Possible Genetic Links Between Hidradenitis Suppurativa and CAD, Diabetes Risks

Higher risks for developing coronary artery disease (CAD) and diabetes, and of having an altered plasma proteome, were seen among a group of older adults who also had a higher genetic susceptibility to hidradenitis suppurativa (HS), according to publicly available summary statistics and individual genotype data from the UK Biobank. Further, HS polygenic risk score (PRS) had a positive correlation with the altered expression of 58 plasma proteins, including those implicated in inflammatory and metabolic processes, the authors explained.

Although PRS has its limitations—including that it should not be used to diagnose or screen,^1,2 and that its accuracy depends on patient DNA similarity to the DNA used to develop the PRS²—research also backs its utility as a promising tool for risk prediction and determining genetic associations.^3,4

The results of this investigation were recently published in JAMA Dermatology,⁵ and they encompassed 391,418 adults of European ancestry—defined by the authors as a combination of self-reported ethnicity and the first 6 genotype principal components—with a median (IQR) age of 58 (51-64) years. Study enrollment took place between January 1, 2006, and December 31, 2010. Follow-up concluded January 1, 2023. The investigators used logistic regression to quantify CAD and diabetes diagnoses, and Cox proportional hazards regression models to quantify incident disease by adjusting for sex, age, body mass index (BMI), and smoking status. Most participants (53%) were female patients.

“Large-scale genome-wide association studies have advanced our understanding of the genetic underpinnings of hidradenitis suppurativa,” the authors wrote, “but the affected genetic pathways are still not fully elucidated.”

During this study's follow-up, 26,994 participants had a first incident CAD and 13,153 developed diabetes. | Image Credit: © syahrir-stock.adobe.com

Patients were placed in genetic risk cohorts that ranged from low (<25th PRS percentile) to mid (25th-49th and 50th-74th PRS percentile) to high (≥75th PRS percentile) genetic risk. CAD was seen in at least 5% of each of these groups, diabetes was seen in at least 4.2% of each group, overall total cholesterol was 220 (85) mg/dL, overall high-density lipoprotein cholesterol was 56.37 (14.67) mg/dL, overall low-density lipoprotein cholesterol was 137.84 (33.59) mg/dL, and overall triglycerides were 154.87 mg/dL. The mean (SD) blood pressure was 140 (19.6)/82.1 (10.6) mm Hg. As for smoking status, 54% had never smoked, 35% were current smokers, and 10% had a history of smoking.

Overall Results

Positive genetic correlations were seen between common genetic variants for HS and the following:

• CAD: genetic r, 0.25 (P = 5.59 x 10⁻⁵)
• Diabetes: genetic r, 0.30 (P = 4.19 x 10⁻⁶)
• Triglyceride plasma levels: genetic r, 0.20 (P = 1.81 x 10⁻⁴)
• C reactive protein plasma levels: genetic r, 0.31 (P = 1.49 x 10⁻⁴)

A negative correlation was seen with high-density lipoprotein cholesterol (genetic r = −0.21; P = 2.78 x 10⁻⁴).

Further, each 1 SD increase in PRS for HS meant there was a 25% greater risk of the chronic inflammatory skin condition (OR, 1.25; 95% CI, 1.09-1.43; P = .002), when the investigators used an independent data set of 1209 cases and 432,686 controls. Next, using 213 unrelated individuals and a reference model that adjusted for sex, age, BMI, and the first 6 genotype principal components, the investigators saw an area under a receiver operating characteristic curve (AUC)—used to evaluate model performance—of 0.753 (95% CI, 0.720-0.786), which rose to 0.761 (95% CI, 0.729-0.794; P = .03 for difference) when they added HS to the model.

An additional model that adjusted for sex, age, and the first 6 genotype principal components determined that per 1 SD increase in PRS for HS, there was a 0.14-unit (95% CI, 0.13-0.16; P < .001) increase in BMI.

The median follow-up for this study was 13.7 (12.8-14.5) years, and during that time, among the 26,994 individuals who had a first incident CAD; the 13,153 diagnosed with incident diabetes; and the 19,911 who died, rates of CAD and diabetes increased as the risk of HS increased. For example, compared with those in a PRS percentile less than 25 considered at low genetic risk of HS, those in the 25th-49th and 50th-74th PRS percentiles had a 5% greater risk of CAD (both OR, 1.05; 95% CI, 1.02-1.08) and those in a PRS percentile of 75th or greater had a 9% greater risk of CAD (OR, 1.09; 95% CI, 1.06-1.12). The risks were more varied for diabetes. Those in the 75th or greater PRS percentile of HS had a 13% greater risk (OR, 1.13; 95% CI, 1.10-1.17) of diabetes, those in the 50th-74th PRS percentile had a 10% greater risk (OR, 1.10; 95% CI, 1.06-1.13), and those in the 25th-49th percentile had a 5% greater risk (OR, 1.05; 95% CI, 1.02-1.09) compared with those in the less than 25th PRS percentile.

The overall cumulative incidence of CAD over the first 15 years of follow-up was highest among those in the high-risk PRS percentile compared with the low-risk PRS percentile, at 8.22% (95% CI, 8.03%-8.42%) vs 7.81% (95% CI, 7.62%-8.01%). Results trended similarly for diabetes, at 4.17% (95% CI, 4.02%-4.31%) vs 3.66% (95% CI, 3.53%-3.79%), respectively.

Principal strengths of this research and its findings are that they echo previous research on links between HS and higher risks of CAD and death attributable to cardiovascular disease vs healthy controls, as well as a higher risk of diabetes for those who have HS. The authors point to potential for the 3 conditions to share pathophysiologic pathways, noting that “obesity is often observed in individuals with HS and is a common risk factor for cardiometabolic disease.” Limitations are the potential for healthy volunteer bias due to the general makeup of the UK Biobank cohort and that the PRS was not sufficiently powered to capture true genetic associations.

The authors say they recommend that future studies delve more into which proteins might hold potential as targets for treatment when they are seen to be altered in the plasma proteome.

References

1. Cambridge University: Cardiovascular Epidemiology Unit YouTube page. Accessed November 7, 2024. https://www.youtube.com/watch?v=BqR_G8DnPJw

2. Polygenic risk scores: how useful are they? Genomics Education Program. October 25, 2018. Accessed November 7, 2024. https://www.genomicseducation.hee.nhs.uk/blog/polygenic-risk-scores-how-useful-are-they/

3. Lambert SA, Abraham G, Inouye M. Towards clinical utility of polygenic risk scores. Hum Mol Genet. 2019;28(R2):R133-R142. doi:10.1093/hmg/ddz187

4. Sun L, Pennells L, Kaptoge S, et al. Polygenic risk scores in cardiovascular risk prediction: a cohort study and modelling analyses. PLoS Med. 2021;18(1):e1003498. doi:10.1371/journal.pmed.1003498

5. Nielsen VW, Vad OB, Holgersen N, et al. Genetic susceptibility to hidradenitis suppurativa and predisposition to cardiometabolic disease. JAMA Dermatol. Published online October 9, 2024. doi:10.1001/jamadermatol.2024.3779