We agree that our conclusions [1] differ from other studies. This may be a result of the large number of respondents but also because of the detailed examination of several areas of the glans penis and the penile shaft. The study focused on the sensitivity not on erectile function etc. as suggested by Morris et al.

According to Morris et al. the results are biased due to multiple testing and numerically small differences of questionable biological significance between the uncircumcised (Group A) and circumcised (Group B). In the referred articles the human penis was examined as a whole. The present study differentiates itself by the extensive examination of the different areas of the human penis. For the purpose of brevity we skipped the overall significances but reported the components of the areas and the subscales (Table 3). In order to compare our results with those from the literature, one must collapse the three separate sides (dorsal, lateral, ventral) into one variable pertaining to the whole penis and not to its components.

Multiple testing concerns testing of multiple uncorrelated independent variables [2]. The analysis of the highly correlated areas of the penis (glans, shaft, dorsal, ventral and lateral sides) and the high correlation between the main variables (sexual pleasure, discomfort and pain) and their subscales were avoided by not testing them all together as suggested by Table 3, but by sequential analysis. Hereby we first tested whether the two main variables (sexual pleasure, discomfort and pain) differed between both groups for overall penile sensitivity. For the penis as a whole the two groups differed for sexual pleasure (P = 0.044) and discomfort and pain (P = 0.018). In a second step and only when either of them was significant did we test what part of the penis was responsible and if that was significant what side was causing the significance (Table 3). Between the uncircumcised and circumcised groups, sexual pleasure for overall penile sensitivity was strongly significant for the penile glans (P < 0.001; with a 7.5 % difference between the means) but did not differ for the penile shaft. For overall penile discomfort and pain, only the penile shaft was strongly significant (P < 0.001; with an 11.7% difference between the means) and it was not different for the penile glans.

To show the robustness of the conclusions, the results for multiple testing were corrected as suggested by Morris et al. With 42 variables the correction of Sidak [3] showed that a significance of 0.00122 rather than 0.05 should have been used. Even applying this too stringent probability, the same differences remained: sexual pleasure, numbness for the glans and discomfort and pain for the penile shaft.

Morris et al. questions whether differences of 11% are biologically significant. As indicated above there was a statistically significant difference between the two main variables. We leave it up to the readers to judge the biological relevance of these differences.

Morris et al. judge our results of ‘questionable biological significance’. Our findings that circumcision significantly diminishes the sexual pleasure induced by the glans and cause significantly more discomfort and pain by the shaft corresponds to the differences in the pattern of the innervations of the penis.

In Belgium the rate of circumcision is 15%, which is consistent with rates quoted by the WHO for non-religious groups in the UK (15.8%) and the USA (12.8%). The proportion of those circumcised worldwide as indicated by the WHO: ‘If 5% of men in other countries are assumed to be circumcised for non-religious reasons, the global prevalence of circumcision is 33%’ [4].


  1. Top of page
  2. References