2 Examining Gender Differences in the Serial Position Effect and its Relationship to Memory Outcomes

Abstract

Objective:

There are gender-related disparities in age of diagnosis of Mild Cognitive Impairment (MCI) or dementia, with women often receiving delayed diagnoses compared to men. These delays may be related to the general female advantage in episodic verbal memory across aging. Thus, it is important to identify methods of examining memory performance that can help to reduce disparities in diagnosis. The serial position effect, a pattern where individuals tend to remember more words at the beginning (primacy) and end (recency) of a list, is predictive of dementia and may provide an avenue for this endeavor. Whereas healthy adults tend to exhibit a prototypical U-shaped serial position profile, those with MCI or dementia tend to show reduced primacy relative to recency (i.e., a J-shaped profile). To date, few studies have examined gender differences in the serial position effect. There is some evidence to suggest that older, cognitively healthy women perform better than men on middle and recency, but more research is needed to clarify the relationship between gender and the serial position effect, which was the focus of this study.

Participants and Methods:

We utilized data across three archival datasets, which included a total of 338 participants (67.5% female; Mage=66.9, SDage=9.4) divided into three age groups (50-64, 65-75, 76+). Scores on the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) immediate and delayed memory indices (IMI and DMI, respectively) were used to assess verbal episodic memory abilities. Performances across learning trials of the List Learning task were utilized to examine the serial position effect. ANCOVA analyses were conducted and utilized regional scoring of list learning performance, which examined the percentage of correctly recalled words in each portion of the list (primacy, middle, and recency). We also calculated the relative strength of primacy to list learning to examine the relationship between J-curve performance and gender. Years of education was included as a covariate in all analyses.

Results:

Consistent with prior literature, men performed worse performance on both IMI F(1, 331)=17.20, p<.001, and DMI, F(1, 331)=6.87, p=.009, across aging. Repeated measured GLM showed that the serial position effect was seen across the full sample, F(1.93, 639.874)=5.66, p=.004, and interacted with gender, F(1.933, 639.874)=5.70, p=.004, and education, F(1.933, 639.874)=6.13, p = .003. Although men and women did not differ in primacy, p = .67, women performed better in middle, p < .001, and recency, p = .03, performance. Higher education was associated with better primacy, but not middle or recency, performance. Additionally, 23.1% of the sample exhibited a J-curve pattern, and there was a main effect of J-curve pattern for both IMI, F(1, 334)=12.33, p<.001, and DMI, F(1, 334)=15.62, p<.001, with those showing a J=curve having worse memory performance.

Conclusions:

Our finding of no gender difference in primacy suggests that focusing on primacy performance in verbal list learning may help to address gender-related disparities in MCI or dementia diagnosis. Additionally, given evidence of education being associated with primacy, but not middle or recency, performance, future research should investigate the development of education-based normative data for primacy performance.