Table 2.

Characteristics at 10 years and median change from baseline, stratified by cohort

	Cohort 1 (1990–1994)		Cohort 2 (2000–2004)		Median difference/OR/relative difference (95% CI)*	Median difference/OR/ relative difference (95% CI)†
	N	Median (IQR)	N	Median (IQR)
Age at 10-year assessment (years)	607	62 (51–72)	340	66 (56–75)	4.00 (1.55 to 6.45)	–
Gender (n (%) female)	422 (69.5)		238 (70.0)		OR 1.02 (0.77 to 1.37)	–
Swollen joint counts (51)	601	1 (0–4)	340	0.5 (0.0–2.5)	−36% (−49% to −20%)	−33% (−47% to −15%)
Median change from baseline	601	−4 (−10 –−1)	340	−2 (−6– 0)
Tender joints counts (51)	601	2 (0–11)	340	2 (0–8)	−5% (−24% to 19%)	2% (−20% to 30%)
Median change from baseline	601	−3 (−11–1)	340	−1 (−6–2)
HAQ	597	0.88 (0.25–1.63)	336	1.00 (0.25–1.88)	0.01 (−0.18 to 0.20)	−0.01 (−0.16 to 0.14)
Median change from baseline	588	0.13 (−0.25–0.75)	330	0.00 (−0.38–0.63)
Current sDMARDs use, n (%)	195 (32.1)		209 (61.5)		OR 3.35 (2.53 to 4.43)	OR 2.71 (1.91 to 3.86)

*Quantile/logistic/negative binomial regression was used to compare the two cohorts on each variable depending on the type of variable. Regressions comparing SJC, TJC, HAQ and current DMARD use between cohorts controlled for age and gender. Cohort 1 is the reference category.

†Quantile/logistic/negative binomial regression used to compare SJC, TJC, HAQ and number on sDMARDs between cohorts at 10 years. These models controlled for baseline: age, gender, symptom duration before baseline, smoking status, SJC (51), TJC (51), RF, anti-CCP, CRP, HAQ score and being on sDMARDs/steroids. Cohort 1 is the reference category.

anti-CCP2, anticyclic citrullinated protein antibodies; CRP, C reactive protein; DMARD, disease-modifying antirheumatic drug; HAQ, Health Assessment Questionnaire; n, number of patients with available data; sDMARD, synthetic disease-modifying antirheumatic drug; SJC, swollen joint count; TJC, tender joint count.