The data from 52 participants with psoriatic arthritis (Study Group) and 50 participants without the disease (Control Group) were analyzed. A total of 1016 nails were analyzed, 517 from the psoriatic arthritis group and 499 from the controls; 4 nails were excluded due to trauma.

The results indicated a significant difference between the groups regarding age (p = 0.012); on average, the age of the patient group was 7 years older than that of the control group, 56.1 ± 12.2 (23–84) versus 49.0 ± 15.6 (22–77) years, respectively. The groups were homogeneous, with no significant differences between patients and controls regarding the percentages of individuals who were male (59.6% versus 48%), Caucasian (84.6% versus 92%) and manual laborers (35.4% versus 34%, respectively) (Table 1).

Table 1 Demographic characteristics of the control and study groups

Among the participants in the psoriatic arthritis group, 57.7% had peripheral arthritis (and all these patients had distal interphalangeal involvement) without associated axial involvement (defined by the care team as an inflammatory axial clinical picture with or without image alteration). Among the 52 patients, all had a personal history of or current psoriasis, and the most common form was vulgaris (46.15%), which presented alone. There was no description of the type of psoriasis in the medical records of 2 patients. Only 1 patient had isolated nail psoriasis (1.92%), and in the other 17.31%, clinical involvement of the nail was associated with psoriasis vulgaris or scalp or inverted psoriasis (9 patients), as shown in Table 2.

Table 2 Descriptive characteristics of the group of patients

Regarding lifestyle, 54.2% of the patients with psoriatic arthritis described having a sedentary lifestyle, 12.2% had a history of previous smoking (≥ 3 months), and only 2% were current active smokers. Most psoriasis patients had at least one cardiovascular risk factor (82.4%), including arterial hypertension, dyslipidemia, obesity, previous cardiovascular events, and diabetes mellitus, the latter present in 31.4% of the patients. Regarding treatment, few patients were using constant-dose anti-inflammatory drugs (3 patients), while 62% were using conventional disease-modifying antirheumatic drugs (DMARDs), and 76.5% were using biological DMARDs; of these, 53.8% were using TNF-alpha inhibitors. Other clinical characteristics of the participants with psoriatic arthritis are presented in Table 3.

Table 3 Clinical features of participants with psoriatic arthritis

Regarding disease activity, 37.3% achieved remission according to the MDA score, while the mean and median DAPSA scores were 12.69 ± 13.11 (0.10–86.06) and 10, respectively, indicating low disease activity. Regarding cutaneous involvement, most patients had low scores, with a BSA 2.01% ± 3.18 (0–18) and PASI 1.74 ± 2.77 (0–13.70), while nail involvement was moderate according to the NAPSI (mean 16.10 ± 13.92 (0–50), median 13) (Supplementary Material).

In the individual evaluations for each finger, according to the Wortsman et al. [18] classification, the finger whose nail plate was most frequently affected was the right thumb (44.2%), followed by the left thumb (36.5%) and the second finger on the right hand (32.7%) (full data available in the Supplementary Material); for patients with psoriatic arthritis THE FINGER SCORE showed the most affected were the right thumb, left thumb, second and third fingers of the right hand (Table 4 and Graphic 1).

Table 4 Ultrasound score of each finger (quirodactyl) in patient groupGraphic 1

Average ultrasound score of each finger of patients in the psoriatic arthritis group. Legend: qd– quirodactyl (FINGER)

Score analysisComparison of the group scores

The score for each finger was calculated as the sum of the number of points obtained in grayscale imaging according to Wortsman et al. [18] of the finger nail (FN), PD score of the bed and matrix, presence of enthesitis, presence of paratendinitis, and DIJ synovitis score according to grayscale imaging and power Doppler.

For the score of each finger and for the mean score of the 10 fingers, the results were different between groups, as shown in Table 5.

Table 5 Nail score of each finger and the mean score of the fingersDetermination of a cutoff point for the mean score (ROC curve)

To determine a cutoff point for the mean score in identifying the disease, ROC curve analysis was performed. The area under the curve was equal to 0.96 (p < 0.001), indicating that the mean score differentiated between the disease and control groups well. The cutoff point obtained from curve fitting was equal to 0.15. Thus, mean score values > 0.15 correspond to the presence of the disease, and mean score values ≤ 0.15 correspond to the absence of the disease. The estimated sensitivity for this cutoff point is 90.4%, and the specificity is 92.0%, as shown in Fig. 2.

Fig. 2

Roc curve for the presence of psoriatic arthritis. Notes

Average score

Groups

Control

Study (patients)

N

%

N

%

≤ 0,15

46

92%

5

9,6%

> 0,15

4

8%

47

90,4%

Total

50

100%

52

100%

Analysis of variables related to ultrasonographyComparison of groups in relation to quantitative variables

The number of fingers showing grayscale changes according to the Wortsman et al. [18] classification, as shown in Fig. 3, was statistically higher in the patient group, both for major changes—grades 3 or 4—and for any grade. The number of fingers with enthesitis, paratendinitis or synovitis according to grayscale imaging was different between the two groups; participants with psoriatic arthritis had more affected fingers than the control participants. Participants with psoriatic arthritis also had a higher mean power Doppler signal in the nail bed and matrix compared to controls; an example positive power Doppler image shown in Fig. 4. On the other hand, the power Doppler signal of the joints and the distance from the nail bed (1.79 mm patients versus 1.67 mm control; p = 0.073) showed no differences between the groups (Table 6).

Fig. 3

Images with changes in the nail plate according to Wortsman (2004). Notes (A) Grayscale (GS 0) and measuring of the distance from the nail bed with 1.1 mm and 1.3 mm, respectively. (B) Grayscale (GS 1). (C) Grayscale (GS 2) and measuring of the distance from the nail bed with 2.5 mm. (D) Grayscale (GS 3). (E) Grayscale (GS 4)

Fig. 4

Power doppler imaging and calculation of the RI (resistivity index) of the nail bed. Notes Power doppler of the nail bed grade 1. Vessel resistivity index 0.70

Table 6 Descriptive statistics of the variables according to the groups

The frequency of involvement according to the presence of a power Doppler signal of the bed and matrix was higher in the psoriatic arthritis group than in the control group (44.2% versus 6%), and none of the calculated resistivity index values were < 0.4 (available in the Supplementary Material). None of the fingers presented with synovitis on power Doppler in the distal interphalangeal joint in either group.

Comparison of categorical variables between groups

The groups were significantly different (p < 0.001) regarding the presence of nail plate alterations, enthesitis, paratendinitis, grayscale distal interphalangeal synovitis and the DIP involvement, all of which were more frequent in the group of psoriatic patients (Table 7).

Table 7 Descriptive statistics of the categorical variables according to the groupsAssociation of MDA and DAPSA with the variables ultrasound

There was no difference for patients with and without MDA and the results of the nail ultrasound findings (available in Supplementary Material). For the analysis with classifications DAPSA: ≤ 4 (disease in remission), > 4 and ≤ 14 (low activity), > 14 and ≤ 28 (moderate activity) and > 28 (high activity) the findings showed that there was no difference between the groups. For the categorical variables, it was not possible to perform the statistical test due to the low frequencies of the findings (available in Supplementary Material).

If two classifications of DAPSA were considered: ≤ 14 (remission or low activity) and > 14 (moderate or high activity) significant differences were found between the groups in terms of the number of fingers with the highest degree of nail plate alteration (type 3 or 4), the average nail plate alteration of all the patient’s fingers and the average score of the 10 fingers. All these parameters were significantly higher among patients with DAPSA > 14 (Table 8). Fifteen patients had DAPSA corresponding to moderate or high activity, and 37 patients had DAPSA corresponding to remission or low activity.

Table 8 Descriptive statistics of the variables according to the dichotomous DAPSA groups

Only a numerical difference was identified between the number of fingers with nail plate changes between patients with moderate and high DAPSA versus those with remission and low DAPSA. There was no significant difference in the presence of nail plate changes, enthesitis, paratendinitis, Doppler or grayscale DIJ synovitis, or DIP involvement between patients with remission/low DAPSA and those with moderate/high DAPSA. Due to the low frequency of cases, it was not possible to apply statistical tests (Supplementary Material).

Correlation between quantitative clinical variables and quantitative us variables (restricted to patients)

In the evaluation between the clinical scores of activity in psoriatic arthritis versus the US quantitative variables of the nail, the NAPSI and ASDAS-CRP showed the highest estimated correlation coefficients (Tables 9 and 10).

Table 9 NAPSI and nail ultrasound findings. The degree of association can be classified as: excellent:|r| >0.90; Good|r| from 0.71 to 0.90; Moderate:|r| from 0.51 to 0.70; Weak|r| from 0.31 to 0.50 Adapted from Mukaka (2012)Table 10 ASDAS-CRP and nail ultrasound findings. The degree of association can be classified as: excellent:|r| >0.90; Good|r| from 0.71 to 0.90; Moderate:|r| from 0.51 to 0.70; Weak|r| from 0.31 to 0.50 Adapted from Mukaka (2012)

The NAPSI showed a weak correlation with the number of fingers with nail plate changes of any classification (1 to 4) or with major changes in the nail plate (grades 3 or 4), as well as with the means of the laminar change score for each patient or the mean of THE FINGER SCORE of the 10 fingers, which took into account nail plate alteration, presence of enthesitis in the distal digit extensor, presence of paratendinitis, grayscale and Doppler synovitis of the distal interphalangeal joint and power Doppler score for the nail bed or matrix as shown in Table 9.

The ASDAS-CRP showed a moderate correlation with the number of fingers with nail plate changes, the number of fingers with more severe nail plate changes (3 or 4) and the mean nail change, and the mean of THE FINGER SCORE of the 10 fingers as shown in Table 10.

The BSA and BASDAI showed no correlations with the nail ultrasound findings. The other correlations found were a weak correlation between DAPSA and the number of fingers with more severe alterations (grades 3 or 4) of the nail plate (Spearman correlation coefficient 0,3; p = 0,032), and between the PASI and the number of fingers with paratendinitis (Spearman correlation coefficient 0,41; p = 0,003). The corresponding tables and graphics are found in the Supplementary Material.

Analysis of the agreement of the two examinersAgreement between the two examiners in the presence of a finger with GS grade 1, 2, 3, 4, PD > 0, enthesitis, paratendinitis, synovitis, grayscale > 0, DIP involvement (binary variables) and quantitative variables related to US

This analysis was not performed for power Doppler synovitis because no patient or control presented with the condition. To assess the level of agreement of the two examiners, kappa coefficients of agreement were estimated for the binary variables, and intraclass correlation coefficients (ICCs) for the quantitative variables, and the 95% confidence intervals (95% CI) were calculated. The results were good for both of them (kappa > = 0.85 and ICC > = 0.904) (Supplementary Material).