Gender-specific hemodynamics in prefrontal cortex during a verbal working memory task by near-infrared spectroscopy

Abstract

The presence or absence of gender differences in working memory, localized in the prefrontal cortex (PFC), has been debated in a few fMRI studies. However, the hypothesis of gender differences in PFC function has not been elaborated, and comparisons among hemodynamic parameters designed to test for gender differences are scarce. We utilized near-infrared spectroscopy during verbal N-back tasks on 26 male and 24 female healthy volunteers. Changes in the concentrations of oxy- (Δ[oxy-Hb]), deoxy- (Δ[deoxy-Hb]) and total hemoglobin (Δ[tot-Hb]) were recorded simultaneously. Δ[oxy-Hb] and Δ[tot-Hb] exhibited obvious gender differences, but Δ[deoxy-Hb] did not. Males showed bilateral activation with slight left-side dominance, whereas females showed left activation. The activation in males was more wide-spread and stronger than in females. Furthermore, females required a lower hemodynamic supply than males to obtain comparable performance, and only females exhibited positive correlations between hemodynamic parameters and behavioral performance. The results reinforce the existence of a gender effect in hemodynamic-based functional imaging studies. Our findings suggest that females possess more efficient hemodynamics in the PFC during working memory and emphasize the importance of studying the PFC to further a scientific understanding of gender differences.

Introduction

The issue of gender influences on performance has recently been highlighted in neuroscience [7]. Gender differences in performance have been recognized on various cognitive tasks [49], [27]. Females outperform males on verbal tasks [49], [16], [2], while males outperform females on visual–spatial tasks [27].

It has been found that sex hormones may affect human prefrontal cortex (PFC) function. PFC and its neural circuitry are suggested to be chief mediators of the influence of sex hormones on cognition [26]. The replacement of estrogen or progesterone roused robust and reliable activation of PFC in females with low levels of sex hormones [5]. Use of estrogen can enhance the performance of PFC-dependent functions, such as working memory (WM) [8], [10], [17]. Thus, sex hormones may play an important role in maintaining certain PFC functions and, thus, PFC functions may well show sexual differentiation.

One function distinctly linked with PFC is WM. A main feature of WM is to temporarily store and maintain information while simultaneously manipulating it, for example, by reordering or actively updating the stored contents of WM [13], [3]. Given that PFC may well demonstrate gender-specific functions and that WM is the most common function of PFC, it is not unreasonable to expect that PFC may be more likely to show gender differences in WM and should be a prominent brain area for intensive study of gender influences.

Gender differences in WM have been researched using N-back tasks with words, faces, letters and numbers [18], [43], [37], the auditory Q3A-INT task [14] and number recognition [4]. These studies were all based on functional magnetic resonance imaging (fMRI) observations of the blood-oxygen-level-dependent (BOLD) signal. These published gender differences in performance and activation patterns are controversial. In behavior, despite the general consensus that females and males are comparable [14], [4], [18], [37], higher accuracy and slightly slower responses in females have also been observed [43]. In brain activation, some research has supported a greater left-lateralization for females and a greater bias toward right hemisphere for males [4], [43], whereas others have reported bilateral activation in females and left-sided PFC activation in males [15] or similar lateralization patterns between the sexes [18], [37].

Gender differences in brain function have been observed in signals of blood flow [8], oxygen saturation of hemoglobin [24], and concentration change of oxy-hemoglobin [25] on the basis of neurovascular coupling. The fMRI BOLD signal is altered by many hemodynamic factors, such as concentration changes in oxy-, deoxy- and total hemoglobin (Δ[oxy-Hb], Δ[deoxy-Hb], and Δ[tot-Hb], respectively), but it cannot identify which hemodynamic factor is altered or the quantitative contributions of each of those hemodynamic factors [9], [30], [12]. Accordingly, the BOLD signal cannot address gender effects on concrete hemodynamic factors. In addition, in some cases, the BOLD signal might mistakenly be interpreted as showing a gender difference in brain activation. For example, females and males alter [deoxy-Hb] equally but alter other hemodynamic factors differently (e.g., [oxy-Hb] in Ref. [25]), but since [deoxy-Hb] is the major factor determining the BOLD signal [46], [22], [40], the BOLD signal would merely obscure the gender difference [18], [37]. Therefore, measuring multiple hemodynamic parameters is recommended for fully and precisely exploring gender effects on hemodynamics-based brain imaging studies.

Both functional near-infrared spectroscopy (fNIRS) [19], [23], [33], [51] and fMRI measure hemodynamic responses to interpret brain activity. Compared to fMRI, fNIRS has the advantage of directly providing multiple measures, including Δ[oxy-Hb], Δ[deoxy-Hb] and Δ[tot-Hb] [46], [44]. In addition, fNIRS can be measured real-life situations, improving the precision of interpreting brain activations [45].

Here, fNIRS was employed to measure hemodynamic-interpreted PFC activation under N-back verbal WM (VWM) tasks. Our hypothesis was that females and males might differ in PFC activation in terms of Δ[oxy-Hb] or Δ[tot-Hb], but not clearly differ in Δ[deoxy-Hb], at the same level of behavioral performance. To obtain a deeper insight into gender influences, we further attempted to investigate gender differences in the brain–behavior correlation between the hemodynamic parameters and behavioral data.