Gender differences in empathy: The role of the right hemisphere

Abstract

The relationship between activation of the right cerebral hemisphere (RH) and empathy was investigated. Twenty-two men and 73 women participated by completing a chimeric face task and empathy questionnaire. For the face task, participants were asked to pick which of the two chimeric faces looked happier. Both men and women were significantly more likely to say the chimera with the smile to their left was happier, suggesting activation of the RH. As expected, men scored significantly lower than women on the empathy questionnaire, p = .003. A correlation was found between RH activation on the face task and empathy for women only, p = .037, suggesting a possible neural basis for gender differences in empathy.

Introduction

The neurological basis of empathy has been a popular subject matter in recent years (e.g. Decety and Jackson, 2004, Decety and Jackson, 2006, Eslinger, 1998). According to the Random House Dictionary (Flexner, 1980), empathy is “identification with or vicarious experiencing of the feelings or thoughts of another person”. Many authors have differentiated between different types of empathy. The most common distinction is between “emotional” and “cognitive” empathy (Davis, 1980, Davis, 1983, Decety and Jackson, 2004, Mehrabian and Epstein, 1972). Cognitive empathy involves the ability to know what another person is thinking or feeling and is similar to the concept of “theory of mind”. Emotional empathy, on the other hand, involves actually experiencing a similar emotion.

Empathy is believed to be important for developing pro-social behavior and for appropriate moral development. It is also seen as important for psychotherapists (Chessick, 1998, Ickes, 1993, Rogers, 1959), parents, and anyone who works with young children (Mehrabian, Young, & Sato, 1988). Deficits in empathy have been found in children with autism and Asperger’s syndrome (Charman et al., 1997, Dapretto et al., 2006, Romanowski-Bashe and Kirby, 2005), in adults with multiple sclerosis (Benedict, Priore, Miller, Munschauer, & Jacobs, 2001), and in psychopaths (Mullins-Nelson, Salekin, & Leistico, 2006).

A number of brain regions have been suggested to be involved in empathy. The strongest evidence suggests involvement of the medial frontal lobes. Patients with lesions to this area have been reported to show deficits in empathy (Eslinger, 1998, Grattan and Eslinger, 1992, Shamay-Tsoory et al., 2003, Varney and Menefee, 1993). A role for the medial frontal lobes has also been suggested by neuroimaging studies of normal subjects. Farrow et al. (2001) used fMRI to study healthy individuals who were asked to judge another’s state of emotion when encountering a social scenario. Activation occurred in the left superior frontal gyrus and orbitofrontal cortex. Vollm et al. (2006) also used fMRI and found activation in numerous regions, including medial prefrontal cortex, the temporoparietal junction, temporal pole, cingulate and amygdala, in men asked to judge the intentions and emotions of characters presented in a cartoon story. Shamay-Tsoory et al. (2005), using PET, found activation of the medial and superior frontal gyrus, occiptotemporal cortices, thalamus and cerebellum in participants undergoing an interview designed to elicit an empathic response about a story character in distress. Ruby and Decety, 2003, Ruby and Decety, 2004 also used PET and asked subjects to answer questions based on their own perspective, or on the imagined perspective of another person. Imagining another’s perspective resulted in activation in frontopolar cortex and right inferior parietal lobe.

In addition to Ruby and Decety, 2003, Ruby and Decety, 2004 finding of right parietal involvement, other studies have suggested that the right hemisphere (RH) may be more involved in empathy than the left hemisphere (LH). Perry et al. (2001) looked at individuals with anterior temporal lobe atrophy and found evidence suggesting greater empathy deficits in those whose lesions involved the RH. Shamay-Tsoory et al. (2003) also found greater empathy deficits in patients with frontal or posterior lesions involving the RH. Rankin et al. (2006) examined patients with various forms of neurodegenerative disease and found that empathy deficits correlated with a decrease in volume in the right temporal pole, right fusiform gyrus, right caudate and right subcallosal gyrus.

Spinella (2002) studied empathy in healthy individuals by using an odor identification task. He found that right-nostril smell identification showed a higher correlation with empathy than left-nostril smell identification. Since the sense of smell relies on ipsilateral pathways, it was suggested that the RH is more involved in empathy than the LH. To measure empathy Spinella used the Mehrabian and Epstein (1972) Empathy Questionnaire (MEEQ). This empathy questionnaire has been validated in a number of previous studies and shown to correlate with empathy-related behaviors such as helping and aggression (for a review, see Mehrabian et al., 1988).

The role of the RH in empathy is congruent with its role in the ability to interpret emotional expression in faces (Adolphs et al., 2000, Etcoff, 1984, Fried et al., 1982, Gur et al., 1994, Warrington and James, 1967) and other social behaviors (Adolphs, 2001, Happe et al., 1999, Rueckert and Pawlak, 2000).

A number of studies have suggested that women may be more empathetic than men. In a literature review and meta-analysis, Eisenberg and Lennon (1983) found that gender differences, favoring women, were highly reliable in a large number of studies utilizing self-report, but less reliable in studies using alternative measures of empathy, such as facial expression or physiological arousal. They suggested that the gender difference may be due to demand characteristics; women may think they are expected to be more caring towards other people, and so are more likely to endorse those items. However, this explanation is called in to question by recent studies that have found evidence of a biological basis for gender differences in empathy. Knickmeyer, Baron-Cohen, Raggatt, Taylor, and Hackett (2006) found a correlation between prenatal testosterone levels and empathy-related behaviors at age four in normal children. Fukushima and Hiraki (2006), using EEG, found a medial-frontal negative component in both men and women 300 ms after they experienced a negative outcome in a gambling game. But only women showed that component after observing a similar negative outcome for their opponent in the game. In addition, Singer et al. (2006) used functional MRI to measure brain activity while participants received mild electric shocks or witnessed a confederate receiving a similar shock. They manipulated participants’ liking of the confederates by having them watch two confederates play a “prisoner dilemma” type of game. In each case, one of the confederates played the game fairly and the other played unfairly. They found that both men and women showed bilateral activation in pain-related areas of the brain (anterior insula and anterior cingulate) when they received a shock, and when they witnessed a “fair” confederate receive a shock. However, only women showed this activation when the “unfair” confederate received a shock. The men, on the other hand, showed activation in reward-related areas (left ventral striatum/nucleus accumbens and orbitofrontal cortex) when they saw the “unfair” confederate shocked.

While some previous studies have suggested a special role for the RH in empathy, others have not found this asymmetry. Given the fact that many studies have reported gender differences in empathy, it is quite possible that the relative role of the RH could differ by gender. Although gender differences in regional activation have generally not been reported in the neuroimaging literature, those studies have typically involved relatively small sample sizes, which may preclude comparisons across gender. The purpose of the present study was to further examine individual differences in RH activation and empathy in a large sample, and to test for potential gender differences. It was hypothesized that there would be a correlation between empathy and the activation of the RH. To measure RH activation, we chose the Levy Chimeric Faces Task (LCFT; Levy, Heller, Banich, & Burton, 1983). This task requires participants to choose the happier of two chimeric faces (faces with one side smiling and the other side showing a neutral expression). Numerous previous studies have demonstrated that right-handed participants (regardless of gender) tend to chose the chimeric face with the smile to their left more often than the chimeric face with the smile to their right (Banich et al., 1992, Levy et al., 1983, Luh et al., 1991, Rueckert, 2005). This is most likely due to the fact that, in right-handers, the RH is dominant for the perception of faces and emotional expression. So when asked to judge the emotional expression of faces, the RH becomes more active. This causes a bias in attention to the left-side of space (Levy et al., 1983). Since the LCFT specifically measures RH involvement in emotion, a stronger correlation with empathy might be expected than that found by Spinella (2002), using lateralized odor identification. As in the study by Spinella, empathy will be measured using the MEEQ, which has shown gender differences in past research (Mehrabian et al., 1988). Although Spinella did not find gender differences, that could have been due to his small sample size.