In their target article, Dijkerman & de Haan (D&dH) review results suggesting that cortical somatosensory processing is organised into distinct functional streams, and distinguish between (a) processing for action versus perception, on one hand, and between (b) processing for external target versus stimuli related to the body itself, on the other hand. These distinctions are related to different neural pathways and neuroanatomical correlates.

According to numerous ethologists, sexual behaviour is one of the most important goal-directed behaviours, and the brain plays a key role at each step of its development (Pfaus Reference Pfaus1999). Ignored during a long period, the main results regarding this question have come from animal studies, which often remain difficult to extrapolate to humans. However, research on the neural correlates of human male sexual arousal has been developing in recent years thanks to the use of modern neuroimaging tools such as positron emission tomography (PET) and functional magnetic resonance imaging (fMRI). Until recently, a classic approach used for the statistical analysis of data followed a subtractive approach that contrasted brain activations recorded during a period of sexual stimuli presentation with those recorded during a control condition. Such an approach has produced many interesting results (for reviews, see Mouras & Stoleru Reference Mouras, Stoleru, Kandeel, Lue, Pryor and Swerdloff2007; Stoleru & Mouras, in press). The presentation of sexual stimuli induces a great number of behavioural responses, ranging from emotional to autonomic reactions. However, neuroimaging regression models often contain hypothetical temporal functions (block-type functions, for example, to describe periods of different experimental conditions). Because of the large set of neural responses involved, one can argue that models that include more specific and objective functions would be more suited to identifying the brain circuits involved.

Penile plethysmography, an objective method of measuring genital response, was used successfully within an fMRI scanner (Arnow et al. Reference Arnow, Desmond, Banner, Glover, Solomon, Polan, Lue and Atlas2002; Ferretti et al. Reference Ferretti, Caulo, Del Gratta, Di Matteo, Merla, Montorsi, Pizzella, Pompa, Rigatti, Rossini, Salonia, Tartaro and Romani2005; Moulier et al. Reference Moulier, Mouras, Pelegrini-Issac, Glutron, Rouxel, Grandjean, Bittoun and Stoleru2006). Pertinent to the purpose of this commentary, Arnow et al. (Reference Arnow, Desmond, Banner, Glover, Solomon, Polan, Lue and Atlas2002) did not identify many brain areas by performing subtractive analyses, whereas in many areas the blood-oxygen level–dependent (BOLD) signal was correlated with the penile response as measured by circumferential plethysmography. In this study, the highest correlation was reported in the right insular/subinsular region, in accordance with the role of insula in somatosensory processes (Augustine Reference Augustine1996). In a recent study (Moulier et al. Reference Moulier, Mouras, Pelegrini-Issac, Glutron, Rouxel, Grandjean, Bittoun and Stoleru2006), we used the higher sensitivity of volumetric plethysmography (Kuban et al. Reference Kuban, Barbaree and Blanchard1999) to distinguish brain areas more involved in penile-response command from those involved in genital response cortical representation, by manipulating the time window separating erectile and cerebral responses. Interesting results have been found pertaining to the distinctions made by D&dH.

Whereas the correlational analyses clearly identified insulae bilaterally, our data suggested a role of the anterior portion of the insula in mediating the erectile response (i.e., the genital response, a cross-correlation led by the cerebral signal). This was in accordance with the report of diastolic blood pressure and heart rate variations after electrical stimulation of the anterior insula (Oppenheimer et al. Reference Oppenheimer, Gelb, Girvin and Hachinski1992). Conversely, a role of the posterior portion of the insula in sensory integration was suggested by a cross-correlation led by the plethysmographic signal (for a voxel located in x=36, y=−23, z=15 mm). Accordingly, sexual stimulation of the penis has been shown to strongly activate the right posterior insula (Georgiadis & Holstege Reference Georgiadis and Holstege2005), especially for a very close location (x=36, y=−24, z=14 mm).

Different areas were also distinguished with regard to penile cortical representation. A positive correlation was found in the paracentral lobules, which are located on the medial surface of the hemisphere and correspond to the first localization of the penile cortical representation of the homunculus (Penfield & Rasmussen Reference Penfield and Rasmussen1950). However, in the same correlational analysis a second location was found more on the lateral surface of the somatosensory cortices; this location became larger and had a higher correlation coefficient when the plethysmographic signal preceded the BOLD signal by 10 seconds. This result was in accordance with previous recordings of evoked cortical potentials by stimulation of the dorsal nerve of the penis (Bradley et al. Reference Bradley, Farrell and Ojemann1998). Following these results, we then proposed a distinction between two distinct penile representations: (a) one on the medial surface of somatosensory cortices, corresponding to the original one and related to the perception of genital response proper; and (b) one on the lateral surface of the cortices, which would be more related to superficial skin stimulation (Kell et al. Reference Kell, von Kriegstein, Rosler, Kleinschmidt and Laufs2005).

Finally, our correlational approach allowed us to identify high correlations between the cerebral and genital responses in regions related to motor preparation and imagery processes, such as left inferior parietal lobule, frontal opercula, ventral premotor area, and supplementary motor areas.

Thus, many of the target article's main arguments are largely supported by studies of the neural correlates of human penile response using, in particular, correlational analyses between autonomic genital response and hemodynamic cerebral response. Overall, the results of such studies support the central role of the insula in penile sensation awareness – that is, in higher-order somatosensory processing. Consistent with D&dH's viewpoint (see sect. 3.2 of the target article), the insula could be involved in the emotional component of sexual arousal (Redoute et al. Reference Redoute, Stoleru, Gregoire, Costes, Cinotti, Lavenne, Le Bars, Forest and Pujol2000). Another striking point of convergence between data reviewed by D&dH and our results in the Moulier et al. (Reference Moulier, Mouras, Pelegrini-Issac, Glutron, Rouxel, Grandjean, Bittoun and Stoleru2006) study is that we found that the magnitude of penile tumescence drives, with a lag of about 10 seconds, the BOLD signal recorded from the paracentral lobule (part of the posterior parietal cortex). This suggests that this area is involved in the command of the penile response – which is again clearly consistent with the role of Brodmann 5 in goal-directed actions as reported by D&dH.

In conclusion, the recent application of penile plethysmography within the neuroimaging setting allows the study of the correlation between the cerebral and autonomic responses. Moreover, the high sensitivity of volumetric plethysmography enables brain areas more related to the conduction of the erectile response to be distinguished from those areas more related to the encoding of penile sensation, once the autonomic reactions appear. The high correlations found for areas related to motor imagery underline the central role of physiological reactions in goal directed behaviours.