Late diagnosis of HIV is still common in South China [18, 19], which affects patients receiving highly active ART and increasing mortality [20]. This cohort study provided evidence for a linear relationship between serum albumin and 12-week mortality in patients with late HIV/AIDS diagnosis, which was independent of age, sex, hemoglobin, thrombocyte count, CD4, CD8, TBIL, ALT, Cr, CRP, PCT and BMI. The risk of 12-week mortality decreased by 6% with every 1 g/L increase in the serum albumin concentration. Furthermore, stratified analysis reflected a persistent positive association between serum albumin concentration and 12-week mortality.
In recent decades, many studies have explored the association between serum albumin levels and mortality. Some studies have shown that serum albumin levels upon admission are inversely associated with short-term mortality in various diseases [20, 21]. It is not only used as a predictor of short-term mortality in patients with HIV/tuberculosis (TB) co-infection [22], but also as a predictor of in-hospital mortality in those with HIV [23]. However, it has been suggested that increasing serum albumin levels by infusion does not alter the survival rate [24]. In patients with sepsis, additional albumin infusion did not improve survival compared with that of crystalloid infusion alone [25].
AIDS-defining illnesses are the main cause of death in patients with late HIV diagnosis [26, 27]. Blood markers traditionally used to assess disease severity in patients with AIDS include CD4 and CD8 [28], inflammation markers, such as neutrophil-lymphocyte ratio [29], and CRP [30, 31]. The development of T-cell exhaustion in chronic HIV infection is due to high levels of HIV antigen, strong proinflammatory immune activation, and impaired T-cell homeostasis [28]. However, researchers have increasingly recognized that serum albumin is also an important indicator for evaluating the degree of inflammation and disease prognosis, and low baseline serum albumin levels and impaired immunity reliably predict mortality [32]. The main causes of hypoproteinemia due to chronic inflammatory states include albumin leakage during the inflammatory response [33] resulting from increased capillary permeability [34]. Meanwhile, the total albumin mass decreases with the short half-life of albumin [34]. HIV infection also reduces serum concentrations through different mechanisms, such as HIV-1 Tat-1 protein, which damages the barrier function of the endothelium [35]. On the contrary, some studies used cohort studies to find a moderate-weak negative correlation between serum albumin level and the inflammatory index CRP; however, serum albumin is more affected by serum sodium and hemoglobin [36]. This may be because different experts used different indicators to evaluate inflammation.
Our study had several strengths. Cases were collected continuously from January 2018 to December 2021 to avoid selection bias. To ensure an independent association of serum albumin and outcomes, the study was adjusted for several potential confounding factors, including age, sex, CD4, CD8, CRP, PCT, ALT, TBIL, Cr and BMI. In this study, we conducted a subgroup analysis using age, gender, hemoglobin, TBIL, ALT, PCT, CRP and Cr levels as stratification factors and tested the interaction. The effect size of serum on 12-week mortality in most subgroups was stable, and no interaction was found between stratified factors and 12-week mortality. The selection of mortality at 12 weeks after admission in this study could reduce missing samples and ensure outcomes to truly reflect the early mortality of patients with late HIV/AIDS diagnosis in the real world. We included all newly diagnosed patients and evaluated the mortality rate within 12 weeks after admission to reduce sample loss and calculation bias caused by admission criteria.
This study has some limitations. First, age, sex, CD4, CD8, CRP, PCT, ALT, TBIL, Cr and BMI were adjusted in this study, and we aimed to prevent the influence of other factors; however, some influences were not recorded or observed. Therefore, the effect of residual confounding factors could not be eliminated. Second, we could not conclusively determine the causal relationship between serum albumin concentration and 12-week mortality. Third, all participants were Chinese, and the findings of this study may not apply to other ethnicities.