Baseline characteristics of the study population
Basic characteristics of the population studied are presented in Table 1. This cross-sectional study included 811 elderly Iranian participants aged 60 years and above, consisting of 369 (45.5%) athletes and 442 (54.5%) non-athletes. The study population had an almost equal distribution of males (49.6%) and females (50.4%). The majority were aged 60-69 years (63.4%). Most were single/divorced/widowed (83.8%) and had 1-8 years of education (48.6%). Regarding lifestyle factors, 90% reported not smoking, while 12.6% consumed alcohol, and 22.9% reported substance addiction (with opium or hookah/pipe usage). The prevalence of medical conditions was high, with 92.2% having at least one condition, most commonly cardiovascular disease (56.6%), metabolic disease (34.5%), and gastrointestinal disease (26.9%).
In terms of anthropometrics, the majority of participants were either overweight (39.1%) or obese (25.5%) based on their body mass index. The median calorie intake was 2,203 kcal, with athletes consuming higher calories (2,330 kcal) compared to non-athletes (2,056 kcal). The median protein and carbohydrate intake were also higher in athletes. For liver enzyme levels, the median ALT was 15.3 U/L and AST was 18.0 U/L, with similar values between athletes and non-athletes. Other characteristics like depression, anxiety, sleep quality, and medication/supplement use are also described in the table.
Relationship between Energy and Macronutrient Intake with Liver Enzymes
The study revealed notable differences between elderly athletes and non-athletes in the associations of macronutrient intake with serum liver enzyme levels.
Energy and Macronutrient Intake with ALT in Elderly Athletes and Non-Athletes
The individual models for the athletes, non-athletes, and the entire sample indicated that a higher intake of calories is associated with increased levels of ALT. Notably, this association seems to be stronger among athletes compared to non-athletes. In the fully adjusted models (Model 2; Table 2), it was observed that for athletes, each 1-unit increase in calorie intake was linked to a 0.0020 (95% CI: 0.0006, 0.0035; p = 0.005) increase in ALT levels, while among non-athletes this association was weaker and not statistically significant (β = 0.0009, 95% CI: -0.0003, 0.0021; p = 0.140). Furthermore, in the fully adjusted model for all participants (Model 5; Table 2), the positive relationship between calorie intake and ALT levels remained consistent and statistically significant (β = 0.0015, 95% CI: 0.0006, 0.0025; p = 0.001).
In athletes, higher protein intake was substantially associated with increased ALT levels in adjusted models (Model 2: β=0.0549, 95% CI: 0.0143, 0.0955; p=0.008). In non-athletes, the association between protein intake and ALT was weaker and not statistically notable in adjusted models (Model 2: β=0.0353, 95% CI: -0.0007, 0.0713; p=0.055). In the fully adjusted model for all participants (Model 5; Table 2), higher protein intake remained significantly associated with increased ALT (β=0.0481, 95% CI: 0.0210, 0.0752; p=0.001).
In athletes, higher carbohydrate intake was significantly associated with increased ALT levels in adjusted models (Model 2: β = 0.0107, 95% CI: 0.0033, 0.0181; P = 0.005). In non-athletes, the association between carbohydrate intake and ALT was weaker and not statistically significant in adjusted models (Model 2: β = 0.0037, 95% CI: -0.0026, 0.0101; P = 0.248). In the fully adjusted model for all participants (Model 5; Table 2), higher carbohydrate intake remained significantly associated with increased ALT (β = 0.0072, 95% CI: 0.0024, 0.0121; P = 0.003).
For fat intake in the group of athletes, there was no notable association with ALT levels in the adjusted models. Specifically, in Model 2 for athletes, the regression coefficient for fat intake was 0.0024 (95% CI: -0.0386, 0.0434; p=0.907), indicating a non-substantial relationship. In Model 2 for non-athletes, the regression coefficient for fat intake was 0.0258 (95% CI: -0.0181, 0.0699; p=0.249), indicating a non-ignificant relationship. When looking at the total study population in the fully adjusted Model 5, fat intake was not significantly associated with ALT levels (β=0.0203, 95% CI: -0.0098, 0.0505; p=0.185).
Energy and Macronutrient Intake with AST in Elderly Athletes and Non-Athletes
The individual models for the athletes, non-athletes, and the entire sample indicated that a higher intake of calories is associated with increased levels of AST, although the association was weaker compared to ALT. In the fully adjusted models (Model 2; Table 3), it was observed that for athletes, each 1-unit increase in calorie intake was linked to a 0.0009 (95% CI: 0.0001, 0.0017; p = 0.018) increase in AST levels, while among non-athletes this association was not statistically significant (β = 0.0004, 95% CI: -0.0004, 0.0013; p = 0.353). Additionally, in the fully adjusted model for all participants (Model 5; Table 3), the positive relationship between calorie intake and AST levels remained statistically significant (β = 0.0006, 95% CI: 0.0000, 0.0012; p = 0.028).
For athletes, protein intake was not notablely associated with AST levels in adjusted models (Model 2: β=0.0136, 95% CI: -0.0085, 0.0357; p=0.228). Similarly, in non-athletes, the association between protein intake and AST was not statistically significant in adjusted models (Model 2: β=0.0160, 95% CI: -0.0097, 0.0418; p=0.222). In the fully adjusted model for all participants (Model 5; Table 3), protein intake was also not significantly associated with increased AST (β=0.0148, 95% CI: -0.0022, 0.0319; p=0.089).
In athletes, higher carbohydrate intake was substantially associated with increased AST levels in adjusted models (Model 2: β = 0.0061, 95% CI: 0.0021, 0.0100; p = 0.003). In non-athletes, the association between carbohydrate intake and AST was not statistically significant in adjusted models (Model 2: β = 0.0018, 95% CI: -0.0026, 0.0063; p = 0.429). In the fully adjusted model for all participants (Model 5; Table 3), higher carbohydrate intake remained significantsubstantially associated with increased AST (β = 0.0038, 95% CI: 0.0007, 0.0068; p = 0.014).
For fat intake in the group of athletes, there was no significant association with AST levels in the adjusted models (Model 2: β=-0.0056, 95% CI: -0.0282, 0.0169; p=0.623). For fat intake in the group of non-athletes, there was also no significant association with AST levels in the adjusted models (Model 2: β=0.0023, 95% CI: -0.0289, 0.0335; p=0.884). When looking at the total study population in the fully adjusted Model 5, fat intake was not substantially associated with AST levels (β=-0.0000, 95% CI: -0.0191, 0.0189; p=0.992).