This is the first study that evaluated sex-differences in the influence of acoustic noise on brain activation associated with VA. The major findings are that for the VA task: 1) men have higher activation in the superior PFC, occipital cortices, and the ATHA and higher PFC↔ATHA connectivity than women, while women have higher PFC↔Ins13 connectivity than men; 2) for men, increased AN reduced brain activation in the SPC and the ATHA; 3) brain activation in women was less affected by increased cognitive load or increased AN compared to men; and 4) AN-modulation in the ATHA was positively related to RTs, while VA-load modulation in the DMTHA was inversely related to performance accuracy.
Men and women have different cognitive abilities that may reflect effects of sex hormones or differential brain development on the functional organization of the brain. Neuropsychological studies have shown that men tend to perform better than women on tasks to evaluate visuospatial skills (mental rotation) (Rilea S et al., 2004
), motor function (Ruff R and S Parker, 1993
), perceptual ability (Tirre W and K Raouf, 1994
), and mathematical reasoning (Benbow C and J Stanley, 1983
; Gallagher A et al., 2000
). Women, on the other hand, tend to perform better than men on tests of verbal (Maitland S et al., 2004
) and spatial (Duff S and E Hampson, 2001
) working memory, precision manual tasks and fine motor coordination (Ruff R and S Parker, 1993
), and mathematical calculation (Carr M et al., 1999
) tests. In this study, women responded faster, without drop in accuracy, during “Loud” than during “Quiet” scans for the 3- and 4-ball tracking conditions (this effect was not statistically significant in men). This suggests that periodic noise from the MRI scanner may be advantageous for visual tracking of moving objects. The larger AN effect on reaction times in women may reflect their larger startle responses (Kofler M et al., 2001
) compared to men. The women’s perception of the “Loud” condition as being even louder could reflect lesser sensory gating (Hetrick W et al., 1996
) and hence lower BOLD responses in the thalamus for women than for men. Women tended to perform better than men during the more difficult condition (4-ball tracking), and especially during the “Loud” scan; this difference was not statistically significant, possibly due to the limited sample size and the relatively large variability of the behavioral responses.
The negative latent connectivity between the MFG6 and the Ins13 during “Quiet” scans was larger for women than for men, which probably contributes to the sex × VA-load interaction on the right latent connectivity of the MFG6 and the Ins13. Conversely, the positive latent connectivity between the MFG6 and the ATHA during “Quiet” was larger for men than for women. The significance of these connections is unclear; however, they might be related to the sensory gating function of the thalamus. Electrophysiological studies have shown that the thalamus controls the flow of sensory-motor information to and from the cortex (McCormick D and T Bal, 1994
). This sensory gating function of thalamic neurons has been demonstrated experimentally in rodents (Krause M et al., 2003
) and other species (Ciancia F et al., 1988
; Schall U et al., 1999
). Sensory gating is modulated by a dopaminergic/glutamatergic mechanism (Schall U et al., 1999
) involving the brainstem, hypothalamus, and cerebral cortex. In sleeping or anesthetized animals, this mechanism generates the spontaneous spindle waves (0–15 Hz) of slow-wave sleep (Steriade M et al., 1993
) that disconnect the cerebral cortex from sensory input, probably to minimize sensory interference. Thus, it appears that the lesser auditory gating of the thalamus in the females might reduce the latent connectivity between the thalamus (ATHA) and the PFC (MFG6), but increase inhibitory connectivity between the PFC (MFG6) and auditory cortices (Ins13). This finding could also reflect a reduction in the flow of auditory information to the PFC.
The induced Ins13↔MFG6 connectivity was modulated by the VA-load and also exhibited a VA-load × sex interaction. This suggests that deactivation (defined as a negative BOLD response) of the Ins13 reflects inhibition (Tomasi D et al., 2006
) rather than a purely hemodynamic “blood stealing” mechanism (Raichle ME and DA Gusnard, 2002
During the VA task, women also responded faster during louder conditions (), and showed AN × Sex interactions in the ATHA (), indicating that brain activation was less affected by acoustic noise in women than in men. The positive association between AN-modulation in the ATHA and increased RTs () further support our conclusion that men had to reduce activation in this hyperactive region ( and ) in order to keep their speed during the louder condition. Thus, it is possible that the lack of activation in the ATHA (), or lesser sensory gating, may be advantageous for women to cope with the increased AN.
We recently evaluated the AN effect on working memory processing using the same precise and reproducible 12dBA spl
-difference between “Loud” and “Quiet” scans used in this study, and found that for healthy volunteers, louder scanner noise increased activation in the occipital cortices, PFC, and the cerebellum (Tomasi D et al., 2005
). We further demonstrated that acoustic noise during working memory tasks has differential effects on brain activation in HIV patients compared to control subjects (Tomasi D et al., 2006
) and a right lateralization of AN-activation for female, but not for male subjects (Tomasi D et al., 2005
). In the present study on the VA task, increased AN led to reduced brain activation in the parietal cortices and the ATHA in men but not in women. We cannot determine whether the activation differences reflect differences during “task” or “rest” because the BOLD-fMRI signal only reflects relative changes between two conditions (e.g. from “rest” to “task”). However, this finding might reflect higher hemodynamic baseline (“rest”) during louder conditions in these regions. During the demanding sustained attention “TRACK” periods of this VA task, the hemodynamic response in these regions may have reached its maximum value for “Quiet” scans; therefore, additional AN-related hemodynamic increases were not possible. In contrast, during “rest” periods, the lower hemodynamic demands could have enabled a higher baseline for “Loud” than for “Quiet” scans. Therefore, the AN-related activation decreases might reflect lower hemodynamic bandwidth in parietal cortices and ATHA in men compared to women.
A model of limited dynamic range of hemodynamic responses in this study is supported by the negative correlations of VA-load and AN responses in the cortical and subcortical regions, which may reflect the limited capacity of the VA network (Tomasi D et al., 2005
; Tomasi D et al., 2006
). The larger VA-load effect on activation in the right DLPFC (IFG47) and the lower VA-load effect on deactivation in the insula for men than for women highlight sex-specific differences. Specifically, under increased VA-load conditions, men may shift attention to the balls to be tracked, causing larger parietal activation. In contrast, women may reduce attention to the AN, resulting in larger deactivation of auditory cortices (posterior insula), but allowing more focused attention to perform the task and having faster reaction times. This differential brain response could underlie the tendency to better performance accuracy during the more demanding 4-ball tracking task for women compared to men ().
Some of the sex differences observed in other fMRI studies might be attributable to the higher hematocrit in men than in women, which could alter the BOLD responses (Levin J et al., 2001
). However, in our analysis, the higher hematocrit in men than in women cannot explain the larger activation in the former group because we co-varied for the hematocrit level. Therefore, the larger activation for men than for women in this study reflects sex-specific differences in the functional organization of the brain beyond that observed with hematocrit-related increased in BOLD signals in men than women.
Men also had larger thalamic volume than women, even after controlling for differences in intracranial volume. This finding is consistent with previous findings that showed larger thalamic volume for men than for women (Chang L et al., 2005
). Larger thalamic volume may be related to a larger number of neurons supporting greater neuronal activity in the thalamus and may partially explain the larger hemodynamic responses in the thalamus () for men than for women during the VA task. The lack of correlation between BOLD responses and the normalized thalamic volume suggests that higher thalamic activation in men does not simply reflect a larger thalamic volume. However, this preliminary finding could also reflect the small sample size and the large variability of the fMRI signals. Finally, the origin of behavioral and functional sex differences here reported could be partially attributed to cultural differences between men and women.