Non-alcoholic fatty liver disease (NAFLD) is one of the most prevalent chronic liver diseases in our setting and, in most cases, is closely linked to metabolic risk factors such as obesity, diabetes mellitus, or dyslipidemia.1 The prevalence of NAFLD has increased in recent years, reaching approximately 25–30% in the general population.2,3

NAFLD occurs 1.5–3 times more frequently in psoratic patients, with prevalences of up to 65% in some cohorts.1 Of note, this association is more evident in patients with severe psoriasis, regardless of other risk factors.1

Mantovani et al. proposed a hepato-cutaneous axis, where hepatic macrophages and psoriatic skin T cells produce pro-inflammatory cytokines that are released into the systemic circulation, playing a significant role in the pathogenesis of both diseases.4

Hepatic steatosis is diagnosed in the presence of fat deposition in ≥5% of hepatocytes. Although the gold standard test for diagnosis is histological examination through a liver biopsy, given that it is an invasive method and associates a high consumption of resources, it is necessary to look for other diagnostic methods for routine clinical practice. Transient elastography and the controlled attenuation parameter (CAP) have shown a good correlation with histological study.5

Imaging modalities such as abdominal ultrasound, magnetic resonance imaging, or hepatic transient elastography are a non-invasive alternative for the diagnosis of hepatic steatosis.6 However, these modalities are not usually available in the Dermatology consultation, and serological indices are useful for the initial assessment of the patient in the consultation. In fact, the NAFLD Fibrosis Score (NFS), FIB-4, or AST-to-Platelet Ratio Index (APRI) are widely used in the routine clinical practice as predictors of fibrosis in patients with NAFLD.7 However, non-invasive indices developed to diagnose steatosis, such as the Hepatic Steatosis Index (HSI) or the Fatty Liver Index (FLI), can be influenced by the presence of inflammation or fibrosis and are not as widely used, as they do not provide great advantages vs imaging modalities and routine blood tests.5

In this regard, 2 recent studies conducted in the United States evaluated the association of the TyG index with NAFLD and its predictive value. Their findings suggested that the TyG index along with anthropometric parameters is useful for the early diagnosis of NAFLD.8,9 However, these indices have not been specifically evaluated in psoriatic patients.

The main objective of this study is to evaluate and compare the diagnostic validity of the TyG index along with different anthropometric parameters in the diagnosis of NAFLD in patients with moderate/severe psoriasis eligible for systemic treatment.

A total of 154 patients were initially recruited consecutively. Of these, a total of 5 participants were excluded from the final statistical analysis for meeting any of the exclusion criteria: 3 patients for chronic hepatitis C, 1 for autoimmune hepatitis, and 1 for primary biliary cholangitis. A total of 149 patients were included in the final statistical analysis.

Diagnostic validity analysis of TyG index and its combinations with anthropometric measurements

The TyG index along with BMI obtained the highest area under the curve, reaching 0.852; 95%CI (Fig. 2; Table 2). The optimal cutoff point for diagnosis was calculated with the Youden index, resulting in 243.27 (sensitivity, 0.802; specificity, 0.792). The results obtained for the AUC of the TyG index and the TyG indices along with anthropometric parameters were similar in both male and female groups (Figs. 3 and 4; Tables 3 and 4), although the AUC was slightly higher for all indices in the case of women.

Figure 2.

Receiver operating characteristic (ROC) curve of the triglyceride and glucose index (TyG); triglycerides and glucose combined with body mass index (TyG-BMI); triglycerides and glucose combined with waist circumference (TyG-WC); triglycerides and glucose combined with waist-to-height ratio (TyG-HtR) of the sample.

(0.09MB).

Table 2.

Area under the curve of the triglyceride and glucose index (TyG); triglycerides and glucose combined with body mass index (TyG-BMI); triglycerides and glucose combined with waist circumference (TyG-WC); triglycerides and glucose combined with waist-to-height ratio (TyG-HtR) of the sample.

Area under the curve			95% confidence interval
Studied indices	Area	Asymptotic significance	Lower limit	Upper limit
TyG index	0.688	0.000	0.600	0.775
TyG-BMI index	0.852	0.000	0.789	0.914
TyG-WC index	0.832	0.000	0.764	0.899
TyG-HtR index	0.834	0.000	0.767	0.901

Figure 3.

Receiver operating characteristic (ROC) curve of the triglyceride and glucose index (TyG); triglycerides and glucose combined with body mass index (TyG-BMI); triglycerides and glucose combined with waist circumference (TyG-WC); triglycerides and glucose combined with waist-to-height ratio (TyG-HtR) of the women in the sample.

(0.08MB).

Figure 4.

Receiver operating characteristic (ROC) curve of the triglyceride and glucose index (TyG); triglycerides and glucose combined with body mass index (TyG-BMI); triglycerides and glucose combined with waist circumference (TyG-WC); triglycerides and glucose combined with waist-to-height ratio (TyG-HtR) of the men in the sample.

(0.09MB).

Table 3.

Area under the curve of the triglyceride and glucose index (TyG); triglycerides and glucose combined with body mass index (TyG-BMI); triglycerides and glucose combined with waist circumference (TyG-WC); triglycerides and glucose combined with waist-to-height ratio (TyG-HtR) of the women in the sample.

Area under the curve			95% confidence interval
Studied indices	Area	Asymptotic significance	Lower limit	Upper limit
TyG index	0.725	0.002	0.599	0.851
TyG-BMI index	0.880	0.000	0.795	0.966
TyG-WC index	0.850	0.000	0.756	0.943
TyG-HtR index	0.859	0.000	0.768	0.951

Table 4.

Area under the curve of the triglyceride and glucose index (TyG); triglycerides and glucose combined with body mass index (TyG-BMI); triglycerides and glucose combined with waist circumference (TyG-WC); triglycerides and glucose combined with waist-to-height ratio (TyG-HtR) of the men in the sample.

Area under the curve			95% confidence interval
Studied indices	Area	Asymptotic significance	Lower limit	Upper limit
TyG index	0.639	0.044	0.512	0.765
TyG-BMI index	0.823	0.000	0.730	0.915
TyG-WC index	0.807	0.000	0.706	0.909
TyG-HtR index	0.821	0.000	0.730	0.916

A ROC curve was calculated to assess the discriminatory capacity of the index in relation to hepatic fibrosis in patients with hepatic steatosis. An area under the curve of 0.598; 95%CI (p>0.05) was obtained. The optimal cutoff point was calculated using the Youden index, 260.99 (sensitivity, 0.824; specificity, 0.4).

As far as we know, this work is a pioneer in evaluating the diagnostic validity of TyG indices along with anthropometric parameters for the diagnosis of NAFLD in psoriatic patients. The results obtained for the TyG-BMI, TyG-WC, and TyG-HtR indices showed an AUC>0.83. Therefore, these indices could be a useful tool in the screening of NAFLD in psoriatic patients.

NAFLD is the leading cause of liver disease in our setting and has significant consequences on the vital prognosis of affected patients. Up to 30% will develop steatohepatitis, and of these, up to 20% will develop liver cirrhosis.12 NAFLD is the leading cause of liver transplantation in the United States in women and in elderly patients. In addition, it is one of the most frequent risk factors for hepatocellular carcinoma in Latin America.13 Considering that this disease is more prevalent in psoriatic patients vs the general population and the pathophysiological links they share, its evaluation within the comprehensive management of the psoriatic patient is recommended.

A total of 64% of our sample had hepatic steatosis, and 18% of the patients with steatosis had significant hepatic fibrosis. These results are consistent with the prevalence of hepatic steatosis in psoriatic patients reported in other cohorts, as well as with what has been described in other immune-mediated diseases such as inflammatory bowel disease1,14. In addition, the leading cause of mortality in patients with NAFLD is cardiovascular, so the detection of NAFLD in psoriatic patients should prompt us to rule out the presence of other cardiovascular risk factors.15

The main finding of this study was the AUC of the different TyG indices along with anthropometric parameters. The TyG index along with BMI showed the highest AUC, with a value of 0.852; 95%CI (0.789–0.914) for our entire cohort. These results were consistently reproduced in the stratified analysis by sex, with the TyG-BMI index having the highest AUC in both women and men. Our results contrast with those of other studies conducted in the US population, where the TyG-WC index had a higher AUC.9 Our work and the work conducted by Xue et al. agree that TyG indices combined with anthropometric parameters are an accurate method for the initial detection of hepatic steatosis. However, these indices had not been previously evaluated in psoriatic patients. In our study, both the TyG-WC and TyG-HtR indices showed AUC values>0.8. Although the best index may vary depending on the population, we know that if we combine the TyG index with the proposed anthropometric values, we will be able to screen patients with some certainty.

We did not perform any cost-effectiveness analysis to calculate the possible impact of the implementation of any of these indices as a screening method for hepatic steatosis. However, we can affirm that there was no added cost to the collection of data beyond that of routine clinical practice. To calculate the TyG index, only a blood test plus a liver and lipid profile are required, which is part of the routine follow-up of patients with moderate/severe psoriasis. Anthropometric measurements also do not involve any cost or require sophisticated technical support.

Regarding the limitations of our study, firstly, instead of using histopathology as the gold standard for the diagnosis of NAFLD and hepatic fibrosis, transient elastography was used. However, the latest generation transient elastography devices have proven to be effective in detecting steatosis>10%, in addition to correlating well with the severity of the condition. It is a simple, fast, comfortable, and reproducible technique. A 2019 study identified that the optimal cutoff point for diagnosing steatosis is 248dB/m.11

This work presents other limitations, inherent to its observational and cross-sectional design, such as the impossibility of establishing causality or studying the prognosis over time. In addition, the alteration of the TyG index could be related to the presence of other components of metabolic syndrome, not only to the presence of NAFLD per se. On the other hand, consecutive patients participated voluntarily in this study, which could imply a selection bias, so the sample in our study may not be representative of the population with moderate/severe psoriasis. Furthermore, this study did not have a control group of psoriatic patients, which limits the ability to make comparisons with healthy populations. The small sample size of patients with significant hepatic fibrosis may have also influenced the results.

Regarding the strengths of our study, firstly, transient elastography is used as the reference method, which to date has proven to be a useful, as well as widely available, inexpensive, and reproducible tool. Secondly, this is a pioneering study in investigating the role of the TyG index and its variants in psoriatic patients, with results similar to or better than other serological indices that have not been specifically evaluated in the psoriasis population.16,17 Thirdly, the prevalence of NAFLD obtained in our sample is consistent with other studies published in the literature and supports the need to develop simple and widely available strategies for its screening. Finally, our study evaluated and compared the AUC of the TyG index and the TyG index along with anthropometric parameters, not only in the general sample but also by subgroups of women and men, obtaining consistent results in the different subgroups.

The TyG indices combined with anthropometric parameters showed an AUC>0.83, with the TyG-BMI index obtaining the highest AUC, with a value of 0.85. The calculation of the indices did not involve an added cost to routine clinical practice, so it could be an effective and efficient screening system for the detection of NAFLD in psoriatic patients.

None declared.

None declared.