INTRODUCTION

Alcohol abuse among adolescents is a widespread and growing threat to the health of young people in industrialized countries (Bailey and Valery, 1993), where almost 35% of habitual drinkers are under 16 years old (Shope et al, 1994). A growing number of young adolescents drink alcoholic beverages every weekend.

Our group has reported the deleterious effects of alcohol abuse on the reproductive and growth axis of adolescents (Frias et al, 2000a,2000b). An important question to be addressed is why increasing numbers of adolescents consume alcohol, despite the adverse and probably long-lasting effects of this drug on their behavior and physiology.

Owing to the pace of life in modern industrialized society, among other factors, adolescents may feel that their affective needs are not fully met, leading to the development of symptoms of anxiety, irritability, and depression. This could in part explain why these young people seek the anxiolytic and rewarding properties of alcohol, a socially permitted drug.

Alcohol positively modulates gamma-aminobutyric acid type A (GABAA) receptors and produces sedative and anxiolytic effects. There is evidence for interaction between alcohol and neurosteroids at GABAA-receptors (Grobin et al, 1998). The neurosteroid allopregnanolone is one of the most potent endogenous positive modulators of GABAA-receptors and has sedative, hypnotic, anticonvulsant, and anxiolytic effects (Lambert et al, 1995). In animals, it has been suggested that allopregnanolone produced in response to systemic ethanol administration could contribute to several of the effects of ethanol that are associated with the modulation of GABAA-receptors (Morrow et al, 2001a). Neurosteroids have been shown to be rewarding in rodents (Finn et al, 1997). Therefore, neurosteroids may contribute to the anxiolytic and rewarding effects of alcohol.

Many animal studies have investigated the effect of alcohol on neurosteroid concentrations (Barbaccia et al, 1999). However, we know of no studies that have demonstrated the effect of alcohol on neurosteroids in adolescent humans, possibly because of the difficulty to recruit appropriate study subjects. Studies involving the controlled administration of alcohol are evidently ruled out in adolescents on ethical grounds.

The present study aimed to determine allopregnanolone levels in female adolescent humans in situations of acute alcohol intoxication (AAI).

MATERIALS AND METHODS

A study was conducted of 35 white female adolescents (13–17 years old) of variable weight, height, and educational level who arrived at the emergency department of our hospital between 12.00 and 3.30 am. The subjects were assigned to one of the two study groups. One group (AAI, n=20) was formed by those arriving at the emergency department with evident behavioral symptoms of AAI (slurred speech, unstable walking); after remission of these symptoms, all of them reported that they habitually became drunk every weekend. The second group (controls, n=15) comprised young women who arrived at the department for mild trauma (contusions, sprains) and reported no alcohol consumption. All subjects were interviewed regarding their current ovarian phase and any drug treatments, in the case of AAI after the remission of their symptoms. No subjects were receiving medication at the time of the study or suffered any apparent endocrine disorder. All participants, or their parents, gave their informed consent for participation in the study, which was carried out in accordance with the Helsinki Declaration of 1975.

Blood samples were extracted from all subjects on arrival at the emergency area. Serum samples were frozen and stored at −20°C. Assessment of progesterone concentrations was performed by radioimmune assay (RIA), using commercially available Sorin Biomedica kits (Vercelli, Italy) without modifications. Allopregnanolone was measured by RIA as previously described (Torres et al, 2001), using an antiserum kindly provided by Roussel–Uclaf. The plasma samples were extracted with diethyl ether (Genazzani et al, 1998) and the extracts were purified according to Corpechot et al (1993). Since the ovary is an important source of circulating progesterone and allopregnanolone in humans (Rapkin et al, 1997; Genazzani et al, 2002), the study included women in follicular (AAI, n=10; controls, n=7) or luteal (AAI, n=10; controls, n=8) phases of their ovarian cycle. Their consumption of alcohol was confirmed by the measurement of blood alcohol concentrations, using a gas-chromatographic head-space method described previously by our group (Frias et al, 2002).

Statistically significant differences between the groups were analyzed by a two-way ANOVA. Since the Levene test for equality of variance showed significance, logarithmic transformation was performed, meeting the assumption of homogeneity of variance. The Bonferroni method was used in this study. The SPSS version 9.0 for Windows software package was used in the statistical analysis. Results are expressed as mean±SEM.

RESULTS

Serum levels of allopregnanolone and progesterone in the AAI and control groups, respectively, are shown in Figures 1 and 2. Two-way ANOVA of the allopregnanolone results showed significantly higher allopregnanolone levels in the AAI group vs controls (F1,31=75.33, P<0.001) and significantly higher allopregnanolone levels in luteal vs follicular phase (F1,31=47.36, P<0.001). Application of the Bonferroni method showed that the AAI females had significantly higher allopregnanolone levels vs controls in both follicular and luteal phase (P<0.01). With respect to progesterone values, significantly higher levels were found in AAI women vs controls (F1,31=59.14, P<0.001), and progesterone levels were significantly higher in luteal than follicular phase (F1,31=770.07, P<0.001). Application of the Bonferroni method showed that progesterone levels were significantly higher in AAI women vs controls in both follicular and luteal phase (P<0.01).

Figure 1
figure 1

Serum allopregnanolone in AAI and controls (C). The data are expressed as mean±SEM. The Bonferroni method showed statistical differences in AAI vs C P<0.01 in both follicular and luteal phases of the ovarian cycle.

Full size image
Figure 2
figure 2

Serum progesterone in AAI and controls (C). The data are expressed as mean±SEM. The Bonferroni method showed statistical differences in AAI vs C P<0.01 in both follicular and luteal phases of the ovarian cycle.

Full size image

There was no significant difference in the blood alcohol level of the AAI adolescents between follicular and luteal phases (108±4 vs 99±3.5 mg/dl, respectively).

DISCUSSION

According to the present results, AAI raises the plasma concentrations of progesterone and allopregnanolone in adolescent females during both the luteal and follicular phases of their ovarian cycle. These findings are consistent with a previous report that alcohol injection increased plasma concentrations of allopregnanolone in rats (Barbaccia et al, 1999). Although the effects of ethanol on the hepatic metabolism of steroids should be taken into account, our data may indicate that alcohol ingestion promotes neurosteroidogenesis in peripheral organs. A previous study reported that exogenously administered allopregnanolone attenuated the stress-induced release of corticotropin-releasing factor (CRF), adrenocorticotropin hormone (ACTH), and corticosterone and blunted the anxiogenic effect of CRF. Our group earlier reported that AAI produces a high increase in pituitary–adrenal axis hormones in female adolescents (Frias et al, 2000a). Circulating allopregnanolone is increased in humans and rats after CRF and ACTH administration (Genazzani et al, 1998; Torres et al, 2001). Allopregnanolone may therefore represent a novel feedback of the hypothalamic–pituitary–adrenal axis, as previously proposed (Purdy et al, 1991).

The ovary and adrenal cortex are sources of circulating allopregnanolone. Although we could not know the contribution of each source, we observed an increase in serum allopregnanolone in our AAI females in both follicular and luteal phase. Earlier reports demonstrated that circulating allopregnanolone crosses the blood–brain barrier (Morrow et al, 2001b) and may exert its anxiolytic, analgesic, anticonflictive, and rewarding effects in the CNS through allosteric modulation of GABAA-receptors.

The increase in allopregnanolone in response to AAI in female adolescents may have major biological repercussions. Allopregnanolone may increase the anxiolytic or rewarding effects of this socially permitted drug either directly, or by increasing the sensitivity of GABA receptors to alcohol.