Back to XIV Abstract Summary

Back to Invited Symposium Summary


Organizer: Brain, Paul F.
School of Biological Sciences, University of Wales Swansea, Swansea, U.K

Symposium Abstract

There has been much progress since the early ‘primitive’ view that hormones (notably testosterone) ‘cause’ aggression. The papers in this session illustrate some recently revealed complexities in linking physiology and behavior in infrahuman animals and our own species. Sanchez-Martin et al provide tentative evidence that salivary testosterone levels in pre-school (especially boys) may prove a ‘marker’ of conflict behavior. O’Connor et al in a study involving giving testosterone to eugonadal males with partners found no evidence to support the view that supraphysiological levels of the hormone augmented aggression (reported by the patient or the partner). Salvador et al opine that sports competitions are likely to provide socially acceptable situations for studying the links between hormones and behavior in our own species. They also provide tentative support for positive correlations between some measures of offensive behavior in judo contests and testosterone. Haller et al work with rats and suggest that plasma corticosterone levels influence hostile responding in this (and other?) species. They suggest that ultradian variations in this hormone account for changes in expressed behavior.


Sánchez-Martín, J.R., Fano, E. , Cardas, J., Ahedo, L. ,  Brain* P.F. and Azpíroz, A.
Area of Psychobiology, Faculty of Psychology, University of the Basque Country, San Sebastian, Spain. *School of Biological Sciences, University of Wales Swansea, Swansea, U.K.

A study was performed to assess relationships between a series of behavioral measures seen in the interactions of preschool children with their peers (particularly aggressive behavior) and salivary testosterone levels. The subjects were 28 boys and 20 girls of preschool age (4-5 years) videotaped in San Sebastian in free play interactions. Their behavior was subsequently ethologically evaluated, concentrating on levels of isolation, play and social interactions (including giving and receiving aggression and affiliation). Testosterone levels were measured twice using radioimmunoassay but the titers were highly correlated. Significant positive correlations with testosterone on the pooled data (boys and girls) were found for ‘proximity without interaction’; ‘parallel activity’ and receiving aggression in ‘social interactions’. A negative correlation was found between the hormone and ‘play’. When analyzing the data for boys and girls separately, a significant direct correlation was found between testosterone and incidences of aggressive behavior in ‘social interactions’ in boys. This gender also showed inverse relationships between the hormone and ‘play’ and ‘proximity without interaction’. Although the levels of hormone in boys and girls did not differ at this age (most of their androgens are of adrenocortical origin), the girls showed no such relationships. It could be the case, however, that the sexes express their conflict behavior in different ways even at this early age. Testosterone can be a useful biological marker for aggression (and behavioral patterns reflecting different levels of sociability) in children (especially boys). Much more work is needed, however, to evaluate the complex relationships between early hormone titer and current and subsequent behavior.


O’Connor, D.B., Archer*, J., Hair** , W.M. and Wu, F.C.W
Department of Endocrinology, Manchester Royal Infirmary, University of Manchester, Manchester, UK. *Department of Psychology, University of Central Lancashire, Preston, Lancashire, UK. **MRC Centre for Reproductive Biology, Edinburgh, UK.

In nonhuman primates, aggressive behavior correlates directly with testosterone (T) levels. In men, self-reported aggression does not reliably correlate with circulating T levels. This study set out to investigate the behavioral effects of supraphysiological levels of testosterone on self reported and partner reported aggression. Thirty healthy eugonadal men with partners (mean age = 28.2 years; range 19 – 45 years) and 7 hypogonadal men (mean age = 31.43 years; range 23 – 40 years) participated in this single-blind, placebo controlled study. Participants were randomised into two treatment groups (n = 15) to receive: 1) 200 mg testosterone (T) enanthate, intramuscularly, weekly for 8 weeks to raise T levels into the supraphysiological range; the active group or 2) 200 mg sodium chloride, intramuscularly, weekly for 8 weeks; the placebo group. The hypogonadal group received 200 mg T enanthate, intramuscularly, bi-weekly for 8 weeks. All groups completed a battery of behavior measures [Aggressive Provocation Questionnaire (O’Connor et al, in press); Aggression Questionnaire (Buss & Perry, 1992); Aggression Questionnaire-Partner version (O’Connor et al., in press); Rathus Assertiveness Schedule (Rathus, 1973); State Self Esteem Questionnaire (Heatherton & Polivy, 1991); Irritability sub-scale (Buss & Burkee, 1957); Barratt Impulsivity Scale-11 (Barratt & Patton, 1983)] and a blood sample was obtained at baseline, week 4 and week 8. Preliminary analysis has found no statistically significant increases in self reported or partner reported aggression levels in any of the groups, although there was a trend towards increased frequency of aggressive responding to provoking scenarios at week 4 in the active group. No significant changes in assertiveness or self esteem have been found. Surprisingly, the hypogonadal group reported significantly higher levels of hostility at all time points compared to the active and placebo groups. Plasma T levels were generally not significantly correlated with any of the behaviour measures. Cognitive and motor impulsivity emerged from stepwise multiple regression analysis as the most important predictors of self-reported aggression, explaining significant proportions of variance at different time points. These results generally do not offer support to the hypothesis that supraphysiological levels of T lead to increased self and partner reported aggression.


Salvador, A., Suay, A.F.,  Martinez-Sanchis, S., Simon, V.M. and Brain*, P.F.
Area of Psychobiology, Faculty of Psychololgy, University of Valencia, Valencia, Spain.
*School of Biological Sciences, University of  Wales Swansea, Swansea, UK

The role of testosterone in competitive aggression is still open to debate in humans. An important problem is the difficulty of finding adequate behavioral measures. In fact, it has been repeatedly claimed that a main challenge for this topic of research is identifying social situations which permit the study of the role of testosterone in overt behavior comparable to investigation carried out in other species. Sports competitions present several important opportunities to analyse behavior associated with high competitiveness. In particular, contact sports have an especialpropensity to develop into aggressive behavior. With this in mind, the relationships between testosterone levels measured before and after a judo combat and the behaviour displayed during the combat itself were examined. An observational scale was designed from activities usually shown by fighters in judo contests. This scale grouped a number of different elements toform seven behavioural categories that covered the entire period studied. These were threat, fighting, domination, attack/counterattack, defense, observation and stop. Positive correlations were found between offensive behaviours displayed in this kind of competition and testosterone. Furthermore, this relationship supports previous data indicating an association of testosterone with involvement and anger displayed during judo contests as assessed by the coaches. Sports competitions appear to be socially accepted situations that may be used to obtain information about behavior/hormone relationships in our own species.


Haller, J., Fabich, K. and Kruk*, M.R.
Institute of Experimental Medicine, Budapest, Hungary. *Medical University, Leiden, The Netherlands

Ultradian fluctuations in plasma corticosterone have been demonstrated in monkeys and humans. It has been recently shown that corticosterone secretion pulsates in female rats and an attempt was made to assess whether levels also fluctuate in male rats. The amplitude of oscillations was similar with that reported for females but the period was larger in males (a phenomenon perhaps related to the shorter corticosterone half-life in females). Male rats were studied in aggressive encounters lasting only 5 min to reduce interference from fight-induced stress reactions. Male rats were significantly more aggressive in the increasing phase of their corticosterone fluctuation when confronting a male intruder than counterparts in the decreasing phase of their corticosterone fluctuations facing such opponents. Corticosterone fluctuations were artificially mimicked by a combination of treatments with the corticosterone synthesis inhibitor metyrapone and corticosterone. Again, males with increasing plasma corticosterone levels were more aggressive than counterparts with a decreasing plasma corticosterone concentration. It appears that ultradian fluctuations in corticosterone affect the propensity of an animal to behave aggressively. This suggests that the behavioural response to an aggressive challenge may vary in the same animal across the day due to the pulsating nature of corticosterone secretion.