INTRODUCTION

~ [Page 511] ~

The term pedophilia denotes the erotic orientation of individuals (usually men) whose sexual interest in pre-pubescent children exceeds their sexual interest in physically mature adults (Freund, 1981). Most clinical authorities differentiate between

	true pedophiles versus
	men who, despite an erotic preference for the mature physique, have molested children in particular circumstances, for example, while they were intoxicated (Barbaree & Seto, 1997).

Several studies have examined the cognitive neuropsychological functioning of sex offenders against children (less often, homogeneous groups of diagnosed pedophiles), using a variety of non-pedophilic control groups.

The most frequently assessed neuropsychological characteristic has been general intelligence or IQ.

	Some investigators have compared sex offenders against children with non-criminal community control groups, finding the sex offenders to score lower in intelligence (Langevin et al., 1985).
	Other investigators have compared sex offenders against children with men convicted of nonsexual crimes, thus controlling for their status in the legal system and any anti-sociality. The offenders against children again scored lower.

[Page 512]

(Hambridge, 1994; Hucker et al., 1988; Langevin, Wortzman, Dickey, Wright, & Handy, 1988; Langevin, Wortzman, Wright, & Handy, 1989).

Still other investigators have compared sex offenders against children with sex offenders against adults, thus controlling not only for legal status but also for the demonstration of sexually atypical behavior.
Once again, the offenders against children scored lower
(Blanchard et al., 1999; Ellis, 1951; Frosh & Bromberg, 1939; Henn, Herjanic, & Vanderpearl, 1976; Quinsey, Arnold, & Pruesse, 1980).

Similar conclusions have been reached by investigators who used the Halstead-Reitan and Luria-Nebraska Neuropsychological Batteries and who found that sex of-enders against children showed greater impairment than offenders committing nonsexual crimes

(Scott, Cole, McKay, Golden, & Liggett, 1984)

and offenders committing sexual crimes against adult women

(Hucker et al., 1986; Langevin et al., 1988, 1989).

Many of the foregoing studies had one or more methodological weakness and the occasional comparison produced inconsistent results; however, the bulk of evidence indicates that pedophiles show poorer cognitive functioning in comparison with any reasonable control group.

Another neuropsychological variable that is routinely included in clinical neuropsychological assessments is hand-preference. Two studies have assessed the relation between this variable and pedophilia. Bogaert (2001) found that pedophiles have elevated rates of left-handedness. Cantor, Christensen, Klassen, Dickey, and Blanchard (2001) confirmed this relation and showed that it remained significant when they controlled for IQ.

Al-though this is not the main reason we introduce the handedness findings in this paper — our purpose is explained below — it should be noted that these findings argue against the notion that any observed neuropsychological differences between pedophiles and their comparison groups are likely to be mere artifacts of ascertainment bias.

This is relevant because the previously reviewed findings on pedophilia and IQ could conceivably be interpreted as such artifacts: Less intelligent pedophiles are more likely to be apprehended (or be unable to afford the best lawyers); therefore convicted pedophiles will have lower mean IQs. One might plausibly argue that less intelligent pedophiles are more likely to be apprehended, but one cannot plausibly argue that left-handed pedophiles are more likely to be apprehended.

The present significance of the handedness findings and the IQ findings lies in the information they provide about neurodevelopment.

A variety of research has shown that perturbations in neurodevelopment can lead to deficits in cognitive functioning. Data from human studies that support this observation can be gleaned from several sources.

First, acquired neurologic damage during infancy or early childhood has profound and long-lasting cognitive effects. This has been demonstrated among children with

brain tumors
(Radcliffe, Bunin, Sutton, Goldwein, & Phillips, 1994),

traumatic brain injury
(Taylor et al., 1999),

intra-cranial hemorrhage

(Dennis & Barnes, 1994),

peri-natal hypoxia
(Gottfried, 1973), and

epilepsy
(Neyens, Aldenkamp, & Meinardi, 1999).

Second, exposure to neuro-toxic substances, either in utero or early in postnatal development, can have similar robust effects on cognition. Such effects have been associated with several teratogenic substances including

lead
(Needleman, Schell, Bellinger, Leviton, & Allred, 1990),

coumarins
(Wesseling et al., 2001),

alcohol
(Olson, Feldman, Streissguth, Sampson, & Bookstein, 1998), and

tobacco
(Frydman, 1996).

Third, genetic disorders, with known adverse neurobiological effects, have also been connected with low cognitive functioning. For example, children with

fragile X syndrome
(Fisch et al., 1996),

velocardio-facial syndrome
(Kozma, 1998), and

Down’s syndrome (Hayes & Batshaw, 1993)

typically demonstrate significant intellectual impairment.

Considered collectively, the foregoing studies strongly suggest that a variety of ad-verse neuro-developmental events/conditions can lead to poor intellectual functioning.

Another line of research has shown that perturbed neurodevelopment increases the probability of left-handedness (Coren & Halpern, 1991). Typically, these studies have found that rates of left-handedness are elevated among individuals exposed to pre- or peri-natal neuro-developmental disruptions.

Examples of neuro-developmental risk factors include a variety of events and conditions:

	prenatal exposure to neurotoxins (Biro & Stukovsky, 1995),
	ultrasound (Kieler, Cnattingius, Haglund, Palmgren, & Axelsson, 2001),
	extremely low birth weight (Saigal, Rosenbaum, Szatmari, & Hoult, 1992),
	twin births (Coren, 1994),
	prematurity (Marlow, Roberts, & Cooke, 1989), and
	markers of birth stress (Williams, Buss, & Eskenazi, 1992).

The fact that pedophilia is associated with left-handedness and with poor cognitive functioning, two variables that are causally related to neuro-developmental perturbation, suggests that pedophilia may also be causally related to neuro-developmental perturbation. It is possible, in other words, to obtain correlations among pedophilia, left-handedness, and poor cognitive functioning, because neuro-developmental problems predispose a male to develop all three.

Neurodevelopment is not complete at birth. It continues throughout childhood and, in certain regards, up to and during adulthood.

Neurodevelopment can be broadly divided into two phases (Lund, 1997).

[Page 513]

	During the first phase, which occurs pre-natally in most mammals, new neurons are formed (neurogenesis), move to their correct positions (neuronal migration), and elaborate the primary neuritic elements that constitute emerging dendritic fields (Caviness, 1989). The formation of new synapses is also central to this phase (Lewis, 1997).
	During the second phase, which occurs both pre- and postnatally, the fine details of neuronal structure and the organization of circuitry are accomplished via cell death, axonal pruning, and synaptic elimination (Caviness, 1989). In addition, environmental sensory modulation of synaptic connectivity occurs (Lund, 1997).

Although much of the second phase is completed in the first few postnatal years, there are significant refinements in neuronal structure and brain circuitry during later periods of development.

For example, the dentritic spines of pyramidal neurons in cortical layer III exhibit rapid attrition until approximately 4.5 years of age in the monkey

(Anderson, Classey, Conde, Lund, & Lewis, 1995).

Moreover, the density of axospinous synapses declines from approximately 3 years of age through adulthood and the density of GABAergic chandelier neurons declines from the age of 1.5 years to stable adult levels following puberty (

Anderson et al., 1995).

Similarly, several changes in several neurotransmitter systems extend into late postnatal development. Dopamine varicosities undergo a marked increase in density until the time of puberty (Plant, 1988) whereas the density of cholecys-tokinin neurons in Layers 1 and 3 steadily decline post-natally

(Lund & Lewis, 1993).

Additionally, studies utilizing structural neuro-imaging methods (e.g., MRI) have demonstrated that morphometric brain changes continue well into adulthood (Jernigan & Sowell, 1997). For ex-ample, Pfefferbaum et al. (1994) found that overall brain growth persists until 10 years of age, gray matter volume increases until approximately 4 years of age, and white matter volume increases until around 20 years of age. In addition, there was some evidence for changes in sulcal volume until 30 years of age.

The sequelae of brain damage vary as a function of the developmental stage during which the injury occurred

(Kolb, Gibb, & Gorny, 2000).

This has been observed across several laboratory species including rats, cats, and monkeys

(for reviews, see Almli & Finger, 1984; Finger & Almli, 1984).

The question therefore arises as to the time frame during which neuro-developmental perturbations could result in pedophilia.

One indication comes from research on handedness.

Handedness appears to be determined during fetal development

(e.g., Hepper, Shahidullah, &White, 1991).

It follows that if the observed correlation between pedophilia and left-handedness is correctly attributed to neuro-developmental perturbations that produced both, then those perturbations most likely occurred during the pre-natal or peri-natal period.

The possibility that neuro-developmental problems before birth may increase a male’s risk of pedophilia does not preclude the possibility that neuro-developmental problems after birth may also increase the risk of pedophilia.

In fact, our review of postnatal brain development indicates that the latter possibility is quite feasible. The present study was therefore conducted to look for evidence that might support it. We were interested, in particular, in the correlation between pedophilia and traumatic head in-juries.

Head injuries are certainly capable of affecting neurodevelopment and they are relatively accessible through patients’ self-report. The finding of a positive correlation would, of course, have implications for the etiological theory of pedophilia. It would also, however, have implications for normal psychosexual development, in that it would suggest that at least one dimension of erotic object-choice — the age of the persons to whom a man is most attracted — is not completely determined before birth.