Materials and methods

Participants

We recruited female BCS, aged 30 to 65, who were >6 months post-treatment, and age-matched NC. Recruitment was conducted through community outreach, social media advertisements, the University of Ottawa’s Integrated System of Participation in Research community pool, and the Cancer Foundation Maplesoft-Jones Centre at the Ottawa Regional Cancer Foundation. Exclusion criteria included: a mental health diagnosis and/or initiating mood-altering medications within <12 months, a lifetime diagnosis of a neurodevelopmental or neurological disorder, current substance dependence or cannabis use >7 times/month, non-fluency in English, residence outside of Canada, and a Montreal Cognitive Assessment (MoCA) score ≤21 [46]. Of the ninety-three eligible participants, two withdrew (BCS, n=2) and three were excluded due to audio recording failures (BCS, n=2; NC, n=1), resulting in a sample of 88 participants (BCS, n=45; NC, n=43; Tables 1 and 2). The study was approved by the University of Ottawa’s Research Ethics Board (H-04-21-6706) and adhered to the Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans.

 | BCS (n=45) | NC (n=43)
 | Mean (SD) or % (n) | Mean (SD) or % (n)
Age | 51.22 (8.73) | 49.88 (10.67)
Race |  | 

White | 95.6 % (43) | 81.4 % (35)
Black | 0.0 % (0) | 2.3 % (1)
East Asian | 0.0 % (0) | 2.3 % (1)
South Asian/East Indian | 2.2 % (1) | 2.3 % (1)
Indigenous | 0.0 % (0) | 2.3 % (1)
Latin American | 0.0 % (0) | 2.3 % (1)
Person of mixed origin | 0.0 % (0) | 2.3 % (1)
Turkish | 2.2 % (1) | 0.0 % (0)
West Asian, North African, Arab | 0.0 % (0) | 2.3 % (1)
Prefer not to answer | 0.0 % (0) | 2.3 % (1)

Province of residence |  | 

Alberta | 0.0 % (0) | 4.7 % (2)
British Columbia | 4.4 % (2) | 9.3 % (4)
Manitoba | 8.9 % (4) | 0.0 % (0)
New Brunswick | 2.2 % (1) | 0.0 % (0)
Nova Scotia | 4.4 % (2) | 0.0 % (0)
Northwest Territories | 4.4 % (2) | 0.0 % (0)
Ontario | 55.6 % (25) | 69.8 % (30)
Prince Edward Island | 2.2 % (1) | 0.0 % (0)
Quebec | 13.3 % (6) | 11.6 % (5)
Saskatchewan | 4.4 % (2) | 2.3 % (1)
Yukon | 0.0 % (0) | 2.3 % (1)

Completed education level |  | 

High school diploma | 6.7 % (3) | 2.3 % (1)
College degree | 24.4 % (11) | 18.6 % (8)
Bachelor’s degree | 44.4 % (20) | 39.5 % (17)
Master’s degree | 17.8 % (8) | 27.9 % (12)
Doctorate degree | 2.2 % (1) | 7.0 % (3)
Other | 4.4 % (2) | 4.6 % (2)

Income |  | 

<$30,000 | 0.0 % (0) | 7.0 % (3)
$30,000–$50 000 | 2.2 % (1) | 9.3 % (4)
$50,000–$100,000 | 24.4 % (11) | 20.9 % (9)
$100,000–$200,000 | 46.7 % (21) | 39.5 % (17)
>$200,000 | 20.0 % (9) | 11.6 % (5)
I prefer not to answer | 6.7 % (3) | 11.6 % (5)
Current substance use | 46.7 % (21) | 60.5 % (26)

Drug-frequency |  | 

Daily | 20.0 % (9) | 27.9 % (12)
2–3 times per week | 8.9 % (4) | 11.6 % (5)
Once per week | 2.2 % (1) | 9.3 % (4)
Once per month | 22.2 % (10) | 23.3 % (10)
Not applicable | 46.7 % (21) | 27.9 % (12)
Mental health diagnosis (lifetime) | 13.3 % (6) | 16.3 % (7)
Mental health symptoms (past 12 months) | 0.0 % (0) | 4.7 % (2)
Mental health symptoms (past 6 months) | 0.0 % (0) | 0.0 % (0)
Medications | 82.2 % (37) | 51.2 % (22)
Mood altering medications | 13.3 % (6) | 9.3 % (4)
Menopausal status | 82.2 % (37) | 53.5 % (23)

 | %, (n)
Time since diagnosis | 

1–2 years ago | 17.8 % (8)
2–5 years ago | 48.9 % (22)
5–10 years ago | 20.0 % (9)
>10 years ago | 13.3 % (6)

Cancer stage at diagnosis | 

Stage 0 | 2.2 % (1)
Stage 1 | 15.6 % (7)
Stage 2 | 44.4 % (20)
Stage 3 | 33.3 % (15)
Unknown | 4.4 % (2)

Treatment end | 

6–12 months ago | 15.6 % (7)
1–2 years ago | 28.9 % (13)
2–5 years ago | 26.7 % (12)
5–10 years ago | 20.0 % (9)
>10 years ago | 8.9 % (4)

Treatment | 

Chemotherapy | 100.0 % (45)
Hormonal therapy | 60.0 % (27)
Radiation therapy | 80.0 % (36)
Surgery | 100.0 % (45)
Immunotherapy | 4.4 % (2)
Targeted therapy | 17.8 % (8)

Current hormone therapy | 

Yes | 53.3 % (24)
No | 6.7 % (3)
Not applicable | 40.0 % (18)

Procedure

This study involved two online sessions, each lasting up to 60 min (Figure 1). In session one, participants completed the videoconference version of the MoCA. Eligible participants then filled out a sociodemographic and health questionnaire via Qualtrics (RRID: SCR_016728), an online questionnaire tool. In session two, participants began with the immediate recall of two Taler stories [37], followed by 20 min of various cognitive measures (omitted from this paper), including the Survey of Autobiographical Memory (SAM; [47]), and then performed the delayed recall of the Taler Stories and the VPA [48] on Gorilla (RRID: SCR_020991), an online research platform. The National Aeronautics and Space Administration Task Load Index (NASA-TLX) was completed after each task, except the SAM [49]. All tasks were adapted for online administration, with auditory stimuli pre-recorded and participants’ spoken responses recorded via Gorilla. Throughout, participants remained connected on Zoom, with a research team member available for technical support.

Figure 1:: Schematic of the study design. cISPR, Community Integrated System of Participation in Research at the University of Ottawa; DR, Delayed Recall; MoCA, Montreal Cognitive Assessment; IR, immediate recall; n, sample size; NASA-TLX, National Aeronautics and Space Administration Task Load Index; ORCF, Ottawa Regional Cancer Foundation; SAM, Survey of Autobiographical Memory; VPA, Verbal Paired Associates.

Figure description: Flow chart depicting the study design. A recruitment session was followed by session 1 which included cognitive and demographic screening, and session 2 which included cognitive testing.

Measures

Montreal cognitive assessment (MoCA)

The MoCA is a 13-item cognitive screening tool designed to detect mild cognitive impairment by evaluating memory, attention, language, executive functioning, and visuospatial abilities [46]. Scores from each item are summed to yield a total score on 30, with lower scores indicating greater cognitive impairment. A cut-off score of ≤21 was applied for this study, as the standard cut-off of 26 may lead to over-pathologizing cognitive difficulties in this clinical population [50].

Taler stories

The Taler Stories task is designed to assess episodic verbal memory [37] and was developed as an alternative to WMS – LM [51] to refine scoring criteria, enhance interpretation of group differences, and improve inter-rater reliability [37]. For this study, one pair from the 12 Taler Stories pairs was randomly selected. Participants listened to two stories, each comprised of four sentences totaling 65 to 69 words, then recalled them immediately and again after a 20-min delay. Responses were scored as verbatim (exact word-for-word recall), gist (accurate, rephrased content), or distortion (inaccurate details). For example, “One sunny Saturday, Fred and Grace went to visit their two grandchildren in New York”, would be scored as: “went to visit” is verbatim, “went to see” is gist, and “went to pick up” is a distortion. Each story contains 50 informational units, which are scored separately to generate verbatim, gist, and distortion scores – each with a maximum of 50 points. A high verbatim score reflects greater recall of episodically rich and specific details, a high gist score indicates accurate recall of the story’s general content and themes, whereas a high distortion score reflects inaccurate recall of story content. A retrieval success score was calculated by summing verbatim and gist scores and subtracting distortions to capture total accurate recall while accounting for errors.

Verbal paired associates (VPA)

The VPA task is a widely used measure of verbal episodic memory [48]. Participants listened to a list of eight unrelated word pairs (e.g., banana-shoe). Afterward, they heard one word from each pair (e.g., banana) and were asked to recall the corresponding target word (e.g., shoe). The procedure was repeated across four trials, with the stimuli presented in a fixed, trial-specific sequence. The total score out of 32 was derived by summing the correct responses for each word pair across all four trials, with lower scores indicating worse episodic recall.

Survey of autobiographical memory (SAM)

The SAM is a 26-item self-report questionnaire designed to assess perceived abilities related to episodic, semantic, future, and spatial memory [47]. Participants rated each item using a 4-point Likert scale (strongly disagree to strongly agree). Each item response was converted to a numerical score based on an established scoring procedure [47]. Domain-specific scores were calculated by summing relevant items, with higher scores indicating better perceived memory functioning. For this study, only episodic and semantic subscales were analyzed.

National aeronautics and space administration task load (NASA-TLX)

The NASA-TLX is a self-report measure designed to assess cognitive workload during tasks [49]. The six-item questionnaire evaluates various domains of workload: mental demand, physical demand (omitted from this study), temporal demand, performance, effort, and frustration. Participants rated each item on a 20-point scale, which was summed to yield a maximum score of 100, with higher scores indicating higher perceived cognitive workload.

Statistical analysis

Responses were manually transcribed, scored, and double-checked by study researchers blind to conditions. Inter-rater reliability for the Taler Stories scoring exceeded 90 %, with scoring discrepancies resolved through discussion. Statistical analyses were conducted using SPSS 28 (RRID:SCR_002865). Four 2 × 2 repeated measures ANOVAs (RM ANOVA) assessed four memory recall types on the Taler Stories (verbatim, gist, distortions, retrieval success) across time (immediate vs. delayed) for both groups (BCS vs. NC). Bonferroni correction for multiple comparisons was applied when necessary to reduce the risk of type I errors. Independent samples t-tests compared group performance on the VPA, SAM, and NASA-TLX. Pearson correlations (r) assessed the relationship between: NASA-TLX and Taler Stories – Immediate Recall, NASA-TLX and Taler Stories – Delayed Recall, and NASA-TLX and VPA. Statistical significance was set at α=0.05, and partial eta squared (ηp2) was calculated to estimate effect sizes. Figures were generated using GraphPad Prism (RRID:SCR_002798). Final sample sizes varied by task due to missing data: Taler Stories (n=83; BCS=42, NC=41), SAM (n=87; BCS=45, NC=42), and VPA (n=88; BCS=45, NC=43).