Stroke, a prevalent cerebrovascular disease, has emerged as one of the foremost causes of disability and death worldwide, posing significant challenges to public health (1, 2). Post-stroke cognitive impairment (PSCI) is a clinical syndrome characterized by cognitive deficits following a stroke, with approximately one-third of stroke patients experiencing varying degrees of PSCI (3). Patients with PSCI have cognitive dysfunction, which leads to compromised motor, language, and self-care abilities, which in turn increases the family’s healthcare costs and severely impacts the patient’s quality of life (4). Studies have shown that the period from post-stroke to the onset of PSCI can be considered a therapeutic window for early intervention to protect cognitive function and reduce mortality through better early care (5, 6). Therefore, identifying validated predictors for early screening and evaluation of patients with PSCI is particularly crucial.

Inflammatory factors have a strong association with cognitive impairment (7, 8). Recent studies indicate that lymphocytes play a pivotal role in inflammatory repair and brain protection (9). Lymphopenia after stroke also suggests a poor prognosis for neurological disorders (10). Clinically, lymphocyte-associated inflammation indices such as the neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), and HALP (hemoglobin, albumin, lymphocyte, and platelet) scores are strongly correlated with cardiovascular disease and malignancy (11, 12). However, it remains unclear whether these inflammatory markers are associated with cognitive impairment following stroke. Therefore, our study aimed to systematically retrieve and analyze studies on the lymphocyte-associated inflammation index to provide an evidence-based basis for its clinical application in the early prediction of PSCI.

This meta-analysis adheres to the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) 2020 statement (13) and was registered prospectively in PROSPERO (CRD42024541099). The PRISMA 2020 checklist is presented in Supplementary Table 1.

The primary objective of this meta-analysis was to assess the predictive value of the lymphocyte-related inflammation index for PSCI. A total of 16 articles involving 3,406 patients were included in our analysis. The results of the meta-analysis indicated that PLR and NLR levels were significantly higher in the PSCI group than in the non-PSCI group, whereas HALP scores and lymphocyte count in the PSCI group were not significantly different from the non-PSCI group.

At present, a great deal of research suggests a strong link between cognitive impairment or dementia and inflammatory factors (35, 36). Klulesh et al. revealed that patients with executive cognitive impairment had significantly higher cerebrospinal fluid concentrations of (Interleukin) IL-1β, IL-10, and serum IL-6 levels than cognitively normal patients (37). Qi et al. discovered a strong link between inflammatory factors (Tumor necrosis factor-α, C-reactive protein, and IL-6) and the development of vascular dementia (38). However, these inflammatory factors require additional testing to obtain, which greatly limits their use as an effective means of early screening and detection of PSCI. In recent years, an increasing number of studies have focused on inflammatory indices associated with lymphocytes, such as PLR and NLR. These two novel inflammatory indices reflect the balance between lymphocyte and platelet and neutrophil levels, respectively, and have the advantage of being more readily available and more widely used in clinical practice. In previous studies, patients with post-stroke cognitive impairment had higher levels of PLR and NLR compared to patients with post-stroke non-cognitive impairment (26, 33), which is consistent with our findings.

When a stroke occurs, numerous platelets and neutrophils accumulate in the damaged area of the brain. Platelets secrete pro-inflammatory factors into the damaged area, further exacerbating the inflammatory response and causing damage to blood vessels and neurons (39, 40), while neutrophils also induce pro-inflammatory factors, including IL-6 and matrix metalloproteinase-9, which can disrupt the brain’s blood-oxygen barrier, impeding the flow of oxygen and nutrients to the brain (41–43). Both the above-mentioned platelet and neutrophil damage to the central nervous system after a stroke cause cognitive decline, contributing to the development of PSCI. Lymphocytes, by contrast, play a role in tissue repair and neuroprotection. It has been shown that during a stroke, regulatory T lymphocytes play a neuroprotective role by producing anti-inflammatory factors and thus inhibiting the inflammatory process. However, lymphocytes undergo a corticosteroid response, resulting in their decreased numbers, which hampers the repair of post-stroke damage and accelerates the development of post-stroke cognitive impairment (44, 45). Considering the findings from previous studies indicating that PLR and NLR are strongly associated with prognosis, mortality, and severe atherosclerosis in post-stroke patients (42, 46, 47). Therefore, we suggest that PLR and NLR levels can be used as a new and effective tool for predicting PSCI, which can help in early clinical detection and prediction of PSCI.

The HALP scores is another lymphocyte-related indicator that reflects not only the patient’s inflammatory state but also their nutritional status. It is derived from the weighted sum of four items: hemoglobin, albumin, lymphocytes, and platelets, which excludes the interference of other factors, with the advantages of being easy to obtain, simple, and effective (24). The albumin and hemoglobin included in the HALP scores are important indicators of the nutritional status of the body. Studies have shown that reduced hemoglobin leads to decreased oxygen transport capacity and brain oxygen delivery, which can lead to neuronal damage, mitochondrial disorders, oxidative stress, and inflammation (48, 49). Secondly, reduced albumin levels impair the body’s ability to fight oxidation and capture oxygen-free radicals. All these factors increase the risk of poor prognosis and cognitive impairment in post-stroke patients (50). Zuo et al. (51)study results showed that low HALP scores increase the risk of PSCI. However, in our study, the HALP scores of patients in the PSCI group were not significantly different from those of patients in the non-PSCI group. We believe that there may be the following reasons for this difference: first, the number of studies included was small; second, the cognitive assessment scales used in the studies we included were not the same, which may have led to differences in the diagnosis of cognitive impairment in the patients, thus affected the final pooled results. In conclusion, we believe that the HALP scores may be able to be used as another lymphocyte-associated inflammation index to reflect the body’s inflammatory and nutritional status for early screening of PSCI, but studies with larger sample sizes are needed to validate this conclusion.

Finally, our study showed no significant difference in lymphocyte count between the PSCI and non-PSCI groups. In the results of sensitivity analysis, when the lymphocyte count data of Li et al. were excluded, there was a change in the combined WMD. The reason for this may be that unlike other studies where patients were assessed for cognitive function at a short period of time, the time point for PSCI assessment in the study by Li et al. was the 6th month after stroke, which may have led to an alteration in the combined WMD. Meanwhile, Wang et al. (25) mentioned in their study that lymphocyte count could not be identified as an independent risk factor for PSCI. Based on these findings, lymphocyte count cannot yet be considered strongly correlated with the occurrence of PSCI. Thus, it is not recommended to rely solely on lymphocyte count as an indicator for predicting the occurrence of PSCI.

Our study also has several limitations. Firstly, most of the studies we included were from China, thus limiting the generalization of the findings to other regions and ethnicities. Secondly, we used only published articles in English and Chinese, which may have led to selection bias in our findings. Also, despite our sensitivity and subgroup analysis, we still did not find a source of high heterogeneity in our study results, taking into account potential confounders, which reduces the reliability of our study results. Finally, PLR, NLR, and HALP scores in the included studies were measured only once, failing to capture dynamic changes associated with PSCI. Therefore, future studies should involve continuous monitoring to establish the association between dynamic changes in these indices and PSCI. Certainly, our study is the first meta-analysis about the predictive value of lymphocyte-associated inflammation index in PSCI, providing evidence-based medical support for early screening and prediction of PSCI in clinical practice.

In conclusion, increased levels of NLR and PLR were significantly associated with PSCI. However, there was no strong evidence of a direct relationship between HALP scores, lymphocyte count, and PSCI. Considering the limitations of this paper such as regional selective bias, potential heterogeneity, and retrospective study design, more prospective studies with large sample sizes and multicenter are needed in the future to further confirm the predictive value of lymphocyte-associated inflammation index for PSCI.