You are here
Home > Blog > Aging in Place > Cognitive Compromise In Prostate Cancer Patients Receiving Hormone Therapy

Cognitive Compromise In Prostate Cancer Patients Receiving Hormone Therapy

Cognitive Compromise In Prostate Cancer Patients Receiving Hormone Therapy


A comprehensive review and meta-analysis delved into the potential cognitive effects of hormone therapy, a common treatment for prostate cancer. While prior research yielded conflicting outcomes, this study aimed to provide clarity by synthesizing longitudinal data on both objective (cognitive tests) and subjective (questionnaire-based) cognitive changes induced by hormone therapy.


The investigation involved a thorough search across PubMed, Web of Science, and PsycINFO databases for studies examining cognitive changes in patients undergoing androgen-deprivation therapy and new-generation hormone therapy over time. A total of twenty studies, encompassing 1440 patients, were reviewed, with fifteen studies (1093 patients) meeting the criteria for inclusion in the meta-analysis.


Prior to treatment initiation, between 20% to 50% of patients exhibited objective cognitive impairment. The meta-analysis revealed a decline in subjective cognition (g = -0.44; p = 0.03) associated with both androgen-deprivation therapy and new-generation hormone therapy. However, all other effect sizes were small (ranging from g = -0.02 to g = 0.18), with none indicating a significant decline in objective cognition. Notably, significant heterogeneity was observed across all domains of objective cognition, suggesting varied responses among patients.


This study underscores the importance of considering both objective and subjective measures when assessing the cognitive impact of hormone therapy in prostate cancer patients. While subjective perceptions of cognitive decline were evident, objective cognitive function generally remained stable. Further research is warranted to elucidate individual differences in treatment response and the underlying mechanisms driving cognitive changes in this population.



Hormone therapy (HT) stands as a cornerstone treatment modality for prostate cancer (PC), particularly in hormone-sensitive disease. Androgen-deprivation therapy (ADT), a fundamental component of HT, entails reducing testosterone production via luteinizing hormone-releasing hormone agonists or antagonists, along with impeding testosterone binding to androgen receptors in tumor cells using antiandrogens. New-generation hormone therapy (NGHT), targeting the androgen receptor axis, complements ADT for castration-resistant or metastatic PC. Tailoring treatment strategies based on factors like patient life expectancy, comorbidities, and sensitivity to side effects is paramount, guided by clinical presentation and individual preferences.

Related : Prostate Cancer Detection Using Urine Glycoproteins

Previous studies have illuminated the potential central nervous system (CNS) side effects induced by ADT and NGHT, critical considerations given the prevalence of PC. Cognitive impairment, a notable side effect, can impact patient autonomy, complicate care, and correlate with increased cancer mortality. The relationship between androgens and cognition remains debated, with low testosterone levels in the elderly potentially linked to poorer cognitive functioning directly or indirectly via estradiol conversion.


Preclinical research in murine models has identified androgen receptors in brain regions implicated in memory and executive functions. However, clinical studies investigating estradiol supplementation to ameliorate cognition in ADT patients have yielded inconclusive results. While earlier meta-analyses identified lower visuomotor performance in ADT patients compared to non-cancer controls, subsequent reviews reported heterogeneous or contradictory findings, with varying prevalence rates of cognitive decline during treatment.


Notably, the synthesis aimed to assess both objective (cognitive tests) and subjective (questionnaire-based) measures of cognition longitudinally in PC patients receiving HT, including ADT and initial reports on NGHT. Objective impairments are discerned from abnormal cognitive test scores, while perceived cognitive changes are captured through patient-reported questionnaires. Given the limitations of cognitive tests in reflecting real-world cognitive functioning, questionnaires assessing perceived cognitive impairment have gained prominence in research.


This synthesis endeavors to shed light on the nuanced longitudinal impact of HT on both objective and subjective cognition in PC patients, addressing a critical gap in existing literature.



Inclusion Criteria

– Population: Patients diagnosed with prostate cancer.

– Intervention: Receipt of hormone therapy (HT), including androgen-deprivation therapy (ADT) or new-generation hormone therapy (NGHT).

– Study Design: Longitudinal studies with pretreatment assessment.

– Outcome: Assessment of objective or subjective cognition.

– Reports involving case reports, case-control studies, cross-sectional studies, reviews, meta-analyses, study protocols, letters, meeting reports, interviews, and studies exclusively utilizing nonspecific questionnaires or cognitive screening tests were excluded.

– Eligible reports included those providing sufficient data for effect size computation and available for meta-analysis. 


Exclusion Criteria

– Reports not meeting the inclusion criteria.

– Studies solely utilizing nonspecific questionnaires or cognitive screening tests, such as the Montreal Cognitive Assessment.

– Case reports, case-control studies, cross-sectional studies, reviews, meta-analyses, study protocols, letters, meeting reports, and interviews were excluded.

– Reports not providing adequate data for effect size computation.

– Studies not assessing cognitive changes longitudinally.

– Reports not involving patients with prostate cancer undergoing hormone therapy.


Additionally, different reports from the same study were eligible if they provided new data. The eligibility of reports was assessed through a systematic search of PubMed, Web of Science, and PsycINFO databases, complemented by a search on for ongoing studies. Two independent reviewers screened abstracts, followed by full-text eligibility assessment, with conflicts resolved by consensus. Data extraction included study details, patient information, follow-up duration, cognitive domains assessed, statistical analysis, and main results. Risk of bias assessment utilized Pullens and colleagues’ adapted checklist, with studies categorized into high, moderate, or low risk of bias based on total scores.


The statistical analysis employed R 4.2.0, utilizing packages including dplyr, esc, and meta. A cognitive domain approach guided the analysis, encompassing several key steps. Firstly, means, standard deviations (SDs), and sample sizes were aggregated according to the cognitive domain assessed, with data recoded to reflect an increase in cognitive status. Individual effect sizes and associated standard errors were then computed for each cognitive measure, comparing pretreatment with follow-up assessments using the Hedge’s correction for Cohen’s effect sizes, a recommended measure for cognitive changes. Hedge’s values were interpreted as small, medium, and large effects.


In cases where multiple results were provided for an objective cognitive domain within a single study, effect sizes and associated standard errors were combined to obtain a unique measure. For subjective cognition, only the perceived cognitive impairment scale of the Functional Assessment of Cancer Therapy-Cognitive Function questionnaire was considered, given its consensus as the preferred tool for assessing subjective cognition in cancer patients.


Weighted mean effect sizes (Hedge’s g) and their 95% confidence intervals were computed using a random-effect model, alongside values of τ (with a restricted maximum likelihood estimator), Cochran’s Q, and Higgins and Thompson’s I² to quantify heterogeneity. Outliers were identified through a search for effect sizes with non-overlapping 95% confidence intervals with the pooled effect size, with subsequent steps repeated if necessary. Subgroup analyses were conducted using a mixed-effect model to assess differences in cognitive change magnitude according to treatment generation, treatment duration at follow-up, and level of risk of bias.


Meta-regression was employed to examine variations in effect sizes as a function of patient age. Funnel plots were obtained for each domain to estimate publication bias, with Duval and Tweedie’s trim-and-fill method applied if necessary to adjust for funnel plot asymmetry. These steps were repeated for studies including at least one eligible control group assessed at baseline and during follow-up, with weighted mean effects obtained separately for each group (HT vs. mixed PC- and HC) and compared in subgroup analysis.



The systematic review and meta-analysis encompassed 30 selected reports from 20 studies, published between 2002 and 2021, involving a total of 1440 unique patients receiving hormone therapy (HT) and 608 controls (252 in the PC- group and 356 healthy controls). Among the HT groups, the mean age was 72 years, with 39% to 45% having metastatic cancer. The most assessed period was 6 months after treatment initiation, with a mean follow-up duration of 11 months.


A variety of cognitive measures and analytical methods were employed across the studies. Despite methodological variations, the overall risk of bias was moderate, with 12 studies considered to have a low risk and 8 with a moderate risk. Notably, information about patient and disease characteristics of non-responders, consideration of confounding factors, participation and response rates, and sample size were among the least frequently met criteria.


Among studies without a control group, those investigating advanced prostate cancer found no significant cognitive change in patients. However, studies assessing NGHT revealed differences in subjective cognition between treatments, with more perceived cognitive impairment reported in the enzalutamide group compared to abiraterone acetate. Three studies demonstrated cognitive improvement.


In studies with control groups, lower cognitive performance was observed in the ADT groups compared to PC- or healthy control groups across various cognitive domains. Notably, subjective cognition was associated with ADT administration, albeit with variations in findings. Meta-analysis of 15 studies indicated a significant decrease in subjective cognition scores, while no significant decline was observed in objective cognition. Subgroup analyses showed differences in cognitive change magnitude related to follow-up duration and risk of bias.


Visual inspection of funnel plots suggested asymmetry in several cognitive domains, with adjustment made using Duval and Tweedie’s trim-and-fill method to address publication bias. Ultimately, while some cognitive improvements were noted, the overall significance of cognitive changes during treatment was not sustained across all cognitive domains.



This systematic review and meta-analysis delved into the cognitive ramifications of hormone therapy (HT) in patients with prostate cancer (PC), aiming to incorporate findings on new-generation HT (NGHT) and subjective cognition. Notably, an increase in perceived cognitive impairment scores during treatment was observed, marking the first meta-analysis to scrutinize ADT and NGHT’s effects on subjective cognition.


While no consensus emerged on a decline in objective cognition, baseline cognitive impairment was prevalent among patients slated for ADT alone or in combination with NGHT. This impairment primarily manifested in memory tasks, paralleling observations in elderly patients with breast cancer. Factors like cancer-related issues and pre-existing conditions such as anxiety and depression might contribute to pre-treatment cognitive deficits.


During follow-up, several studies indicated a decline in objective cognition or significant between-group disparities favoring non-HT groups, particularly under ADT. Conversely, meta-analysis findings suggested minor effects of ADT or ADT combined with NGHT on objective measures, with some improvements observed in visual memory and executive functions possibly due to practice effects.


While subjective cognition assessments unveiled greater perceived impairment during ADT treatment alone or with enzalutamide, comparisons with non-HT groups were lacking for NGHT. Notably, patients on enzalutamide reported more cognitive impairment than those on abiraterone acetate. Such findings underscore the importance of considering treatment-specific cognitive effects and the age-related cognitive changes prevalent in PC patients, which might complicate care and decision-making processes.


Oncology Related Tools


Latest Research

Prostate Cancer

About Author

Similar Articles

Leave a Reply