A total of 169 subjects were enrolled in the study. 80% were women and 20% were men. Nearly half of the sample was diagnosed with DLE (46%), 25% with SCLE, 9% with LET, 7% with ACLE, 7% with SLE and nonspecific skin manifestations, 4% had >1 CLE subtype, and 2% had other forms of chronic CLE ([other CCLE], including panniculitis or chilblains). The >1 CLE subtype category was comprised of 4 subjects with SCLE and DLE and one of each with SCLE and ACLE, SCLE and LET, and DLE and panniculitis. Two-thirds of subjects were Caucasian, 28% were African American, 5% were Asian, and 1 subject was Hispanic ().
Prevalence of photosensitivity
Of the 169 subjects, 91 had both a history of and ongoing photosensitivity, 11 had a history of photosensitivity, 13 had new onset photosensitivity, 40 did not experience photosensitivity, and in 14 subjects we were unable to determine photosensitivity status because of incomplete information. Overall, 68% of the sample reported symptoms of photosensitivity while only 24% of subjects denied photosensitivity. Approximately 82% of ACLE, 76% of SCLE, 71% of >1 CLE subtype, 63% of LET, 58% of SLE with nonspecific skin manifestations, 46% of DLE, and 0% of other CCLE patients (panniculitis or chilblains) reported both a history of and current photosensitivity (and were classified into PS group) ().
Prevalence of photosensitivity within each LE type
PS group characteristics
Of the 91 subjects comprising the PS group, 35% had DLE, 31% had SCLE, 11% had LET, 10% had ACLE, 8% had SLE with nonspecific skin manifestations, and 5% had >1 CLE subtype. There was a significant association between two lupus subtypes (DLE, SCLE) and photosensitivity grouping (Pearson’s X2 = 17.92; p = 0.006). SCLE patients were more likely to be categorized into the PS group while DLE patients were more likely to be classified into the NOT PS group. There was no significant relationship between any of the other LE diagnoses (ACLE, LET, SLE with nonspecific skin manifestations, >1 CLE subtype, and other CCLE) and photosensitivity grouping.
Sun exposure-related characteristics in PS group vs. NOT PS group
The PS group developed lesions in sun-exposed areas significantly more frequently than the NOT PS group (71% vs 31%, X2 = 5.88, p=0.015). The PS group engaged in sun protective behaviors significantly more frequently than the NOT PS group (X2 = 16.63, p = 0.002).
Photosensitivity items of the Modified Skindex-29+3
ANOVA showed that the effect of photosensitivity was significant for Item 31 (F=19.51, p=0.000), Item 33 (F=16.18, p=0.000), and the PS scale (F=21.58, p=0.000). The PS group scored significantly worse on Item 31: I worry about going outside because the sun might flare my disease (M=74.2, SD=28.9 vs. M= 40, SD=29.3 in NOT PS group), and on the PS scale (average of the items 31 and 33) (M=55.7, SD=27.1. vs. M= 35.9, SD=26.4 in NOT PS group) compared with the NOT PS group. For Item 33: My skin disease prevents me from doing outdoor activities, the PS group scored significantly worse (M=65.9, SD=33.2) than the PS Suggestive (M=46.8, SD=34.0) and even more so compared with the NOT PS group (M= 31.8, SD=28.9) ().
Mean scores for photosensitivity items of Modified Skindex-29+3
CLASI activity and photosensitivity
ANOVA showed that the effect of photosensitivity was significant for CLASI activity score (F=8.30, p=0.005). Subjects in the PS group had significantly increased lupus specific cutaneous disease activity as measured by the CLASI activity score compared with subjects in the NOT PS group (M=9.2, SD=9.4 vs. M=4.6, SD=5.6). A linear regression analysis with CLASI activity as the response variable and photosensitivity (yes/no) as the independent variable was statistically significant (F = 8.30, p =0.005). CLASI activity was moderately correlated (r=0.36) with the presence/absence of photosensitivity ().
Effect of photosensitivity and CLASI activity on Modified Skindex 29+3 subscales
Two-way ANOVA yielded a main effect for photosensitivity (yes vs. no), on all three subscales of the modified Skindex-29+3, symptom (p=0.007), emotion (p=0.003), and function (p=0.000) such that the PS group had significantly worse scores on average compared with the NOT PS group (). A main effect for CLASI activity [mild (0–9) vs. moderate and severe (>9)] was also significant for the symptom (p=0.007) and emotion subscales (p=0.003) but not for the function subscale (p=0.10) of the modified Skindex-29+3 such that the subjects with more severe (higher) CLASI activity scores had significantly worse scores on average compared with subjects with mild (lower) CLASI activity scores.
Mean Modified Skindex-29+3 subscale scores between the PS and NOT PS group
Within PS Group strata: Hi/Low PS and Mild/Moderate-Severe CLASI activity
ANOVA showed that the effect of PS group strata was significant for symptom, emotion, and function (F=11.90, p=0.000; F=13.76, p=0.000; F=14.97, p=0.000). Post hoc analyses using Bonferroni criterion for significance showed that the Hi PS group scored significantly worse on the symptom scale compared to the Low PS group within each disease activity strata [(Mild Activity Group: Low PS M=26.8, SD=19.2 vs. Hi PS M=47.7, SD=22.9, p=0.001)(Mod-Severe Activity Group: Low PS M=37.7, SD=12.5 vs. Hi PS M=60.8, SD=16.7, p=0.006)]. The same was true for emotions with the Hi PS group scoring significantly worse than the Low PS group within each disease activity strata [(Mild Activity Group: Low PS M=29.7, SD=23.0 vs. Hi PS M=61.4, SD=24.5, p=0.000)(Mod-Severe Activity Group: Low PS M=47.5, SD=25.0 vs. Hi PS M=70.8, SD=18.1, p=0.031)]. In terms of functioning, the Hi PS group scored significantly worse than the Low PS group within and across activity strata such that the Hi PS, Mild Activity group scored worse than the Mod-Severe Activity, Low PS group (Hi PS, Mild Activity M=41.8, SD=25.5 vs. Low PS, Mod-Sev Activity M=19.0, SD=14.8, p=0.010) ().
Mean Modified Skindex-29+3 subscale scores across Hi/Low PS groups and Mild Activity/Moderate-Severe Activity strata