Psoriasis Among Adolescents in Kuwait and the Role of Siblings, Breastfeeding, and Household Cat and Secondhand Smoke Exposure: A Cross-Sectional Study

The current study estimated the lifetime and current prevalence of psoriasis in a large school-based sample of adolescents in Kuwait for the first time and assessed the role of early life risk factors. This investigation showed that 3.6% of the enrolled adolescents have ever had psoriasis and 1.1% have had active psoriasis in the past 12 months, with the scalp and extensor surfaces of the knees and elbows reported as the most commonly affected body areas by psoriasis lesions. Currently having eczema was strongly associated with elevated lifetime and current psoriasis prevalence. Exposure to household SHS and cat during infancy were associated with an elevated lifetime psoriasis prevalence, and showed a trend for association with current psoriasis prevalence. Direct breastfeeding was associated with a significantly reduced lifetime prevalence of psoriasis but was not related to current psoriasis prevalence. Overall, prevalence of both lifetime and current psoriasis demonstrated increasing trends with higher numbers of total, older, and younger siblings. These observed associations highlight the potential role of environmental exposures and sibship size (a marker of prenatal and/or postnatal exposures) in psoriasis development.

Globally, the prevalence of psoriasis is scarcely investigated, with around 81% of countries around the world lacking epidemiologic data on psoriasis [23]. Moreover, existing studies reporting on psoriasis prevalence are difficult to compare due to different methodologies, types of reported prevalence estimate (point, period, or lifetime), methods of ascertaining psoriasis (self-reported or physician/dermatologist diagnosed), population age (children, adults, or total population), and study settings (population-based, existing electronic health records, clinical settings) [21, 28]. In the current study, the lifetime prevalence of parent-reported doctor-diagnosed psoriasis was estimated to be 3.6%, which is higher than the most extreme estimate reported in the literature among children, i.e., 2.1% in Italy [14]. Moreover, the prevalence of current parent-reported doctor-diagnosed psoriasis was estimated to be 1.1% among the enrolled adolescents in our study, which is twice as high as the most extreme period prevalence of 0.45% reported among children in Germany [15, 29]. At the regional level, a study conducted among Saudi schoolchildren (average age 10.3 years) estimated the prevalence of psoriasis to be 0.6% [30], which is comparable to our findings of a current prevalence of 1.1%. Given that genetic factors are the largest contributor to psoriasis development [2, 3, 31], a plausible explanation of the observed elevated prevalence of psoriasis in our sample is the high prevalence of consanguinity (inbreeding) in Kuwait, which was previously estimated to be between 22.6 and 42.1% [32]. Consanguineous marriages are associated with higher rates of genetically determined diseases due to increased homozygosity of recessive alleles [33]. For instance, a meta-analysis demonstrated that offspring of consanguineous marriages were more likely to suffer from primary immunodeficiency diseases with autosomal recessive pattern of inheritance than children with no parental consanguinity [34]. In our sample of adolescents with current psoriasis, 50% reported family history of psoriasis (i.e., grandparents, parents, and/or siblings), where 24.4% reported history in grandparents, 33.3% reported history in siblings, 7.1% reported history in father, and 11.9% reported history in mother (the prior proportions are not mutually exclusive). Our estimate of positive family history (50%) is consistent with prior studies showing 47% of psoriasis patients in Spain [35] and 45.9% of psoriasis patients in Italy [36] reported a positive family history. Moreover, out of the 42 subjects with current psoriasis, 42.9% were born to couples related as second cousins or closer (data not shown). Hence, such factors may underlie the elevated prevalence estimates in our study sample. Moreover, our study might have included more mild cases of psoriasis due to the population-based sample, where such cases are usually less likely to be included in clinical-based samples. Also, parents’ inability to differentiate between psoriasis and eczema (atopic dermatitis)/seborrheic dermatitis may have contributed to a higher prevalence of psoriasis in our report. However, eczema and psoriasis show clear difference in their anatomical location (flexor vs. extensor surfaces, e.g., of the elbows and knees).

Prior investigations have reported conflicting findings regarding sex distribution in psoriasis, with reports suggesting that psoriasis affects males and females equally [14, 15, 21], psoriasis is slightly more common in males than females [20], and psoriasis is slightly more prevalent in females than males [37, 38]. Our findings showed that the lifetime prevalence of psoriasis did not differ between males (3.3%) and females (3.8%). However, current psoriasis prevalence was more common among females (1.4%) compared to males (0.8%). Hence, indicating that the type of prevalence measure (lifetime, point, or period) might explain the inconsistencies in the reported sex disparities. Another possible explanation for the marginal female predominance is that females experience, on average, an earlier age of onset of psoriasis compared to males [23]. Several studies have shown consistent positive associations between adiposity and psoriasis [10, 39‐41]. However, our analysis did not show any such association, rather a statistically non-significant increase in lifetime and current psoriasis prevalence among thin (underweight) subjects compared to those with normal weight was observed. This finding contradicts the current literature and should be corroborated in future investigations. Exposure to household SHS was associated with increased prevalence of lifetime and current psoriasis in our report, which is consistent with the existing scientific literature [11, 12, 42]. Moreover, the observed strong association between current eczema and both lifetime and current psoriasis prevalence is in agreement with prior investigations [29, 43], and is further supported by the shared genetic predisposition [44, 45].

We have specifically investigated the association of breastfeeding, household pet (cat/dog) exposure during infancy, and sibship size with psoriasis due to the ability of these factors to influence the development of the immune system (i.e., immune maturation and response). Breastfeeding was strongly associated with reduced prevalence of lifetime psoriasis, but not current psoriasis prevalence. Hence, indicating that the protective effect of breastfeeding is limited to psoriasis that develops early in childhood. We identified only one prior study that showed no exclusive breastfeeding for at least 3 months was associated with increased odds of psoriasis (odds ratio = 1.46, 95% CI 1.09–1.95) [42]. This finding further supports our observation that breastfeeding might provide early life protection for children susceptible to psoriasis. In contrast, we observed positive associations between family ownership of a household cat during a child’s first year of life and lifetime and current psoriasis prevalence, though the association with current psoriasis did not gain statistical significance. Dog-keeping during child’s infancy showed a trend for association with lifetime psoriasis and showed no effect on current psoriasis. This observation might have been obscured by the fact that a very limited proportion (2.2%) of families in our study sample reported having dogs (i.e., lack of statistical power). Moreover, we showed that increased numbers of total, older, and younger siblings are associated with an elevated lifetime and current psoriasis prevalence. In particular, an elevated prevalence of lifetime and current psoriasis was observed in the groups with a large total number of siblings (e.g., 4 or ≥ 5) and a large number of older siblings (e.g., ≥ 4). Whereas, having at least one younger sibling was associated with an increased lifetime and current psoriasis prevalence. Overall, when analyzed as quantitative variables to infer trends per additional sibling, numbers of total, older, and younger siblings demonstrated statistically significant positive associations with lifetime and current psoriasis. These reported associations of household cat-keeping in infancy and sibship characteristics with psoriasis are novel and are exclusively based on empirical findings. Since immune dysregulation, specifically the stimulation of T helper type 1 and type 17 (Th1 and Th17) cells, in addition to resident skin cells (e.g., keratinocytes and dendritic cells), in response to external triggers/insults [3, 4, 46], characterize psoriasis development, we speculate that immune-mediated mechanisms might underlie the aforementioned observed associations. For instance, breastfeeding has been shown to support the proper maturation of an infant’s immune system, which might explain the observed protective effect of breastfeeding on psoriasis in early life [47].

The observed positive association between household cat exposure during infancy and psoriasis could be explained by the fact that keeping pets, including cats, has been linked to increased endotoxins and microbial exposures in homes; such exposures in early life may promote exaggerated Th1 immune responses and, in turn, possibly increase the risk of psoriasis in susceptible individuals [48, 49]. In general, sibship is a marker of unknown exposure; however, the effects of older siblings could reflect prenatal and/or postnatal programming, whereas associations with younger siblings are likely explained by postnatal mechanisms [50]. An investigation demonstrated that immune priming occurs in utero and depends on the number of previous maternal pregnancies [51]. Moreover, a study demonstrated that the presence of siblings is associated with in utero programming of exaggerated Th1 and Th17 immune responses in the airways of asymptomatic neonates [52]. Such findings support the observed elevated psoriasis prevalence with increased numbers of total, older, and younger siblings. In addition, the association observed between sibship characteristics and psoriasis prevalence is opposite from what has been reported for eczema and allergic diseases, namely a decreasing risk with increasing number of siblings has been widely reported [53, 54]. Given that the pathophysiology of psoriasis involves Th1/Th17 pathways and allergic diseases are driven by Th2 activation, the opposite effects of siblings on psoriasis (more siblings is associated with higher risk) and allergies (more siblings is associated with lower risk) can be speculatively explained by different Th1/Th17 and Th2 patterns. However, more studies are needed to corroborate our findings.

A major strength of the current study is the large and representative study sample that allowed for the estimation of psoriasis prevalence among schoolchildren throughout Kuwait. A potential limitation to our study is the inaccuracy of parent/guardian-reported weight and height of their children, which could have led to misclassification of children into the BMI-for-age groups. However, the estimated prevalence of overweight (25.3%) and obesity (28.8%) in the current report did not substantially differ from that in a previous investigation conducted among schoolchildren aged 6–18 years in Kuwait that used objective weight and height measurements (overweight, 21.6%; and obesity, 30.5%) [55]. Moreover, the possibility of misclassification of the psoriasis status cannot be excluded because it was parent-reported. However, to increase the validity of parent-reporting, we asked whether psoriasis was doctor-diagnosed. A prior study has shown that the validity of self-reported psoriasis increases when asking if the condition was physician-diagnosed [56]. To minimize the effect of recall bias, when defining current psoriasis, we incorporated having had active psoriasis lesion(s) and/or used psoriasis treatment in the past 12-month, in addition to lifetime history of doctor-diagnosed psoriasis. The validity of the current psoriasis definition used in this study is supported by the fact that a high proportion of subjects with current psoriasis showed typical manifestation of psoriasis in usual anatomical sites (scalp [47.6%] and extensor surfaces of the knees [50%] and elbows [38.1%]), which is supported by previous studies [3, 57]. A prior validation study showed that compared to the gold standard (i.e., clinical skin examination performed by a dermatologist), self-reported psoriasis had an estimated sensitivity of 56%, specificity of 99%, positive predictive value of 78%, and negative predictive value of 96%. [56]. These results indicate that self-reporting is a valid method for ascertaining psoriasis in population-based studies that may underestimate rather than overestimate the true prevalence. Similarly, another study found that self-reporting of psoriasis tends to underestimate the prevalence of cases identified in clinical settings [38]. However, the generalizability of the aforementioned validation study results to our study population should be cautiously interpreted since those studies were conducted in different settings and populations (i.e., Norwegian and Danish adults). Another issue is that differential diagnosis of psoriasis with eczema is common and may lead to misclassification. However, of the 42 children classified as currently having psoriasis in our report, 35 (83.3%) reported that their diagnosis was made by a dermatologist, and 4 (9.5%) and 3 (7.2%) were diagnosed by general practitioner and pediatrician, respectively. Hence, given that the majority were diagnosed by dermatologists, the effect of misdiagnoses on results of this report should be minimal. Given that 0.8% of the children with no current eczema had a diagnosis of psoriasis and 3.6% of children had current eczema, there may be a small overlap in the diagnoses of the two diseases. A further limitation to our study is the lack of statistical power to detect statistically significant effects when assessing associations with current psoriasis prevalence (e.g., effect of SHS: aPR = 1.77, 95% CI 0.89–3.53); this is due to the limited number of subjects (n = 42) with current psoriasis. Moreover, family history of psoriasis was only sought from subjects who reported current psoriasis; hence, we were not able to adjust for its effects in our association analyses. Given that psoriasis and eczema share, to some extent, similar clinical and pathologic features, including immune dysregulation and epidermal barrier defects [58], our association analyses adjusted for the effect of current eczema. Hence, the reported associations were independent of eczema effect. Moreover, it is essential to indicate that our analysis aimed to assess associations between different exposures and psoriasis rather than to infer causal relationships.