In 2018, it is estimated that about 25% of the world population has NAFLD. Several studies have investigated the progression of NAFLD and its subtypes. Although most studies have concluded that non-alcoholic steatohepatitis (NASH) is the progressive form of NAFLD, recent data indicates that a small subset of NAFLD, those with no histologic features of NASH, can progress into advanced liver disease [31].
Adiponectin is the most abundant peptide secreted by adipocytes as well as other cell types including skeletal, cardiac, and endothelial cells. It can protect the liver from injury by enhancing fatty acid β-oxidation, thus reducing hepatic triacylglycerol content and hepatic insulin resistance. Adiponectin has an anti-inflammatory role by inhibiting the production of pro-inflammatory cytokines such as interleukin-6 (IL-6) and tumor necrosis factor-a (TNF-a) in macrophages and/or reduces their phagocytic capacity. This may be a paramount feature in reversing metabolic dysfunction. Adiponectin potentially has a protective role in insulin resistance (IR) and pathophysiologic procedure of type 2 diabetes mellitus (T2DM). Reduction in adiponectin level plays a central role in obesity-related diseases, including IR/T2DM and cardiovascular disease [33].
In this study, a highly significant correlation was found between adiponectin level and liver fibrosis. We found that adiponectin levels were significantly lower in the high-grade fibrosis group with a P-value of 0.000. Similar findings were reported by Nazal et al. [25] who concluded that low adiponectin levels are associated with more severe liver fibrosis in a study that was conducted on seventy (70) obese patients and sixty-nine (69) controls. This can be attributed to the hepatoprotective and antifibrogenic effects of adiponectin in liver injury. It has anti-inflammatory action by neutralizing TNF-α and antifibrotic action by inhibition of HSC proliferation and migration.
On the contrary, Lucero et al. [22] concluded that adiponectin levels were lower in NAFLD patients, but there was no significant relation between adiponectin and the degree of liver fibrosis. That study was conducted on sixty patients with metabolic syndrome of which thirty-six (36) were biopsy-proven NAFLD and twenty-four (24) with no even ultrasound NAFLD evidence. This different result may be due to the difference in the type of patients included in the study.
In addition, adiponectin level showed a highly significant correlation with obesity. Its mean values in both overweight and obese patients were 4.68 ± 1.09 and 3.33 ± 1.21, respectively, with a P-value of 0.000. This was consistent with Zhang et al. [34] who found a significant negative correlation between adiponectin and obesity (P < 0.01) in a prospective study that was done on a large scale.
This can be attributed to the disproportional accumulation of white adipose tissue in overweight and obese patients which is accompanied by a generalized change in the circulating levels of adiponectin [6].
This study showed that HTN has a significant negative correlation with adiponectin levels and a significant positive correlation with the grade of liver fibrosis (P = 0.01). As discussed before, patients with high-grade liver fibrosis have a state of hypo-adiponectinemia, which means loss of the protective role of adiponectin on vascular function.
Highly significant negative correlation was found between SBP and adiponectin levels with a P-value of 0.003. Significant negative correlation was also found between DBP and adiponectin levels with a P-value of 0.023. This agreed with Brzeska et al. [4] who found that plasma levels of adiponectin were significantly lower in the patients with hypertension (P = 0.0026).
Also, Baden et al. [1] has found that adiponectin showed a significant negative correlation with systolic and diastolic blood pressure (P < 0.01). This can be explained by the fact that adiponectin attenuates the phenotype of M1 macrophage which displays upregulation of pro-inflammatory cytokines including TNF-α, IL-6, and monocyte chemotactic protein-1 (MCP-1), and it promotes the phenotype of M2 macrophage which upregulates arginase-1 (ARG-1) and interleukin-10 (IL-10). In addition, adiponectin improves the endothelial cell function via increasing nitric oxide (NO) and prostaglandin I2 (PGI2) production, thus exerts a protective action on vascular function through its ability to improve the function of macrophage and endothelial cells [28].
In the present study, a highly significant correlation was found between adiponectin levels and DM with P-values of 0.000 being lower in diabetic compared to non-diabetic patients and a highly significant positive correlation between diabetes and high-grade fibrosis with a P-value of 0.000.
This was in agreement with Davis et al. [10], who stated that adiponectin level was negatively correlated with T2DM (P = 0.04) and Jaafar et al. [15] who found that diabetic patients presented with more severe fibrosis than non-diabetic NAFLD patients (P < 0.001).
T2DM patients can present with high-grade liver fibrosis that could be attributed to glucose intolerance, insulin resistance, and increased serum free fatty acids; there is ectopic fat deposition in visceral organs, notably the liver. Subsequently, fatty liver status increases the liver vulnerability to oxidative stress and mutations, and hence increased inflammation, leading to liver fibrosis [24].
This can be attributed to the role of adiponectin in regulating the metabolism of both glucose and lipid in the liver and has been implicated in inhibiting gluconeogenesis, as well as activating fatty acid oxidation and glycolytic pathways which overall improves insulin sensitivity [30].
As for lipid profile, we found that there was a highly significant negative correlation between adiponectin concentration and total cholesterol as well as TG (P < 0.001), while there was a highly significant positive correlation between adiponectin and HDL levels (P < 0.001). We found that total cholesterol, LDL, and triglycerides were higher among the high-grade fibrosis group.
This was found to be consistent with Brzeska et al. [4], who reported a significant negative correlation between TG and adiponectin level (P = 0.025), while a highly significant positive correlation between adiponectin and HDL levels (P = 0.0001), but they reported no correlation between adiponectin and total cholesterol levels. Also, Maghsoudi et al. [23] found a significant negative association between adiponectin and TG as well as LDL levels (P < 0.05) and a significant positive correlation with HDL (P < 0.05).
This agreed with Kumar et al. [20] who found that high stages of liver fibrosis correlated with a significant increase in levels of serum total cholesterol (P-value 0.005), TG (P-value 0.002), and LDL (P-value 0.001), while an associated significant decrease in HDL (P-value 0.001) in NAFLD patients. As explained before, patients with high-grade liver fibrosis have a state of hypoadiponectinemia which has an important role in dyslipidemia.
This is explained by the role of adiponectin in lipid metabolism. It reduces the secretion of hepatic apolipoprotein E and apolipoprotein B from the liver. Also, it increases insulin activity, improves glucose tolerance, and plays an important role in fatty acid oxidation; thus, it prevents dyslipidemia [14].
In the current study, more patients with low-grade liver fibrosis were found among the overweight group compared to the obese group (92.95% compared to 63.6%), while more high-grade fibrosis was found among the obese group compared to the overweight group (36.4% compared to 7.1%). Hence, high-grade fibrosis was positively correlated with obesity in a statistically significant correlation with a P-value of 0.01. This was consistent with Caballería et al. [7] who reported that liver fibrosis in NAFLD patients is correlated with obesity and Ratziu et al. [26] who found that liver fibrosis is significantly correlated with BMI (P = 0.01).
This can be explained by the increased visceral fat in obese patients that has a primary role in the pathogenesis of NAFLD. Liver fat accumulation (steatosis) is largely dependent on recirculated free fatty acids (FFA) from the adipose tissue pool. Steatosis leads to lipotoxicity which causes apoptosis, necrosis, generation of oxidative stress, and inflammation. The resulting chronic injury activates a fibrogenic response [11].
In addition, worth noting is another highly significant correlation found between relatively low platelet count and higher grades of liver fibrosis where platelet count were 191.7 ± 15.02 in this group compared to 238.31 ± 30.45 in the lower grade group with a P-value of 0.000. This agrees with Cao et al. [8] who found that platelet count negatively correlated with the stage of fibrosis in NAFLD patients.
The peripheral platelet production is mainly regulated by thrombopoietin, which is a glycoprotein hormone predominantly synthesized in the liver. During the development and progression of NAFLD, excessive lipid deposition and oxidative stress could impair the mitochondrial functions through inflammatory mediators and then affected thrombopoietin synthesis. Reduced platelet counts could happen in this process [21].
Another similar significant correlation with a P-value of 0.001 was established between relatively lower albumin levels and high-grade fibrosis (4.1 ± 0.21) compared to low-grade fibrosis (4.44 ± 0.29). This was consistent with Kumar et al. [20] who stated that liver fibrosis is inversely related to serum albumin level in NAFLD patients (P = 0.004).
We found that obesity was common among diabetic compared to non-diabetic patients (58.6% compared to 23.8%) as expected with a P-value of 0.014 indicating a significant correlation. Consequently, higher BMI values were reported among diabetic patients compared to the non-diabetic group (30.56 ± 2.6 compared to 28.6 ± 1.81, respectively) with a highly significant correlation as indicated by a P-value of about 0.005. This was consistent with Bae et al. [2] who found that a higher BMI than 23.7 kg/m2 is associated with increased risk of incident diabetes (P = 0.02) in a prospective study over 8 years conducted on eight-thousand nine hundred patients. Also, this was in agreement with Riaz et al. [27] who found that BMI in diabetic patients was significantly higher than controls (P < 0.05).
In the current study, we found that there is a significant correlation between obesity and high total cholesterol and LDL levels with a P-value of 0.031 and 0.023, respectively, as well as with low HDL levels with a P-value of 0.04. These findings were consistent with Maghsoudi et al. [23] who found a significant positive correlation between BMI and total cholesterol and LDL (P < 0.05) and a significant negative correlation with HDL (P < 0.05). Also, this was in agreement with Hertelyova et al. [13] who found that there was a significant positive correlation between BMI and total cholesterol as well as LDL levels (P < 0.001), while there was a significant negative correlation between HDL levels and BMI (P < 0.001) in a study conducted on four-hundred nineteen patients.
This can be explained by obesity-related uncontrolled increased FFA release from adipose tissue via lipolysis that results in enhanced delivery of FFA to the liver. This leads to increased TG and very low-density lipoprotein (VLDL) production in the liver and inhibition of lipoprotein lipase in adipose tissue and skeletal muscle leading to hypertriglyceridemia. Moreover, the increased VLDL in the liver can inhibit lipolysis of chylomicrons, which also contributes to hypertriglyceridemia. The TG in VLDL is exchanged for cholesteryl esters from LDL and HDL by the cholesteryl ester transport protein, producing TG-rich LDL and HDL. The TG in the LDL and HDL is then hydrolyzed by hepatic lipase, producing both small, dense LDL and HDL [17].
We also found a highly significant correlation between diabetes and total cholesterol, LDL, and TG values which were significantly higher in diabetic compared to the non-diabetic group with P-values of 0.000 for all. Higher HDL levels were also reported among non-diabetic compared to the diabetic group (55.1 ± 14.58 compared to 42.67 ± 14.42) with a P-value of 0.004 indicating a highly significant correlation. Kolhar and Priyanka [18] found that dyslipidemia is highly prevalent in diabetics and in particular more prevalent in those with poorly controlled diabetes. Also, Bhowmik et al. [3] reported that diabetic patients had a significant correlation with high total cholesterol, LDL, and TG levels with P-values lower than 0.001 for each, while low HDL levels were significantly correlated with diabetes (P = 0.048).
The origins of dyslipidemia in diabetes are complex but derived from specific abnormalities in lipoprotein metabolism and abnormalities in insulin action.
In this study, we found that there were low platelet counts in the obese group compared to the overweight group with a P-value of 0.003 indicating a highly significant correlation. This did not come in line with Jamshidi and Seif [16] who found that there is a significant relationship between central and general adiposity and higher platelet count. That study was performed on a greater number of cases (four-hundred eighty-six patients) while our study was conducted on fifty patients only. Also, in that study, they excluded diabetic and hypertensive patients. In our study, 77.3% of our obese patients were diabetics, and 45.5% of them were hypertensive, while Coban et al. [9] found that there is no significant correlation between platelet count and obesity in a study that was conducted on one-hundred obese patient and one-hundred controls.
Another noted highly significant correlation was found between lower platelet count and diabetes with a P-value of about 0.000. This may be because of various factors such as high production and turnover rate in T2DM with diminished mean platelet survival [19].
This goes with Kshirsagar et al. [19] who reported decreased platelet count in diabetic compared to non-diabetic patients but was not statistically significant (P = 0.08), as well as with Buch et al. [5] and Hekimsoy et al. [12].
We also found relatively lower serum albumin levels in the diabetic compared to the non-diabetic group with a P-value of 0.000. This agrees with Riaz et al. [27] who found that serum albumin for diabetic individuals was lower than control subjects which was statistically significant correlation (P < 0.05). Mostly due to the fact that patients with T2DM with long duration and/or poor control may develop diabetic nephropathy where they lose albumin in urine.