHOA progression in >90%), although DIP progression 8
199
was slightly lower among women in the JoCo OA than in the FOA (74% vs. 78%, respectively)
200
but higher in men, (72% in JoCo OA vs. 59% in FOA). Additionally, PIP and MCP progression
201
were more common in JoCo OA compared with FOA. Thumb base OA progression was more
202
frequent in FOA (<50% in JoCo OA, >60% in FOA), again with similar patterns by sex. Among
203
individuals with hand pain but without baseline rHOA in the Clinical Assessment Studies of the
204
Hand study14, the KLG sum (for 20 joints; no MCP) increased by 9 in women and 7 in men over
205
7 years, not dissimilar to the increase of 12 and 11 in the current analysis over 12 years
206
(although the latter sum was for 30 joints). In the Genetics, Arthrosis and Progression study15, a
207
family study of multiple joint OA, the pattern of progression was similar to the present study, with
208
HOA progression most frequent in the DIPs, then CMC, PIP/IP and MCPs. In a few smaller
209
studies in more specific populations, progression has been shown to be a common occurrence,
210
although often modest in severity16-18. Of course, these are very different populations and
211
utilized somewhat different methodologies so are not directly comparable. The fact that there
212
are not more studies, and that there are none in diverse populations (all of the above studies
213
were among whites), highlights the need for additional work in this area.
214 215
We have previously reported on the cross-sectional frequency of multiple joint OA phenotypes,
216
including HOA, by race and sex in the JoCo OA.19, 20 We found that, compared to whites, African
217
Americans were less likely to have phenotypes that included rHOA20 or sxHOA19, although they
218
were more likely to have phenotypes involving one or more large joints. Additionally, the lifetime
219
risk of sxHOA was lower among African Americans compared with whites in this cohort (29.2%
220
vs. 41.4%).2 A large cross-sectional study in young (age 42-53) African American and white
221
women nearly two decades ago found a similar prevalence of hand OA by race, although MCP
222
OA was more common among African American (12%) compared with white (3%) women;
223
African American women were also more likely to have both hand and knee OA (8% vs 2% of
224
white women)21. Unfortunately, we are not aware of any other community-based studies 9
225
including African Americans that are available for comparison, and the mechanisms behind
226
these differences are not known.
227 228
Limitations of this work include small numbers in some categories which precluded more
229
detailed modeling. As in most large cohorts, we have relatively infrequent assessments of
230
symptoms, and are not able to model detailed fluctuations in symptoms over time. We focused
231
on incidence and progression and did not separately analyze improvements in symptoms over
232
time. However, the strengths, including the large overall sample size, the unique biracial cohort
233
including both men and women, and the community-based nature of the cohort, make this an
234
important contribution to the HOA literature.
235 236
In this large, longitudinal, community-based cohort inclusive of African American and white men
237
and women, we found overall high frequencies of both incidence and progression of HOA over
238
an average of 12 years of follow up. Although common across the board, the frequencies of
239
incident and progressive HOA by a variety of definitions were generally higher in women
240
compared with men and among whites compared with African Americans; differences were
241
overall more marked for definitions of incidence than for progression. Additional studies are
242
needed to understand potential mechanisms for the identified differences (e.g., occupational,
243
genetic, nutritional, biomechanical, or other factors).
244 245
10
246
Acknowledgments: We would like to thank the staff and participants in the Johnston County
247
Osteoarthritis Project, without whom this work would not be possible.
248
Author contributions:
249
Conception and design: EAS, CA, YMG, AEN
250
Acquisition of data: JBR, JMJ
251
Analysis and interpretation of data: EAS, CA, YMG, JBR, JMJ, AEN
252
Drafting the article or revising it critically for important intellectual content: EAS, CA, YMG, JBR,
253
JMJ, AEN
254
Final approval of the version to be submitted: EAS, CA, YMG, JBR, JMJ, AEN
255
Dr. Nelson takes responsibility for the integrity of the work as a whole.
256
Role of the funding source:
257
This work is funded in part by: CDC U01 DP006266 and DP003206, NIH/NIAMS P60
258
AR049465 and AR064166. The funding agencies had no role in the design, analysis, or
259
interpretation of data, nor in the writing of the manuscript or the decision to submit it for
260
publication.
261
Competing interest statement:
262
None of the authors has any competing interests in relation to this work.
263 264 265 266 267 268
Figure Legend Figure 1. Flow chart of participant inclusion in the current analysis
11
269 270 271
References 1.
Gullo TR, Golightly YM, Cleveland RJ, Renner JB, Callahan LF, Jordan JM, et al.
272
Defining multiple joint osteoarthritis, its frequency and impact in a community-based
273
cohort. Semin Arthritis Rheum 2019;48:950-957. doi:10.1016/j.semarthrit.2018.10.001.
274
2.
Qin J, Barbour KE, Murphy LB, Nelson AE, Schwartz TA, Helmick CG, et al. Lifetime
275
Risk of Symptomatic Hand Osteoarthritis: The Johnston County Osteoarthritis Project.
276
Arthritis Rheumatol 2017;69:1204-1212. doi:10.1002/art.40097.
277
3.
Zhang W, Doherty M, Leeb BF, Alekseeva L, Arden NK, Bijlsma JW, et al. EULAR
278
evidence-based recommendations for the diagnosis of hand osteoarthritis: report of a
279
task force of ESCISIT. Ann Rheum Dis 2009;68:8-17. doi:10.1136/ard.2007.084772.
280
4.
Michon M, Maheu E, Berenbaum F. Assessing health-related quality of life in hand
281
osteoarthritis: a literature review. Ann Rheum Dis 2011;70:921-928.
282
doi:10.1136/ard.2010.131151.
283
5.
Oliveria SA, Felson DT, Reed JI, Cirillo PA, Walker AM. Incidence of symptomatic hand,
284
hip, and knee osteoarthritis among patients in a health maintenance organization.
285
Arthritis Rheum 1995;38:1134-1141.
286
6.
Pereira D, Peleteiro B, Araujo J, Branco J, Santos RA, Ramos E. The effect of
287
osteoarthritis definition on prevalence and incidence estimates: a systematic review.
288
Osteoarthritis Cartilage 2011;19:1270-1285. doi:10.1016/j.joca.2011.08.009.
289
7.
Zhang Y, Niu J, Kelly-Hayes M, Chaisson CE, Aliabadi P, Felson DT. Prevalence of
290
symptomatic hand osteoarthritis and its impact on functional status among the elderly:
291
The Framingham Study. Am J Epidemiol 2002;156:1021-1027.
292
8.
Jordan JM, Helmick CG, Renner JB, Luta G, Dragomir AD, Woodard J, et al. Prevalence
293
of knee symptoms and radiographic and symptomatic knee osteoarthritis in African
294
Americans and Caucasians: the Johnston County Osteoarthritis Project. J Rheumatol
295
2007;34:172-180. 12
296
9.
Felson DT, Nevitt MC. Blinding images to sequence in osteoarthritis: evidence from
297
other diseases. Osteoarthritis Cartilage 2009;17:281-283.
298
doi:10.1016/j.joca.2008.09.008.
299
10.
Gensburger D, Roux JP, Arlot M, Sornay-Rendu E, Ravaud P, Chapurlat R. Influence of
300
blinding sequence of radiographs on the reproducibility and sensitivity to change of joint
301
space width measurement in knee osteoarthritis. Arthritis Care Res (Hoboken)
302
2010;62:1699-1705. doi:10.1002/acr.20311.
303
11.
304 305
Kellgren JH, Lawrence JS. Radiological assessment of osteo-arthrosis. Ann Rheum Dis 1957;16:494-502. doi:10.1136/ard.16.4.494.
12.
Marshall M, Dziedzic KS, van der Windt DA, Hay EM. A systematic search and narrative
306
review of radiographic definitions of hand osteoarthritis in population-based studies.
307
Osteoarthritis Cartilage 2008;16:219-226. doi:10.1016/j.joca.2007.06.004.
308
13.
Haugen IK, Englund M, Aliabadi P, Niu J, Clancy M, Kvien TK, et al. Prevalence,
309
incidence and progression of hand osteoarthritis in the general population: the
310
Framingham Osteoarthritis Study. Ann Rheum Dis 2011;70:1581-1586.
311
doi:10.1136/ard.2011.150078.
312
14.
Marshall M, Peat G, Nicholls E, Myers HL, Mamas MA, van der Windt DA. Metabolic risk
313
factors and the incidence and progression of radiographic hand osteoarthritis: a
314
population-based cohort study. Scand J Rheumatol 2019;48:52-63.
315
doi:10.1080/03009742.2018.1459831.
316
15.
Bijsterbosch J, Meulenbelt I, Watt I, Rosendaal FR, Huizinga TW, Kloppenburg M.
317
Clustering of hand osteoarthritis progression and its relationship to progression of
318
osteoarthritis at the knee. Ann Rheum Dis 2014;73:567-572. doi:10.1136/annrheumdis-
319
2012-202461.
13
320
16.
Paradowski PT, Lohmander LS, Englund M. Natural history of radiographic features of
321
hand osteoarthritis over 10 years. Osteoarthritis Cartilage 2010;18:917-922.
322
doi:10.1016/j.joca.2010.04.008.
323
17.
Balblanc JC, Mathieu P, Mathieu L, Tron AM, Conrozier T, Piperno M, et al. Progression
324
of digital osteoarthritis: a sequential scintigraphic and radiographic study. Osteoarthritis
325
Cartilage 1995;3:181-186. doi:10.1016/s1063-4584(05)80052-4.
326
18.
Harris PA, Hart DJ, Dacre JE, Huskisson EC, Spector TD. The progression of
327
radiological hand osteoarthritis over ten years: a clinical follow-up study. Osteoarthritis
328
Cartilage 1994;2:247-252. doi:10.1016/s1063-4584(05)80076-7.
329
19.
Nelson AE, Golightly YM, Renner JB, Schwartz TA, Kraus VB, Helmick CG, et al. Brief
330
report: differences in multijoint symptomatic osteoarthritis phenotypes by race and sex:
331
the Johnston County Osteoarthritis Project. Arthritis Rheum 2013;65:373-377.
332
doi:10.1002/art.37775.
333
20.
Nelson AE, Renner JB, Schwartz TA, Kraus VB, Helmick CG, Jordan JM. Differences in
334
multijoint radiographic osteoarthritis phenotypes among African Americans and
335
Caucasians: The Johnston County Osteoarthritis project. Arthritis Rheum 2011;63:3843-
336
3852. doi:10.1002/art.30610.
337
21.
Sowers M, Lachance L, Hochberg M, Jamadar D. Radiographically defined osteoarthritis
338
of the hand and knee in young and middle-aged African American and Caucasian
339
women. Osteoarthritis Cartilage 2000;8:69-77. doi:10.1053/joca.1999.0273.
340 341
14
Table 1. Descriptive characteristics of the sample overall and by baseline hand OA status Sample
Overall (n=800)
No baseline hand OA
Baseline hand OA
(n=473)
(n=327)
mean (SD) or %
mean (SD) or %
mean (SD) or %
Age, years
59.6 (7.5)
57.3 (6.6)
63.1 (7.4)
Female, %
68.0
65.3
71.9
African American, %
31.5
39.5
19.9
<12 years education, %
13.8
13.8
13.8
31.0 (6.4)
31.6 (6.7)
30.2 (5.9)
32.6
33.4
31.5
Characteristics
Baseline body mass index, kg/m2 >5% weight gain from baseline to follow up, %
15
Table 2. Proportions by hand OA incidence definition over ~ 12 (±1.2) year follow-up for those without baseline hand OA, by sex or race Hand OA incidence definition
Overall
Women
Men
(n=473)
(n=309)
(n=164)
Women vs Men*
White
African
(n=286)
American
White vs AA*
(n=187) % or median
effect (95% CI)
% or median
(interquartile
(interquartile
range: Q1-Q3)
range: Q1-Q3)
effect (95% CI)
59.4%
60.5%
57.3%
1.26 (0.84, 1.90)
65.0%
50.8%
2.26 (1.49, 3.43)
– 5 ) OA
32.1%
34.0%
28.7%
1.60 (1.03, 2.49)
40.2%
19.8%
3.16 (1.97, 5.07)
PIP (2nd – 5th) OA
20.1%
21.4%
17.7%
1.24 (0.82, 1.87)
22.7%
16.0%
1.17 (0.77, 1.77)
Thumb IP OA
24.5%
24.9%
23.8%
1.02 (0.64, 1.61)
23.1%
26.7%
0.87 (0.55, 1.37)
MCP (1 – 5 ) OA
13.5%
11.7%
17.1%
0.59 (0.34, 1.04)
13.6%
13.4%
1.08 (0.60, 1.95)
Thumb base OA
20.7%
22.3%
17.7%
1.59 (0.96, 2.64)
24.1%
15.5%
2.11 (1.24, 3.58)
Radiographic HOA (≥2 joint(s))
31.7%
33.0%
29.3%
1.18 (0.78, 1.79)
35.3%
26.2%
1.60 (1.05, 2.45)
Symptomatic HOA (≥1 joint(s))
12.9%
15.9%
7.3%
2.98 (1.50, 5.91)
17.1%
6.4%
4.29 (2.08, 8.86)
Number of radiographic HOA
1 (0-3)
1 (0-3)
1 (0-2)
1.33 (1.03, 1.72)
1
1 (0-3)
1 (0-2)
2.00 (1.54, 2.61)
1
0 (0-0)
0 (0-0)
0 (0-0)
4.74 (2.08, 10.8)
1
0 (0-0)
0 (0-0)
1.69 (0.41, 7.06)
1
Max=93
Max=9
Max=2
Max=9
Max=3
At T3:
At T3:
At T3:
At T3:
At T3:
Radiographic HOA (≥1 joint(s)) DIP (2
nd
st
th
th
joints Number of symptomatic HOA joints Total sum KL score (30 joints)
0.99 (0.95, 1.03)2
1.00 (0.96, 1.04)2
16
11 (7-19)
12 (8-19)
11 (7-18)
12 (8-20)
11 (7-16)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
7 (4-12)
7 (4-12)
8 (4-11)
8 (5-14)
6 (3-9)
Distal interphalangeal joint (DIP), Hand osteoarthritis (HOA), Interphalangeal joint (IP), Kellgren-Lawrence grade (KLG), Metacarpophalangeal joint (MCP), Osteoarthritis (OA), Proximal interphalangeal joint (PIP); ∆=change from T1 to T3 (baseline to follow up) Effects shown are adjusted for baseline: age, sex, race, education, BMI, and weight gain—additionally adjusted for HOA baseline values for total sum KL score. *Logit models for categorical definitions producing adjusted odds ratios (aOR) and 1 zero inflated negative binomial models for continuous definitions producing adjusted incidence density ratios (aIDR) and 2 scaled negative binomial model for continuous definitions producing adjusted incidence density ratios (aIDR)
17
Table 3. Proportion by hand OA progression definition over ~ 12 (±1.2) year follow-up for those with baseline hand OA, by sex or race Hand OA progression definition
Overall
Women
Men
(n=327)
(n=235)
(n=92)
Women vs Men*
White
African
(n=262)
American
White vs AA*
(n=65) % or median
effect (95% CI)
(interquartile
effect (95% CI)
(interquartile
range: Q1-Q3) Radiographic HOA
% or median
range: Q1-Q3)
94.2%
94.0%
94.6%
0.70 (0.22, 2.22)
95.4%
89.2%
1.15 (0.38, 3.45)
n=306
n=218
n=88
1.26 (0.67, 2.37)
n=242
n=64
1.39 (0.70, 2.76)
77.8%
79.4%
73.9%
81.0%
65.6%
n=318
n=228
n=90
n=254
n=64
63.5%
66.2%
56.7%
65.0%
57.8%
n=286
n=205
n=81
n=228
n=58
56.3%
54.2%
61.7%
57.9%
50.0%
MCP (1 – 5 ) OA
44.0%
38.7%
57.6%
0.61 (0.36, 1.05)
44.7%
41.5%
0.96 (0.52, 1.77)
Thumb base OA
n=274
n=193
n=81
2.56 (1.44, 4.55)
n=215
n=59
3.39 (1.74, 6.60)
54.7%
60.6%
40.7%
60.5%
33.9%
80.4%
81.3%
78.3%
82.4%
72.3%
(additional in ≥1 joint(s)) nd
th
DIP (2 – 5 ) OA
nd
th
PIP (2 – 5 ) OA
Thumb IP OA
st
th
Radiographic HOA
1.38 (0.80, 2.37)
0.80 (0.46, 1.37)
1.11 (0.59, 2.10)
0.94 (0.51, 1.76)
1.30 (0.71, 2.40)
1.08 (0.53, 2.18)
18
(additional ≥2 joint(s))
Symptomatic HOA
31.0%
31.9%
28.6%
1.25 (0.72, 2.16)
32.4%
25.0%
At T3:
At T3:
At T3:
1
At T3:
At T3:
9 (4-14)
9 (5-15)
8 (4-12)
10 (5-15)
6 (3-10)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
5 (2-8)
6 (3-9)
5 (2-7)
6 (3-8)
4 (2-8)
Number of symptomatic
At T3:
At T3:
At T3:
At T3:
At T3:
HOA joints
0 (0-1)
0 (0-2)
0 (0-1)
0 (0-1)
0 (0-0)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
0 (0-1)
0 (0-1)
0 (0-0)
0 (0-1)
0 (0-0)
At T3:
At T3:
At T3:
At T3:
At T3:
34 (21-45)
35 (22-47)
32.5 (19-42)
35 (23-47)
27 (17-37)
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
∆ T3-T1:
16 (9-22)
16 (10-23)
14.5 (8.5-20.5)
17 (10-22)
13 (8-18)
1.96 (1.01, 3.83)
(additional ≥1 joint(s)) Number of radiographic HOA joints
Total sum KL score (30 joints)
1.17 (0.99, 1.39)
0.23 (-0.34, 0.81)
1.09 (0.97, 1.24)
2
1
1.12 (0.92, 1.36)
1
0.46 (-0.21, 1.12)
1.09 (0.94, 1.26)
2
1
Distal interphalangeal joint (DIP), Hand osteoarthritis (HOA), Interphalangeal joint (IP), Kellgren-Lawrence grade (KLG), Metacarpophalangeal joint (MCP), Osteoarthritis (OA), Proximal interphalangeal joint (PIP); ∆=change from T1 to T3 (baseline to follow up) Effects shown are adjusted for baseline: age, sex, race, education, BMI, weight gain, and HOA baseline values. *Logit models for categorical definitions producing adjusted odds ratios (aOR) and 1 scaled negative binomial model for continuous definitions producing adjusted incidence density ratios (aIDR) and 2 scaled regression models for the number of symptomatic HOA joints producing adjusted betas.
19
STROBE Statement—Checklist of items that should be included in reports of cohort studies
Title and abstract
Item No 1
Line or page # in ms Recommendation (a) Indicate the study’s design with a commonly used term in the
Line 1
title or the abstract (b) Provide in the abstract an informative and balanced summary of
Page 2
what was done and what was found Introduction Background/rationale
2
Explain the scientific background and rationale for the
Page 3
investigation being reported Objectives
3
State specific objectives, including any prespecified hypotheses
Methods Study design
4
Present key elements of study design early in the paper
Setting
5
Describe the setting, locations, and relevant dates, including
Lines 68-70 Page 3-5 Lines 73-82
periods of recruitment, exposure, follow-up, and data collection Participants
6
(a) Give the eligibility criteria, and the sources and methods of
Lines 73-82 and 109-
selection of participants. Describe methods of follow-up
115
(b) For matched studies, give matching criteria and number of
n/a
exposed and unexposed Variables
7
Clearly define all outcomes, exposures, predictors, potential confounders, and effect modifiers. Give diagnostic criteria, if
Lines 84-107
applicable Data sources/
8*
measurement
For each variable of interest, give sources of data and details of
Lines 84-115
methods of assessment (measurement). Describe comparability of assessment methods if there is more than one group
Bias
9
Describe any efforts to address potential sources of bias
Study size
10
Explain how the study size was arrived at
Lines 78-82
Quantitative
11
Explain how quantitative variables were handled in the analyses. If
Lines 92-107
variables Statistical methods
n/a (descriptive)
applicable, describe which groupings were chosen and why 12
(a) Describe all statistical methods, including those used to control
Lines 109-115
for confounding (b) Describe any methods used to examine subgroups and
Lines 109-115
interactions (c) Explain how missing data were addressed (d) If applicable, explain how loss to follow-up was addressed (e) Describe any sensitivity analyses Results Participants
13*
(a) Report numbers of individuals at each stage of study—eg numbers potentially eligible, examined for eligibility, confirmed
Line 82 Line 77-82 n/a Lines 77-82 and 118120
eligible, included in the study, completing follow-up, and analysed (b) Give reasons for non-participation at each stage (c) Consider use of a flow diagram
Lines 77-82 Unable due to journal constraints
Descriptive data
14*
(a) Give characteristics of study participants (eg demographic,
Lines 118-122
clinical, social) and information on exposures and potential confounders (b) Indicate number of participants with missing data for each variable of interest
Complete case analysis, line 82
1
(c) Summarise follow-up time (eg, average and total amount)
Lines 109-111 and
Report numbers of outcome events or summary measures over time
Lines 120-122 and
Tables 1&2 Outcome data
15*
Tables 1&2 Main results
16
(a) Give unadjusted estimates and, if applicable, confounder-
Tables 1 & 2
adjusted estimates and their precision (eg, 95% confidence interval). Make clear which confounders were adjusted for and why they were included (b) Report category boundaries when continuous variables were
n/a
categorized (c) If relevant, consider translating estimates of relative risk into
n/a
absolute risk for a meaningful time period Other analyses
17
Report other analyses done—eg analyses of subgroups and interactions, and sensitivity analyses
Pages 5-6 and Tables 1&2
Discussion Key results
18
Summarise key results with reference to study objectives
Lines 154-166
Limitations
19
Discuss limitations of the study, taking into account sources of
Lines 197-200
potential bias or imprecision. Discuss both direction and magnitude of any potential bias Interpretation
20
Give a cautious overall interpretation of results considering
Lines 202-208
objectives, limitations, multiplicity of analyses, results from similar studies, and other relevant evidence Generalisability
21
Discuss the generalisability (external validity) of the study results
Other information Funding
22
Give the source of funding and the role of the funders for the
Lines 168-187 Lines 222-5
present study and, if applicable, for the original study on which the present article is based *Give information separately for exposed and unexposed groups. Note: An Explanation and Elaboration article discusses each checklist item and gives methodological background and published examples of transparent reporting. The STROBE checklist is best used in conjunction with this article (freely available on the Web sites of PLoS Medicine at http://www.plosmedicine.org/, Annals of Internal Medicine at http://www.annals.org/, and Epidemiology at http://www.epidem.com/). Information on the STROBE Initiative is available at http://www.strobe-statement.org.
2
Inclusions N=3085
Exclusions
JoCoOA Project T1/T1* clinic participants (baseline, 1999-2004) No hand X-ray films read (n=1559)*
N=1526
JoCoOA Project T1/T1* clinic participants (1999-2004) with hand X-rays Did not participate at T3 (n=1071) • Lost to follow-up (n=616) • Deceased (n=455)
N=908
JoCoOA Project T3 clinic participants (follow up, 2013-2015) Missing complete x-ray information (n=109) • 83 missing x-rays • 25 missing specific joint scores
N=800
Available KL grades for all 30 hand joints at baseline and follow up
*Based on priority, hand x-rays for JoCo OA participants without follow-up have not yet been read.