Sex differences in human virtual water maze performance: Novel measures reveal the relative contribution of directional responding and spatial knowledge

Abstract

Sex differences in humans on virtual water maze navigation are well established when overall performance is measured, e.g., by the total time taken to find the hidden platform, total path length, or quadrant dwell time during probe trials. Currently, it is unknown whether males are better spatial learners than females, or if overall performance differences reflect other aspects of the task unrelated to spatial memory. Here, males and females were tested on a virtual analogue of the Morris water maze. We devised a novel method of analysis in which each trial was divided into an initial trajectory phase and search phase. We also implemented a new measure of spatial learning during early and late training, by including trials in which subjects were only required to indicate where they thought the hidden target zone was located. Consistent with previous reports, males outperformed females on overall measures of task performance. Males also performed significantly better on all initial trajectory phase variables. Interestingly, only small (non-significant) differences were observed during the search phase and when spatial learning was tested without the constraints of a typical water maze trial. Our results suggest that spatial knowledge regarding the location of the hidden target zone is not the main factor responsible for overall sex differences in virtual water maze performance. Instead, the largest sex differences were observed during the initial trajectory phase of the trial, which is thought to depend on effective processing of distal features of the environment.

Introduction

The Morris water maze is commonly used in rodent research to study the neurobiology of learning and memory (for a review, see [5]). The task requires subjects to learn the spatial position of an escape platform that is located in a circular pool of opaque water and hidden just beneath the surface. The importance and widespread use of the rodent version of the task has prompted an increasing number of researchers to investigate the performance of humans on analogous virtual versions [1], [2], [11], [18], [19], [22].

Previous studies in both rodents [4], [24] and humans [1], [2], [18], [19] suggest males perform better on the water maze compared to females. At present it is unclear which mechanisms underlie this sex difference. Are males better spatial learners, or is the advantage derived from other behavioural aspects of the task important to overall performance? Here, we examined which specific elements of behaviour comprising an individual trial lead to sex differences in human subjects.

In humans, it is well established that males outperform females on measures assessing overall virtual water maze performance. For example, Astur et al. [1], [2] reported that adult males were faster in locating the hidden platform during training, a finding replicated in children [19]. These studies also found that when the platform was removed on a probe trial following training, males displayed a stronger search preference for the former platform quadrant. One potential mechanism mediating apparent sex differences on overall task performance in humans is the extent to which distal features of the environment can be used effectively. Two sources of distal information thought to be important to navigation in the water maze are (1) distal cues, defined as distinct visual features of the environment located outside the pool wall and (2) the geometry of the room in which the pool is located [22]. The pool wall itself and landmarks located within its boundaries can also be used, and are referred to as proximal cues. Sandstrom et al. [22] observed that males performed better than females when the geometry of the room was the only useful source of distal information available. Saucier et al. [23] demonstrated in a real-world navigation task that females preferred using landmark information, whereas males displayed a preference for cardinal directions (e.g. North, East) and distances. To determine which visual features in the environment were used during virtual water maze performance, Mueller et al. [18] recorded eye tracking data during the first 3 s of each trial. While both sexes directed gaze towards distal features of the environment significantly longer than the pool wall or search area, shorter trial latencies in males corresponded with a slightly longer focus on distal cues, and faster visual exploration of the environment early during training. These findings suggest that overall sex differences might arise from the ability of males to faster learn and discriminate important features of the distal environment.

Understanding precisely how distal and proximal cue processing influences behavioural water maze performance is receiving increased attention. While self-localisation within the room at the beginning of a water maze trial must be based on distal room cues, the initial trajectory towards a hidden platform can either be determined only by its spatial relationship to distal cues (place learning), or by combining distal and proximal information to select an initial trajectory towards the hidden platform based on its location relative to the pool wall (directional responding) [8], [9], [10], [11]. Hamilton and colleagues argue that each trial begins with a directional response, followed by a local search for the platform controlled by judging distance to the pool wall. This idea is supported by a series of experiments that examined trajectories on probe trials in which the location of the pool was shifted within the room. In this situation, the absolute location of the learned platform position within the room is placed in competition with the relative location of the platform position within the pool. A trajectory towards the absolute location is consistent with place learning within the room, while a trajectory towards the relative location is consistent with a directional response based on proximal cues provided by the pool wall. They found that rats [9], [10] and humans [11] develop a preference to respond in the relative direction of the platform. Eye tracking data obtained from human subjects during performance of a virtual water maze showed that those subjects able to make direct trajectories to a hidden platform location directed their gaze towards distal cues early in the trial, before gradually shifting focus to the pool wall for the remainder of the trial [11]. In contrast, subjects classified as learning a non-direct strategy directed their gaze towards distal cues for a shorter duration early in the trial, before rapidly shifting focus to the pool wall. Although the authors indicated that a similar proportion of males and females were assigned to direct and non-direct trajectory groups, an analysis directly assessing sex differences was not reported.

The present study assessed if overall differences between human males and females on virtual Morris water maze performance could be attributed to specific components of the trial. New behavioural measures and analysis methods were developed in order to assess initial trajectory and local search performance independently. We hypothesised that males would perform better during initial direction selection, since this behaviour appears dependent on distal cues. It was an open question whether sex differences would be observed to a similar extent when accurate searching becomes more dependent on the pool wall.