Abstract
Psoriasis mimics uric acid in terms of inflammation, but the association has not been well defined. This study aimed to identify the causal link between serum uric acid (SUA) and psoriasis in an observational study using the National Health and Nutrition Examination Survey (NHANES, 2004–2006, and 2011–2014) and transethnic Mendelian randomization (MR) analyses. We utilized weighted multivariable-adjusted logistic regression and transethnic MR in European and East Asian populations to assess the association. Inverse variance weighted (IVW) was the main analysis. To test the robustness and pleiotropy, further sensitivity analyses were also conducted. Weighted regression analysis suggested that SUA positively related to psoriasis risk (OR = 1.339, 95% CI: 1.092–1.642, P = 0.006) in women. For all participants and males, neither association was significant. IVW showed that SUA levels were not significantly associated with psoriasis in Europeans (OR = 1.099, 95% CI: 0.963–1.254, P = 0.159) or East Asians (OR = 1.297, 95% CI: 0.576–2.918, P = 0.528). Furthermore, sensitivity analyses confirmed the robustness of the present MR results. In females, SUA and psoriasis were significantly correlated; findings from transethnic MR analysis did not indicate a causal relationship between SUA and psoriasis.
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-024-77222-y.
Keywords: Uric acid, Psoriasis, NHANES, Mendelian randomization, Causality
Subject terms: Genetics, Diseases
Introduction
Psoriasis is a chronic inflammatory, immune-mediated skin disease that affects more than 7.5 million people among adults in the United States, with an estimated 3.0% prevalence1. According to the Global Burden of Disease study, it imposes a significant disease burden on the elderly population aged 60 to 692.
The characteristics of psoriasis have decreased cellular turnover time3. Indeed, the hyperproliferation of keratinocytes leads to increased uric acid synthesis. Uric acid has been demonstrated to mediate inflammation involved in the pathogenesis of hypertension and atherosclerosis4, and it also acts as a beneficial antioxidant for psoriasis5. Hyperuricemia has been correlated with markers of systemic inflammation, including C-reactive protein6, and is also considered as an independent risk factor for other diseases7. However, given the similarity in inflammation pathways between SUA and psoriasis, no definitive unifying relationship exists. This study aims to evaluate this association.
Previous studies have shown that the correlation is either ethnicity- or region-dependent8,9. With respect to East Asian ancestry, several studies have reported a correlation between SUA levels and psoriasis10,11. However, the direction of association varies between Europeans and East Asians, and conflicting results12,13 do not allow us to draw definitive conclusions. Psoriasis is a multifactorial disease from a biological perspective, and the impact of ethnicity-diversity remains largely understudied. We thus perform additional validation with independent ancestry rather than single ancestry for causal relationships.
The NHANES is an ongoing population based, cross-sectional study. This is unique in that the NHANES survey was selected via a complex, multistage, stratified sampling design14, which provides large sample data to evaluate the correlation between SUA and psoriasis.
Mendelian randomization studies use genetic variation associated with modifiable exposures to assess their possible causal relationship15. MR studies have the potential to avoid some of the limitations of observational epidemiology and RCTs16. This approach effectively reduces associations with confounders of the exposure-outcome and mitigates reverse causation from the outcome or exposure to the genotypes17. All MR approaches are based on three core assumptions to estimate the causal effects of the exposure on the outcome minimize18: (i) the instrumental variables (IVs) have a strong relationship with exposure, (ii) the IVs are independent of factors that confound the exposure-outcome relationship, and (iii) the exclusion restriction, the IVs affects the outcome only through the exposure (as shown in Fig. 1).
Fig. 1.
Flowchart of the Mendelian randomization study showing the causal association between SUA and psoriasis. IVW: Inverse variance weighted; MR: Mendelian randomization; PRESSO: Pleiotropy residual sum and outlier; SNP: Single nucleotide polymorphism; SUA: Serum uric acid.
Given the different levels of SUA metabolism, we employed an observational study in the NHANES to assess the relationship between SUA and psoriasis in different sexes. We also attempt to find multiple endings in Europeans and East Asians, and approaches to strengthen causal inference on psoriasis using transethnic MR to explore risk factors.
Materials and methods
Study population in the NHANES
The study protocol was authorized by the Research Ethics Review Board of the National Centre for Health Statistics, CDC, USA. The NHANES was approved by the NCHS Ethics Review Board (Protocols# 98 − 12, 2005-06 and 2011-17). Informed consent was obtained from all the participants. The study data are publicly available through the NHANES database (https://www.cdc.gov/nchs/nhanes/index.htm).
The study population included 40,401 individuals from four NHANES cycles (2004–2005, 2005–2006, 2011–2012, and 2013–2014). After the demographic variables, questionnaire information, examination, and laboratory test data were merged, 21,349 participants met the following inclusion criteria: aged 20 years or above. Finally, a total of 13,570 participants were included in the analysis (see Supplementary Fig. S1). The exclusion criteria were as follows: (1) missing information on psoriasis; (2) incomplete serum uric acid data; and (3) missing data on other covariates.
Assessment of psoriasis and SUA in NHANES
In this study, data on the presence or absence of psoriasis were obtained from the “medical conditions” and “dermatology” sections of the questionnaire. The participants who answered “yes” to the question “Ever been told you have psoriasis?” or “Have you ever been told by a doctor or other health care professional that you had psoriasis?” were also classified as having psoriasis14,19.
The distribution of SUA status was examined via quartile division. The SUA concentration was equally categorized into ≤ 25th percentile, > 25th and ≤ 50th percentile, > 50th and ≤ 75th percentile, and ≥ 75th percentile in the female and male groups. The concentration of uric acid in the serum was measured by DxC800 with the timed endpoint method20.
Ascertainment of covariates in NHNAES
Race/ethnicity was categorized as Mexican American, other Hispanic, non-Hispanic White, non-Hispanic Black, or other race. The educational level was classified as high school or equivalent, more than high school or less than high school. Marital status was classified as married, living with partner, or living alone. Alcohol status was classified as “yes” or “no” on the basis of the consumption of at least 12 alcohol drinks/year. Smoking status was categorized into never, former, or current. It was dependent on the condition of whether participants had smoked fewer than 100 cigarettes throughout their entire life and now. BMI was classified as underweight (< 25), normal weight (25–30), overweight or obese (> 30).
The American College of Cardiology (ACC) and American Heart Association (AHA) recommended that individuals with SBP (systolic blood pressure, ≥ 130 mmHg) and/or DBP (diastolic blood pressure, ≥ 80 mmHg) in 2 categories should be assigned to the higher blood pressure category21. A specific question concerning the patient’s condition was used: a self-reported physician diagnosis of hypertension or self-reported use of medication for hypertension was also classified as having hypertension. Type 2 diabetes (T2D) was defined as a self-reported diagnosis by a physician, the use of insulin or oral hypoglycemic medication, fasting blood glucose ≥ 7.0 mmol/L (126 mg/dL), postprandial 2-h plasma glucose ≥ 11.1 mmol/L (200 mg/dL) from an oral glucose tolerance test, or glycated hemoglobin A1c (HbA1c) ≥ 6.5% (48 mmol/mol)22. Metabolic syndrome (MetS) was defined by the NCEP ATP III-2005 criterion23.
Sources of MR analysis and single nucleotide polymorphism (SNP) selection
We conducted MR analysis on individuals of European and East Asian ancestry separately to explore the causal effects of SUA on psoriasis. Figure 1 shows the flow chart of the entire analysis.
We obtained summary data for SUA from the MRC Integrative Epidemiology Unit (MRC-IEU) consortium, which included 343,836 European participants and 109,029 East Asian participants in total. For psoriasis, GWAS data from 172,495 East Asian participants were also extracted from MRC-IEU. Moreover, we obtained summary data for psoriasis from the FinnGen GWAS, with a total of 407,876 (10,312 cases and 397,564 controls) participants of European ancestry. Detailed information on the data sources is displayed in Supplementary Table S1. Psoriasis was diagnosed according to the International Classification of Diseases (ICD) -10 L40. All studies were reviewed and approved by local institutional review boards.
We used a two-sample MR method to identify the causal relationship between SUA and psoriasis through the use of exposed genetic variations as IVs. All SNPs associated with SUA were analyzed on the basis of three assumptions18. Those SNPs were selected via the following steps: exposure factors with genome-wide significance parameters were set (p < 5 × 10− 8); linkage disequilibrium (LD) analysis was performed to screen for no linkage effects; and parameters, including the clumping process (r2 < 0.001), clumping distance 10 MB, were also set; and proxy SNPs in linkage disequilibrium of r2 ≥ 0.8.
We calculated the F statistics of the SNPs to exclude the presence of weak instrument bias (F statistics < 10)24 and ensure that the result remained sufficient strength. The following formula was chosen to calculate the F value: F = R2(N-k-1)/[k(1-R2)]25, where k is the number of instruments, R2 is the proportion of the variability of SUA explained by each instrument and N is the sample size. The F-statistic values were all > 10, suggesting that these SNPs are suitable IVs (Supplementary Table S2).
Considering that instruments may exert an effect on psoriasis independent of SUA, we manually checked and eliminated SNPs that were significantly related to potential confounding factors via LDtrait tools and the GWAS catalog. These factors, including smoking or cigarette use, drinking or alcohol intake, BMI (body mass index), cardiovascular disease, hypertension, metabolic syndrome and diabetes26,27, have been strongly evaluated in the existing literature.
Finally, filtered SNPs for SUA were extracted for the subsequent causality analysis. Supplementary Table S2 lists those SNPs used as IVs for European and East Asian ancestry.
Statistical analysis
The data were analyzed according to the NHANES statistical tutorial. Categorical variables are presented as percentages (%), whereas continuous variables are presented as medians (interquartile ranges, IQRs). Owing to differences in SUA metabolism between the sexes, analyses were performed separately by sex. The hypothesis test for variables and chi-square test for categorical variables with continuity correction were used to detect different between sex groups. Continuous variables could be compared via t-test, otherwise, the Kruskal-Wallis H-test was used. Analyses including baseline characteristics and regression models were performed under specific subsample weights.
Weighted multivariable-adjusted logistic regression was used to test the odds ratios (ORs) and 95% confidence intervals (95% CIs) for the association between SUA and psoriasis. Five regression models were constructed to account for the influence of confounding factors. Model I was weighted logistic regression, the independent variable was uric acid, and the dependent variable was psoriasis. Model II was adjusted for sociodemographic characteristics, including age and race/ethnicity. Model III was adjusted for Model II and included other sociodemographic characteristics (education, marital status, family PIR, smoking status, and alcohol status). Model IV was adjusted for sociodemographic characteristics, and BMI and cotinine were added to Model III. Model V was further adjusted for sociodemographic characteristics; that is, hypertension, type 2 diabetes and metabolic syndrome were controlled for as associated covariates. Additionally, five knots (5th, 27.5th, 50th, 72.5th, and 95th quantiles) restricted cubic splines (RCSs) without weights were used to estimate the exposure-response curves of uric acid and psoriasis. A p-value < 0.05 for overall and nonlinear variables indicates a nonlinear relationship between uric acid and psoriasis.
We followed the STROBE-MR guidelines established by Skrivankova15 and conducted two-sample MR analysis to ensure the dependability of our findings. Supplementary Table S3 contains the checklist for the STROBE-MR guidelines. In this study, we applied inverse variance weighted (IVW) as the main statistical method to assess the causal association between SUA and psoriasis. To validate the results from IVW, MR-Egger regression, weighted median (WM), simple mode and weighted mode, MR pleiotropy residual sum and outlier (MR-PRESSO) were used together as supplementary methods for MR analysis.
In addition, we perform random effects IVW and MR-Egger methods with Cochran’s Q-test to examine potential heterogeneity (P < 0.05)28. The MR-Egger intercept and MR-PRESSO were used to test whether multiple IVs have horizontal pleiotropy29. Next, we employed leave-one-out sensitivity analysis to identify whether the causal relationship between SUA and psoriasis was driven by a single SNP. We also carried out the MR-Steiger directionality test to ensure that there was no inverse directionality in the causal estimates.
We conducted all the statistical analyses via R software (version 4.3.2). The packages “TwoSampleMR”, “Mendelianrandomization”, “MRPRESSO” and “rms” were used for MR and NHANES data analysis.
Results
Basic characteristics of the participants
After sampling weight was calculated, the prevalence of all participants with psoriasis was 3.1%, and these individuals were classified into 6,973 (51.4%) women and 6,597 (48.6%) men (Table 1). This cohort included a total of 163,532,283 participants who were enqueued. Table 1 also shows the general characteristics of the participants in the female and male groups, and some variables significantly differed between the female and male groups in terms of education, marital status, family PIR, alcohol status, smoke, BMI, cotinine, hypertension, and type 2 diabetes. There were more female participants with psoriasis than male participants. However, there was no significant difference in the prevalence of psoriasis between those individuals (p = 0.279, Table 1). In males, uric acid was higher than that in females with and without psoriasis, and there was a significant difference in uric acid only in females (p = 0.002). (Supplementary Table S4-S5)
Table 1.
Basic characteristics of the participants in the female and male populations (weighted selection, N = 163,532,283). The percentages and medians were estimated via the U.S. population and study subsample weight. BMI: body mass index; IQR: interquartile range; PIR: poverty income ratio. *p < 0.05; **p < 0.01.
| Variables | Total (n = 13570) |
Female (n = 6973) |
Male (n = 6597) |
P-value |
|---|---|---|---|---|
| Age(year), Median(IQR) | 43.0 (32.0, 54.0) | 44.0 (32.0, 55.0) | 43.0 (31.0, 54.0) | 0.247 |
| Race/ethnicity, n(%) | 0.193 | |||
| Mexican American | 1130 ( 8.3) | 516 ( 7.4) | 614 ( 9.3) | |
| Non-Hispanic Black | 1517 (11.2) | 844 (12.1) | 673 (10.2) | |
| Non-Hispanic White | 9441 (69.6) | 4804 (68.9) | 4637 (70.3) | |
| Other Hispanic | 624 ( 4.6) | 314 ( 4.5) | 310 ( 4.7) | |
| Other Race | 858 ( 6.3) | 495 ( 7.1) | 363 ( 5.5) | |
| Education, n(%) | < 0.001** | |||
| Less than high school | 2072 (15.3) | 997 (14.3) | 1075 (16.3) | |
| High school | 2971 (21.9) | 1374 (19.7) | 1597 (24.2) | |
| More than high school | 8527 (62.8) | 4602 (66.0) | 3925 (59.5) | |
| Marital status, n(%) | < 0.001** | |||
| Living alone | 4763 (35.1) | 2566 (36.8) | 2197 (33.3) | |
| Married or living with partner | 8807 (64.9) | 4407 (63.2) | 4400 (66.7) | |
| Family PIR, Median(IQR) | 3.08 (1.49, 5.00) | 2.94 (1.41, 4.97) | 3.18 (1.56, 5.00) | < 0.001** |
| Alcohol status, n(%) | < 0.001** | |||
| No | 2969 (21.9) | 2085 (29.9) | 884 (13.4) | |
| Yes | 10,601 (78.1) | 4888 (70.1) | 5713 (86.6) | |
| Smoke, n(%) | < 0.001** | |||
| Never | 7462 (55.0) | 4289 (61.5) | 3173 (48.1) | |
| Current | 3089 (22.8) | 1380 (19.8) | 1709 (25.9) | |
| Former | 3019 (22.2) | 1304 (18.7) | 1715 (26.0) | |
| BMI, n(%) | < 0.001** | |||
| < 25 | 4268 (31.5) | 2447 (35.1) | 1821 (27.6) | |
| 25–30 | 4859 (35.8) | 2629 (37.7) | 2230 (33.8) | |
| > 30 | 4443 (32.7) | 1897 (27.2) | 2546 (38.6) | |
| Cotinine (ng/mL), Median(IQR) | 0.05 (0.01, 39.72) | 0.03 (0.01, 0.87) | 0.09 (0.02, 122.77) | < 0.001** |
| Hypertension, n(%) | < 0.001** | |||
| No | 7653 (56.4) | 4170 (59.8) | 3483 (52.8) | |
| Yes | 5917 (43.6) | 2803 (40.2) | 3114 (47.2) | |
| Type 2 diabetes, n(%) | 0.015* | |||
| No | 11,922 (87.9) | 6143 (88.1) | 5779 (87.6) | |
| Yes | 1648 (12.1) | 830 (11.9) | 818 (12.4) | |
| Metabolic syndrome, n(%) | 0.076 | |||
| No | 8252 (60.8) | 4281 (61.4) | 3971 (60.2) | |
| Yes | 5318 (39.2) | 2692 (38.6) | 2626 (39.8) | |
| Uric acid (mg/dL), Median(IQR) | 5.40 (4.40, 6.30) | 4.60 (3.90, 5.40) | 6.00 (5.30, 6.90) | < 0.001** |
| Psoriasis, n(%) | 0.279 | |||
| No | 13,147 (96.9) | 6735 (96.6) | 6412 (97.2) | |
| Yes | 423 ( 3.1) | 238 ( 3.4) | 185 ( 2.8) |
Relationship between SUA and psoriasis in an observational study by sex
The results of weighted regression model I-V analysis revealed that there was a significant relationship between SUA and psoriasis risk in women (Table 2). After full adjustment for potential covariates, no associations with psoriasis were observed either in males (Supplementary Table S6) or in all participants (Supplementary Table S7) when quartiles were used. In women, the adjusted OR for SUA and psoriasis in Q4 (≥ 5.4 mg/dl) of Model II, Model III, Model IV, and Model V was 1.86 (95% CI: 1.24 ~ 2.79, p = 0.003), 1.82 (95% CI: 1.21 ~ 2.74, p = 0.004), 1.66 (95% CI: 1.08 ~ 2.54, p = 0.020), and 1.66 (95% CI: 1.10 ~ 2.51, p = 0.017), respectively, with Q1 (≤ 3.8 mg/dl) as a reference (Table 2).
Table 2.
Weighted regression to determine the odds of psoriasis presence by SUA in the female population. Model I non-adjusted model; model II adjusts for age, Race/ethnicity; Model III adjusts for age, Race/ethnicity, Education, Marital status, Family PIR, Alcohol status, smoke; Model IV adjusts for age, Race/ethnicity, Education, Marital status, Family PIR, Alcohol status, smoke, BMI, cotinine; model V adjusts for age, Race/ethnicity, Education, Marital status, Family PIR, Alcohol status, smoke, BMI, cotinine, hypertension, type 2 diabetes, and metabolic syndrome. (ref) = reference; CI: confidence interval; OR: odds ratio; PIR: poverty income ratio; Q: quartile; SUA: serum uric acid. *p < 0.05; **p < 0.01.
| SUA, mg/dl | SUA(mg/dl), quartile | ||||
|---|---|---|---|---|---|
| Q1 (≤ 3.8) | Q2 (3.9–4.5) | Q3 (4.6–5.3) | Q4 (≥ 5.4) | ||
| Model I | |||||
| OR (95%CI) | 1.421 (1.211–1.667) | 1(ref) | 1.41 (0.93–2.15) | 1.26 (0.81–1.96) | 2.11 (1.43–3.12) |
| P-value | < 0.001** | 0.047* | 0.301 | < 0.001** | |
| Model II | |||||
| OR (95%CI) | 1.363 (1.145–1.622) | 1(ref) | 1.37 (0.90–2.09) | 1.18 (0.76–1.84) | 1.86 (1.24–2.79) |
| P-value | < 0.001** | 0.143 | 0.463 | 0.003** | |
| Model III | |||||
| OR (95%CI) | 1.343 (1.125–1.603) | 1(ref) | 1.37 (0.90–2.09) | 1.18 (0.76–1.84) | 1.82 (1.21–2.74) |
| P-value | 0.002** | 0.144 | 0.464 | 0.004** | |
| Model IV | |||||
| OR (95%CI) | 1.351 (1.131–1.615) | 1(ref) | 1.33 (0.87–2.03) | 1.11 (0.70–1.74) | 1.66 (1.08–2.54) |
| P-value | 0.001** | 0.193 | 0.664 | 0.020* | |
| Model V | |||||
| OR (95%CI) | 1.339 (1.092–1.642) | 1(ref) | 1.34 (0.88–2.05) | 1.12 (0.71–1.75) | 1.66 (1.10–2.51) |
| P-value | 0.006** | 0.171 | 0.627 | 0.017* | |
Compared with low SUA levels (< 5.3 mg/dl), Q3 and Q4 in all participants were positively associated with the risk of psoriasis [OR of 1.42 (95% CI: 1.03–1.96, p = 0.034), OR of 1.45 (95% CI: 1.04–2.03, p = 0.029)] in Model II and [OR of 1.44 (95% CI: 1.04–1.98, p = 0.028), OR of 1.46 (95% CI: 1.04–2.04, p = 0.028)] in Model III (Supplementary Table S7). However, the correlation between SUA (≥ 5.3 mg/dl) and psoriasis became insignificant after adjusting for the covariates in Model V [OR of 1.15 (95% CI, 0.81–1.64), p = 0.424] (Supplementary Table S7).
The RCS results suggested that the dose-response relationship between SUA and psoriasis in women (Fig. 2B) and men (Fig. 2A) was not a curve. The risk of psoriasis tended to increase slightly but not significantly with increasing SUA levels in women (Fig. 2B, p = 0.597). In men, the result was the opposite, and the line almost appeared to plateau (Fig. 2A, p = 0.966).
Fig. 2.
Dose-response relationship between SUA and psoriasis in the male population (A) and the female population (B) adjusted for Model V.
Causal relationships between SUA and psoriasis risk in transethnic MR
We further performed transethnic MR analysis to explain the causal effects of SUA on psoriasis risk. Under the selection criteria for SNPs, 134 and 33 SNPs were used as IVs for effect estimates. In addition, they had minimum F statistic of 29.2 and 30.9 in the European and East Asian cohorts (Supplementary Table S2), respectively, indicating that they were able to efficiently verify the relevance assumption.
We also depict the results of exposure to psoriasis with the classic MR method to measure the effect estimates (see Supplementary Fig. S2). For Europeans, no significant causal effect of SUA on the risk of psoriasis was estimated (OR = 1.099, 95% CI: 0.963–1.254, p = 0.159) in IVW, and the OR was 0.951 and P = 0.636 in MR-Egger (Fig. 3). For East Asians, no significant causal effect of SUA on psoriasis risk was observed in the IVW analysis (OR = 1.297, P = 0.528; Fig. 3).
Fig. 3.
Transethnic MR estimates for the relationship between exposure (SUA) and psoriasis. CI: confidence interval; IVW: inverse variance weighted; MR: Mendelian randomization; MR-PRESSO: MR Pleiotropy residual sum and outlier; OR: odds ratio.
Similar effect estimates were observed, which were consistent among the other five methods. Notably, potential heterogeneity was detected via MR-Egger (p = 0.047, Table 3) in the East Asian cohort. We applied the random effects IVW method to reevaluate the causal associations, and no heterogeneity for the selected IVs was found (p = 0.055).
Table 3.
Transethnic MR estimates from each method of assessing the causal effect of SUA on the risk of psoriasis. CI: confidence interval; IVW: inverse variance weighted; MR: mendelian randomization; OR: odds ratio; PRESSO: pleiotropy residual sum and outlier; SE: standard error.
| MR method | β | SE | OR (95% CI) | P for association | Q for heter- ogeneity |
P for heter- ogeneity |
P for MR-Egger intercept |
P for MR- PRESSO global |
|---|---|---|---|---|---|---|---|---|
| European | ||||||||
| IVW | 0.095 | 0.067 | 1.099 (0.963,1.254) | 0.159 | 153.452 | 0.108 | ||
| MR Egger | -0.050 | 0.105 | 0.951 (0.774,1.168) | 0.636 | 149.833 | 0.137 | 0.076 | |
| Weighted median | 0.075 | 0.103 | 1.078 (0.880,1.320) | 0.465 | ||||
| Simple mode | 0.193 | 0.183 | 1.213 (0.847,1.737) | 0.293 | ||||
| Weighted mode | 0.003 | 0.095 | 1.003 (0.833,1.208) | 0.971 | ||||
| MR-PRESSO | 0.095 | 0.067 | 1.099 (0.963,1.254) | 0.162 | 0.126 | |||
| East Asian | ||||||||
| IVW | 0.260 | 0.414 | 1.297 (0.576,2.918) | 0.528 | 45.715 | 0.055 | ||
| MR Egger | -0.058 | 0.738 | 0.943 (0.222,4.007) | 0.937 | 45.313 | 0.047 | 0.604 | |
| Weighted median | 0.155 | 0.495 | 1.167 (0.442,3.081) | 0.754 | ||||
| Simple mode | 0.527 | 0.844 | 1.693 (0.324,8.849) | 0.536 | ||||
| Weighted mode | 0.316 | 0.505 | 1.372 (0.509,3.695) | 0.535 | ||||
| MR-PRESSO | 0.260 | 0.414 | 1.297 (0.576,2.918) | 0.533 | 0.068 | |||
In the two cohorts, the MR-Egger intercept test revealed no significant pleiotropy (all p values > 0.05), whereas MR-PRESSO global did not detect overall horizontal pleiotropy (Table 3) or any outlier variants (all p values > 0.05). The findings from the main analysis were further substantiated by scatter plots (Supplementary Fig. S2), forest plots (Supplementary Fig. S3) and funnel plots (Supplementary Fig. S4). Additionally, the leave-one-out analysis revealed no single SNP that drove the causal estimates of psoriasis (Supplementary Fig. S5). The MR-Steiger directionality test suggested that there was no inverse directionality in this study (all Steiger directions were TRUE, p values < 0.05).
Discussion
In this study, we selected SNPs that were strongly associated with SUA levels in a large-scale GWAS database as IVs. By combining an observational study with transethnic MR analysis, we assessed the causal relationship between SUA levels and the risk of psoriasis. Given the different levels of SUA metabolism in different sexes, observational analysis was employed for each sex. SUA and psoriasis were significantly correlated only in females. However, our findings from transethnic MR do not support the causal association between SUA levels and the risk of psoriasis.
The role of high levels of SUA in psoriasis is complex and worth exploring on the basis of its underlying mechanisms. Elevated serum uric acid levels deposits uric acid in joints and other tissues, leading to the production of inflammatory crystals. Under these conditions, uric acid may be a mediator that triggers a state of chronic low-grade inflammation, which is associated with the upregulation and release of several proinflammatory cytokines30,31. Hyperuricemia is a cause of epidermal proliferation in psoriasis, and inflammatory crystals also promote cell proliferation in other tissues32. In factor, psoriatic T cells exhibit the ability to decrease epidermal turnover time and affect keratinocyte proliferation3. The abnormal proliferation of keratinocytes leads to increased uric acid synthesis through nucleic acid catabolism. In addition, complications related to SUA, e.g., metabolic syndrome, are also involved13 and may serve as possible precipitating factors for the development of psoriasis.
Moreover, urate crystals stimulate normal human keratinocytes (NHKs) and monocytes, including three inflammasomes in macrophages, and produce inflammatory cytokines/chemokines, indicating the promoting effects of uric acid on psoriasis. This bidirectional relationship between SUA and psoriasis provides support for the treatment of psoriasis. Conventional treatments can modify psoriasis course in the early stage, remitting the symptoms, and are useful for treatment. Notably, biologic therapies that target tumor necrosis factor (TNF)-α, interleukin (IL)-17, and IL-23 for psoriasis, such as inhibitors and monoclonal antibodies, have demonstrated long-lasting efficacy and good safety profiles in individuals. A recent study revealed the role of deucravacitinib. As a selective tyrosine kinase 2 (TYK2) inhibitor, it exhibits emerging potential as an effective treatment for psoriasis because it directly affects inflammation and is well-tolerated33. In addition to improving rash, monoclonal antibodies, e.g., adalimumab and secukinumab, might have beneficial effects on metabolic syndrome and cardiometabolic biomarkers associated with psoriasis. These diseases are associated with classic risk factors along with psoriasis and may share key immune pathways and systemic inflammation. Owing to the anti-inflammatory effects of biologic therapies, the rate of DNA formation and purine synthesis is reduced, and the inhibitory effects of TNF-α on HDL formation are counteracted34. Neutrophil-mediated cardiovascular inflammation is also directly reduced35. Thus, some biomarkers (SUA, HDL, and high-sensitivity C-reactive protein) improve in a positive direction. In summary, biologic therapies are effective treatments for psoriasis. These findings may provide an intriguing therapeutic reference for metabolic syndrome, although results from transethnic MRs do not support the causal association between biomarker SUA levels and psoriasis.
The prevalence and incidence of psoriasis exhibit distinct ethnic and geographic patterns. We analyzed baseline characteristics under specific subsample weights from the four-cycle NHANES database, which suggested that the 3.1% prevalence of psoriasis among U.S. adults has not changed significantly1. In our multivariate analysis, SUA and psoriasis were significantly correlated only in females. For all participants, the correlation became insignificant after adjusting for covariates such as metabolic syndrome, hypertension and type 2 diabetes (Supplementary Table S7). Moreover, there is significant diversity in the incidence of hyperuricemia or gout according to sex. A dose-response RCS study investigated the linear relationship between SUA levels and psoriasis in a US population of different sexes. The performances were displayed in different linear trends between males and females. Hormonal regulation may be the reason for the differences in SUA levels36.
These results are not entirely consistent with those of previous studies. Earlier observational studies reported a high incidence of SUA levels in patients with extensive psoriasis37, whereas other studies reported no association between SUA levels and psoriasis38. A series of studies from different regions have subsequently sought to uncover a link between SUA levels and psoriasis. A prospective study of two cohorts included 27,751 men and 71,059 women in the U.S. The results indicated that psoriasis was associated with an increased risk of subsequent gout39. Similarly, a cross-sectional study revealed that hyperuricemia was closely related to psoriasis in Turkey40. In contrast, a study reported that no association was found between SUA and the severity of skin disease in Hong Kong41. Another research project from the NHANES database also refuted this positive association after adjusting for potential covariates13, which was consistent with our results. Studies from two other regions reported by Stephen Chu-Sung Hu10 and H. H. Kwon11 showed that SUA levels were significantly associated with psoriasis in East Asian. Based on the above research, it seems that more and more results may fall into contradictions and debates around correlations. As shown in the results of two comprehensive studies8,9, we realized that ethnicity- or region-dependent patterns for this correlation were designed necessarily. These above results can be explained by differences in genetic epidemiology possibly, that is, as a multifactorial disease, psoriasis clearly has a strong genetic basis and susceptibility to environmental components42. In the future, investigations are needed to establish a suitable association for psoriasis in different ethnicities or geographic regions.
Several strengths were demonstrated in our studies. First, we integrated transethnic MR and observational study, and ethnicity-dependent correlations were authenticated each other in Europeans and East Asians. Second, the application of large-scale GWAS data from Europeans was parallel to East Asian ancestry, which reduced the impact of population stratification. And F statistics exceeding 10 indicate that the IVs used in this MR study are highly influential. Third, a series of MR estimations and sensitivity analyses were performed, including IVW, WM, MR Egger, Cochran’s Q-test, MR-PRESSO, and the Steiger directionality test. This strong statistical evidence underscores the robustness of the study’s findings.
In assessing causality, MR study uses exposed genetic variations as IVs. It primarily addresses two major limitations in observational study, providing robust evidence that strengthens or challenges observed associations17,43. In MR study, the F statistic reflects the ‘strength’ of the genetic IVs, and is an indicator of the extent (size and probability) of the relative bias that is likely to occur in estimating a causal association using the IVs44. To avoid making causal inferences directly, we report the F statistics and examine the strength of IVs to exclude the presence of weak instrument bias (F statistics < 10)24. The weakest across all included variants is 29.2 in our study (Supplementary Table S2). This ensures that the strength of the IVs is appropriate. However, we admit that we may need adequate sample sizes and understand the (unknown) strength of confounding as much as possible in our MR analysis. In the context of inadequate sample sizes and ‘weak instrument’ scenarios using multiple genetic variants, effect estimates may be substantially biased45,46. Coupled with the winner’s curse47, these factors may affect the sensitivity of MR analysis. If the minimum p-value for association with exposure is low, it mitigates the impact of bias because of winner’s curse48. To obtain suitable and sufficient variants, p-value was set at p < 5 × 10− 8 in sensitivity analysis for MR. Thus, inferences should be interpreted with caution, even though our results did not detect overall horizontal and pleiotropy.
Numerous limitations can be identified. First, in our study, the GWAS datasets for SUA and psoriasis were derived from individuals of European and East Asian ancestry. Considering ethnicity- or region-dependent patterns on correlation, our findings need to be extrapolated to other regions. Second, with respect to the degree of psoriasis for our option, psoriasis was strictly defined on the basis of the questionnaire from the NHANES database. However, most test results including self-reported data from questionnaire surveys, may be influenced by social factors, inevitably producing measurement bias. Third, while we try to control over confounding factors in transethnic MR research, some potential factors and comorbidities cannot be entirely ruled out. Fourth, the winner’s curse can affect genetic associations47. To balance the winner’s curse bias with potential low power due to selecting fewer variants, p-value was set at the typical selection threshold in our study. Finally, a possible biological explanation is necessary. Primarily due to time and funding constraints, inconsistent results from observational and MR study were unable to be validated through additional clinical studies. This happens to be a focal point for our future direction of work.
Conclusion
Despite of an association between SUA and psoriasis in females from observational results, by using large-scale GWAS database of psoriasis in European and East Asian ancestry, our findings from transethnic MR analyses did not indicate a causal relationship between SUA and psoriasis. We need large-scale prospective studies based on larger GWAS samples or further MR analysis to test causal relationships, thus providing a reference for psoriasis and skin health.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Acknowledgements
We want acknowledge the participants and investigators of NHANES, the FinnGen and MRC-IEU study. The authors wish to thank Dr. Li Jiaxin’s wonderful suggestions.
Author contributions
Conceptualization: ZD, ZJR, WS; Data Curation: ZD; Formal Analysis: ZD, WS; Investigation: ZD, WS, ZJR; Methodology: ZD, WS; Project Administration: ZD, SJH; Supervision: ZJR, SJH; Writing - Original Draft Preparation: ZD; Writing - Review and Editing: ZD, ZJR, SJH, WS.
Funding
This work has been supported by Initiation Funds for High-Level Talents Program of Xi’an International University (Grant no. XAIU202402).
Data availability
The data that support the findings of this study are obtained from public databases. The SUA data for European participants and East Asian participants are available through the database at http://ftp.ebi.ac.uk/pub/databases/gwas/summary_statistics/GCST90018001-GCST90019000/GCST90018977/ and https://gwas.mrcieu.ac.uk/datasets/bbj-a-57/, respectively. The psoriasis summary statistics in the MR analysis are downloaded from the FinnGen database (https://storage.googleapis.com/finngen-public-data-r10/summary_stats/finngen_R10_L12_PSORIASIS.gz) and GWAS Catalog (http://ftp.ebi.ac.uk/pub/databases/gwas/summary_statistics/GCST90018001-GCST90019000/GCST90018687/). Further inquiries can be directed to the first author through email xaiu21063@xaiu.edu.cn.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
These authors contributed equally: Dong Zhao and Jin-rong Zhao.
Contributor Information
Jin-rong Zhao, Email: zhaojrr@126.com.
Shuai Wang, Email: 240188452@qq.com.
Ji-hu Sun, Email: ywjkjc1@163.com.
References
- 1.Armstrong, A. W. et al. Psoriasis prevalence in adults in the United States. JAMA Dermatol.157(8), 940–946. 10.1001/jamadermatol.2021.2007 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Damiani, G. et al. The Global, Regional, and National Burden of Psoriasis: Results and insights from the Global Burden of Disease 2019 Study. Front. Med.8, 743180. 10.3389/fmed.2021.743180 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Li, J. et al. Psoriatic T cells reduce epidermal turnover time and affect cell proliferation contributed from differential gene expression. J. Dermatol.42(9), 874–880. 10.1111/1346-8138.12961 (2015). [DOI] [PubMed] [Google Scholar]
- 4.Kanellis, J. & Kang, D. H. Uric acid as a mediator of endothelial dysfunction, inflammation, and vascular disease. Semin. Nephrol.25(1), 39–42. 10.1016/j.semnephrol.2004.09.007 (2005). [DOI] [PubMed] [Google Scholar]
- 5.Nemati, H. et al. Antioxidant status in patients with psoriasis. Cell Biochem. Funct.32(3), 268–273. 10.1002/cbf.3011 (2014). [DOI] [PubMed] [Google Scholar]
- 6.Ruggiero, C. et al. Uric acid and inflammatory markers. Eur. Heart J.27(10), 1174–1181. 10.1093/eurheartj/ehi879 (2006). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Nishizawa, H., Maeda, N. & Shimomura, I. Impact of hyperuricemia on chronic kidney disease and atherosclerotic cardiovascular disease. Hypertens. Res.45(4), 635–640. 10.1038/s41440-021-00840-w (2022). [DOI] [PubMed] [Google Scholar]
- 8.Li, X. et al. Association of serum uric acid levels in psoriasis: A systematic review and meta-analysis. Medicine95(19), e3676. 10.1097/MD.0000000000003676 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Zhang, Y. et al. Updated evidence of the association between elevated serum uric acid level and psoriasis. Front. Med.8, 645550. 10.3389/fmed.2021.645550 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Hu, S. C., Lin, C. L. & Tu, H. P. Association between psoriasis, psoriatic arthritis and gout: A nationwide population-based study. J. Eur. Acad. Dermatol. Venereol.33(3), 560–567. 10.1111/jdv.15290 (2019). [DOI] [PubMed] [Google Scholar]
- 11.Kwon, H. H., Kwon, I. H., Choi, J. W. & Youn, J. I. Cross-sectional study on the correlation of serum uric acid with disease severity in Korean patients with psoriasis. Clin. Exp. Dermatol.36(5), 473–478. 10.1111/j.1365-2230.2010.03988.x (2011). [DOI] [PubMed] [Google Scholar]
- 12.Gui, X. Y., Jin, H. Z., Wang, Z. J. & Xu, T. D. Serum uric acid levels and hyperuricemia in patients with psoriasis: A hospital-based cross-sectional study. Ann. Bras. Dermatol.93(5), 761–763. 10.1590/abd1806-4841.20187547 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Lai, Y. C. & Yew, Y. W. Psoriasis and uric acid: A population-based cross-sectional study. Clin. Exp. Dermatol.41(3), 260–266. 10.1111/ced.12781 (2016). [DOI] [PubMed] [Google Scholar]
- 14.Kurd, S. K. & Gelfand, J. M. The prevalence of previously diagnosed and undiagnosed psoriasis in US adults: Results from NHANES 2003–2004. J. Am. Acad. Dermatol.60(2), 218–224. 10.1016/j.jaad.2008.09.022 (2009). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Skrivankova, V. W. et al. Strengthening the reporting of observational studies in epidemiology using mendelian randomization: The STROBE-MR statement. JAMA326(16), 1614–1621. 10.1001/jama.2021.18236 (2021). [DOI] [PubMed] [Google Scholar]
- 16.Lawlor, D. A., Harbord, R. M., Sterne, J. A., Timpson, N. & Davey Smith, G. Mendelian randomization: Using genes as instruments for making causal inferences in epidemiology. Stat. Med.27(8), 1133–1163. 10.1002/sim.3034 (2008). [DOI] [PubMed] [Google Scholar]
- 17.Boehm, F. J. & Zhou, X. Statistical methods for Mendelian randomization in genome-wide association studies: A review. Comput. Struct. Biotechnol. J.20, 2338–2351. 10.1016/j.csbj.2022.05.015 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.VanderWeele, T. J., Tchetgen Tchetgen, E. J., Cornelis, M. & Kraft, P. Methodological challenges in Mendelian randomization. Epidemiology.25(3), 427–435. 10.1097/EDE.0000000000000081 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Zhan, J., Tang, X., Wang, F. & Han, J. Association between daily dietary eicosatetraenoic acid intake and the lower risk of psoriasis in American adults. Clin. Cosmet. Inv. Derm.14, 1541–1549. 10.2147/CCID.S333288 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Johnson, C. L. et al. National health and nutrition examination survey: Analytic guidelines, 1999–2010. Vital. Health Stat.2(161), 1–24 (2013). [PubMed] [Google Scholar]
- 21.Whelton, P. K. et al. 2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA Guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: Executive summary: A report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Hypertension71(6), 1269–1324. 10.1161/HYP.0000000000000066 (2018). [DOI] [PubMed] [Google Scholar]
- 22.Wang, S. et al. Cobalamin intake and related biomarkers: Examining associations with mortality risk among adults with type 2 diabetes in NHANES. Diabetes Care.45(2), 276–284. 10.2337/dc21-1674 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Grundy, S. M. et al. Diagnosis and management of the metabolic syndrome: An American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation112(17), 2735–2752. 10.1161/CIRCULATIONAHA.105.169404 (2005). [DOI] [PubMed] [Google Scholar]
- 24.Hartwig, F. P., Davey Smith, G. & Bowden, J. Robust inference in summary data Mendelian randomization via the zero modal pleiotropy assumption. Int. J. Epidemiol.46(6), 1985–1998. 10.1093/ije/dyx102 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Burgess, S., Thompson, SG. & CRP CHD genetics collaboration. Avoiding bias from weak instruments in Mendelian randomization studies. Int. J. Epidemiol.40(3), 755–764. 10.1093/ije/dyr036 (2011). [DOI] [PubMed]
- 26.Lei, H. et al. The effects of unsaturated fatty acids on psoriasis: A two-sample Mendelian randomization study. Food Sci. Nutr.11(10), 6073–6084. 10.1002/fsn3.3543 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Kamiya, K., Kishimoto, M., Sugai, J., Komine, M. & Ohtsuki, M. Risk factors for the development of psoriasis. Int. J. Mol. Sci.20(18), 4347. 10.3390/ijms20184347 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Greco, M. F. D., Minelli, C., Sheehan, N. A. & Thompson, J. R. Detecting pleiotropy in Mendelian randomisation studies with summary data and a continuous outcome. Stat. Med.34(21), 2926–2940. 10.1002/sim.6522 (2015). [DOI] [PubMed] [Google Scholar]
- 29.Verbanck, M., Chen, C. Y., Neale, B. & Do, R. Detection of widespread horizontal pleiotropy in causal relationships inferred from Mendelian randomization between complex traits and diseases. Nat. Genet.50(5), 693–698. 10.1038/s41588-018-0099-7 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Inaba, S., Sautin, Y., Garcia, G. E. & Johnson, R. J. What can asymptomatic hyperuricaemia and systemic inflammation in the absence of gout tell us?. Rheumatology52(6), 963–5. 10.1093/rheumatology/ket001 (2013). [DOI] [PubMed] [Google Scholar]
- 31.Brovold, H. et al. Crystallized but not soluble uric acid elicits pro-inflammatory response in short-term whole blood cultures from healthy men. Sci. Rep.9(1), 10513. 10.1038/s41598-019-46935-w (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Goldman, M. Uric acid in the etiology of psoriasis. Am. J. Dermatopath.3(4), 397–404. 10.1097/00000372-198100340-00014 (1981). [DOI] [PubMed] [Google Scholar]
- 33.Hagino, T., Saeki, H., Fujimoto, E. & Kanda, N. Effectiveness and safety of deucravacitinib treatment for moderate-to-severe psoriasis in real-world clinical practice in Japan. J. Dermatol. Treat.35(1), 2307489. 10.1080/09546634.2024.2307489 (2024). [DOI] [PubMed] [Google Scholar]
- 34.Hagino, T., Saeki, H., Fujimoto, E. & Kanda, N. Effects of biologic therapy on laboratory indicators of cardiometabolic diseases in patients with psoriasis. J. Clin. Med.12(5), 1934. 10.3390/jcm12051934 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Merola, J. F. et al. Effect of secukinumab on traditional cardiovascular risk factors and inflammatory biomarkers: Post hoc analyses of pooled data across three indications. Rheumato. Therapy9(3), 935–955. 10.1007/s40744-022-00434-z (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Li, T., Walsh, J. R., Ghishan, F. K. & Bai, L. Molecular cloning and characterization of a human urate transporter (hURAT1) gene promoter. Biochim. Biophys. Acta.1681(1), 53–58. 10.1016/j.bbaexp.2004.10.001 (2004). [DOI] [PubMed] [Google Scholar]
- 37.Fordham, J. N. & Storey, G. O. Psoriasis and gout. Postgrad. Med. J.58(682), 477–480. 10.1136/pgmj.58.682.477 (1982). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Walton, R., Block, W. D. & Heyde, J. A comparative study of uric acid values of whole blood in patients with psoriasis and other dermatoses. J. Invest. Dermatol.37, 125–33. 10.1038/jid.1961.93 (1961). [DOI] [PubMed] [Google Scholar]
- 39.Merola, J. F., Wu, S., Han, J., Choi, H. K. & Qureshi, A. A. Psoriasis, psoriatic arthritis and risk of gout in US men and women. Ann. Rheum. Dis.74(8), 1495–1500. 10.1136/annrheumdis-2014-205212 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Yilmaz, E., Tamer, E., Artüz, F., Külcü Çakmak, S. & Köktürk, F. Evaluation of serum uric acid levels in psoriasis vulgaris. Turk. J. Med. Sci.47(2), 531–534. 10.3906/sag-1512-5 (2017). [DOI] [PubMed] [Google Scholar]
- 41.Lai, T. L., Yim, C. W., Wong, P. Y., Leung, M. C. & Ng, W. L. Hyperuricemia in Asian psoriatic arthritis patients. Int. J. Rheum. Dis.21(4), 843–849. 10.1111/1756-185X.13265 (2018). [DOI] [PubMed] [Google Scholar]
- 42.Chandran, V. & Raychaudhuri, S. P. Geoepidemiology and environmental factors of psoriasis and psoriatic arthritis. J. Autoimmun.34(3), J314-321. 10.1016/j.jaut.2009.12.001 (2010). [DOI] [PubMed] [Google Scholar]
- 43.Sheehan, N. A., Didelez, V., Burton, P. R. & Tobin, M. D. Mendelian randomisation and causal inference in observational epidemiology. PLoS Med.5(8), e177. 10.1371/journal.pmed.0050177 (2008). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Brion, M. J., Shakhbazov, K. & Visscher, P. M. Calculating statistical power in Mendelian randomization studies. Int. J. Epidemiol.42(5), 1497–501. 10.1093/ije/dyt179 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.Pierce, B. L., Ahsan, H. & Vanderweele, T. J. Power and instrument strength requirements for Mendelian randomization studies using multiple genetic variants. Int. J. Epidemiol.40(3), 740–52. 10.1093/ije/dyq151 (2011). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Bowden, J. et al. Assessing the suitability of summary data for two-sample Mendelian randomization analyses using MR-Egger regression: The role of the I2 statistic. Int. J. Epidemiol.45(6), 1961–1974. 10.1093/ije/dyw220 (2016). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Rajasundaram, S. et al. Independent effects of blood pressure on intraocular pressure and retinal ganglion cell degeneration: A Mendelian randomization study. Invest. Ophthalmol. Vis. Sci.65(8), 35. 10.1167/iovs.65.8.35 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Jiang, T., Gill, D., Butterworth, A. S. & Burgess, S. An empirical investigation into the impact of winner’s curse on estimates from Mendelian randomization. Int. J. Epidemiol.52(4), 1209–1219. 10.1093/ije/dyac233 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are obtained from public databases. The SUA data for European participants and East Asian participants are available through the database at http://ftp.ebi.ac.uk/pub/databases/gwas/summary_statistics/GCST90018001-GCST90019000/GCST90018977/ and https://gwas.mrcieu.ac.uk/datasets/bbj-a-57/, respectively. The psoriasis summary statistics in the MR analysis are downloaded from the FinnGen database (https://storage.googleapis.com/finngen-public-data-r10/summary_stats/finngen_R10_L12_PSORIASIS.gz) and GWAS Catalog (http://ftp.ebi.ac.uk/pub/databases/gwas/summary_statistics/GCST90018001-GCST90019000/GCST90018687/). Further inquiries can be directed to the first author through email xaiu21063@xaiu.edu.cn.