The suppression of appetite and food consumption by methylphenidate: the moderating effects of gender and weight status in healthy adults

Abstract

Females typically show greater behavioural responses to stimulant drugs than males, including loss of appetite; as seen, for example, in those who use methylphenidate (MP) therapeutically for treatment of attention deficit hyperactivity disorder (ADHD). This is a relevant issue because of the strong link between ADHD and obesity. In a sample (n=132) of normal-weight (BMI <25) and obese (BMI >30) men and women we assessed appetite, cravings, and snack-food intake in response to MP (0.5 mg/kg) and placebo. Results indicated a significant three-way interaction for the three dependent variables – food-related responding diminishing in all groups from placebo to MP, except in obese males who showed no decreases to the MP challenge. These data show for the first time the existence of gender differences in the appetite response to MP, and are relevant for finding a dopamine pathway to new weight-loss medications, which would be utilized differently in males than in females.

Introduction

Efforts to identify neurobiological risks for drug abuse have found relatively consistent gender differences in the behavioural response to psychomotor stimulants. For example, compared to male rats, females display higher methylphenidate (MP)-induced hyperactivity (Wooters et al.2006), greater reinforcement response to cocaine and amphetamine (Becker et al.2001), and stronger reinstatement of cannabinoid-seeking after a drug or a cue priming (Fattore et al.2010). The findings that female rats show a greater response to amphetamine on the oestrus phase of their cycle (Becker et al.2001), and that ovariectomy attenuates their cocaine-stimulated locomotor behaviour (Johnson et al.2010), suggest that oestrogen mediates these behavioural sex differences. Such an interpretation is consistent with the evidence that oestrogen inhibits GABA neurons in the striatum and accumbens, which thereby increases mesolimbic dopamine function in females (Becker, 1999). In addition, female rats show a higher activity level of the dopamine transporter (DAT), the vesicular monoamine transporter 2 (VMAT2) (Dluzen & McDermott, 2008), and enhanced nigrostriatal dopamine neurotransmission in response to dopaminergic drugs compared to their male counterparts (Walker et al.2006).

In the human condition, there is also some evidence of sex differences in dopamine functioning and in sensitivity to psychostimulant drugs – although the data are sparse, and sometimes indirect or compromised. For instance, a recent study found that low doses of d-amphetamine functioned as a reinforcer in women, but not men, whereas at high doses the reverse pattern of reinforcement occurred (Vansickel et al.2010). Some clinical research has found that women tend to progress more rapidly from drug use to abuse (Giffin et al.1989), are more sensitive to the subjective effects of cocaine, and report stronger drug cravings (Elman et al.2001). They also show a lower incidence of Parkinson's disease than men because oestrogen is thought to exert neuroprotective effects in women (Shulman & Bhat, 2006).

One of the most common behavioural responses to stimulant drugs is a reduction in food intake with concomitant weight loss, as observed in those using stimulants therapeutically – for example, in the case of attention deficit hyperactivity disorder (ADHD) – and in controlled laboratory studies with healthy adults (Goldfield et al.2007). This finding is clinically significant since ‘loss of appetite’ is a common side-effect impediment to the potential benefits of ADHD stimulant medication (Karabekiroglu et al.2008). It is also a relevant issue because of the strong links between ADHD and obesity (Davis, 2010), and the evidence that stimulants are an effective treatment for intractable obesity in those with symptoms of ADHD (Levy et al.2009).

To date, no studies have examined gender differences in stimulant-induced suppression of appetite. In the present study, we assessed the moderating effects of gender and weight status (normal weight vs. obese) on appetite, cravings, and food intake in a controlled laboratory study after MP (0.5 mg/kg) and placebo using a double-blind, cross-over design. We predicted that women would experience a greater decrement in appetite ratings, in food cravings, and in food intake than men in response to MP. There is no prior evidence to suggest this effect would be different in normal-weight compared to obese participants.

Methods

Participants

One hundred and thirty-two adults between the ages of 24 and 45 yr took part in the study (97 women, 35 men). Participants were fluent in English and had lived in North America for at least 5 yr prior to their enrolment. All female participants were pre-menopausal as identified by the self-reporting of regular menstrual cycles, and did not have a pregnancy within the previous 6 months. Exclusion criteria included a current diagnosis (or history) of any Axis I disorder except unipolar depression, including substance abuse, alcoholism, and ADHD, as determined by the Structured Clinical Interview for the Diagnostic and Statistical Manual – IV (APA, 2000)†, or any serious medical/physical illness such as cancer, heart disease, or paralysis. Participants were also free of all medications contraindicated for MP. Only one participant was currently diagnosed with major depressive disorder. Forty-four percent of the sample were of normal weight (40 women, 17 men) while the remainder (57 women, 18 men) were obese (BMI >30). The obese group was significantly (p=0.001) older (34.3±6.5 yr) than the normal-weight group (30.7±5.3 yr). However, there was no gender difference or a significant gender×group interaction. With respect to body mass index (BMI), there was also no significant gender effect, or a gender×group interaction. Finally, 19% of the sample were cigarette smokers but the proportions did not differ by gender and group (χ²=0.91, p=0.340).

Procedure

On the first day of testing, informed consent and demographic information were obtained, and height, weight, and blood pressure were taken. In preparation for the food challenge, subjects were asked to indicate their ‘favourite snack food’ with the proviso that it did not need to be cooked (like French fries) or kept frozen (like ice cream). If the subject first indicated a ‘healthy’ option like an apple, the interviewer probed further until the subject selected a conventional sweet or salty snack. The most common choices were chocolate bars, potato chips, and cookies.

For each subject, the two drug-challenge testing sessions were scheduled at the same time of day, and the same day of the week, separated by 1 wk, and of 2 h duration. Subjects were instructed to eat a normal meal 2 h before their appointment, and to abstain from drinking a caffeine beverage or smoking a cigarette on these two testing days. Upon arriving at the laboratory, attendance to the dietary restrictions was confirmed. No participants had to be re-scheduled for reasons of non-compliance. Blood pressure was then taken and the capsule was ingested. Since the peak uptake of MP is ∼1 h, subjects were seated comfortably in a quiet room after ingesting the capsule, and provided with light reading material. Seventy-five minutes after administration of the capsule, the snack food was presented to the subject, and the following variables were measured.

All procedures were performed in accordance with the Declaration of Helsinki.

Appetite ratings

Participants were given their ‘favourite snack’ by the experimenter and asked to respond verbally to the following questions using a scale from 1–10 (1=‘not at all’, 10=‘a great deal’). They were also encouraged to try and answer each question without concern for calories or whether they would normally eat such snacks at that time of day. Ratings for the three questions were summed for the placebo day and the drug day to give a total appetite score for each session.

1. How hungry does it make you feel to see your favourite snack?

2. How much would you like to eat some of this snack – even just a small portion? (at this point the participant was asked to eat a piece of the snack)

3. Now that you've had a taste of your favourite snack, how strong is your desire to have some more?

Food cravings

The 15-item state version of the General Food Cravings Questionnaire (Cepeda-Benito et al.2000) was used to assess situational food cravings personalized for each participant's specific snack food. This scale has been validated in a heterogeneous sample of eating-disordered patients (Moreno et al.2008) and with indices of food deprivation and satiety (Nijs et al.2007). In the present study, the α-coefficients for the placebo and drug days were 0.93 and 0.94, respectively.

Snack-food consumption

After responding to the appetite questions and completing the craving questionnaire, participants were told they could eat as much of the snack as they wished. At the end of the 2-h testing session the snack was again weighed. The snack was weighed to the nearest gram at the beginning and the end of each testing session. For each individual, the snack was identical on the placebo and the drug days. Consumption was quantified by subtracting the end-of-session weight from the weight of the whole snack, and then converting the difference to a percentage of the initial weight. Therefore, possible values range from 0 when nothing was eaten to 100% when the entire snack was consumed.

Results

Since the participants chose different types of snacks, we first performed a 2 (gender)×2 (weight status) analysis of variance (ANOVA) to assess group differences in pre-consumption snack weight. There were no significant main effects, or a significant interaction, indicating equivalence in snack weight across the four experimental groups.

Separate repeated-measures ANOVAs were performed for the three food-related dependent variables with gender and weight as the between-subjects factors and day (placebo vs. drug) as the within-subjects factor (Fig. 1).

Appetite ratings

The three-way interaction was statistically significant (p=0.010). Post-hoc analyses indicated no gender difference or a gender×day interaction in the normal-weight group. There was, however, a significant day effect (p=0.017) showing a decrease in consumption from placebo to drug condition. By contrast, in the obese group, there was a highly significant (p<0.007) gender×day interaction, which indicated a highly significant (p<0.0001) decrease in consumption in women from placebo to drug, but no difference between the two conditions in men.

Food cravings

Again, there was a significant three-way interaction (p=0.008). Similar to the previous analysis, there was only a day main effect (p<0.0001) in the normal-weight group – showing a decrease in cravings from placebo to drug – while the day×gender interaction was significant in the obese group (p=0.008). There was a highly significant decrease in cravings (p<0.0001) from placebo to drug in women, but no difference between the two conditions in men.

Snack-food consumption

In accord with the previous two analyses, there was also a significant three-way interaction (p=0.014), indicating as before, no gender difference, or a gender×day interaction, in the normal-weight group, but a significant decrease from placebo to drug (p<0.017). In the obese group, there was a significant day×gender interaction (p<0.0001), which indicated a significant decrease in appetite ratings from placebo to drug in women (p<0.0001), but no difference between the two conditions in men.

At the end of the study we asked participants to indicate which day they thought they had taken the MP, and which day the placebo. Seventy-eight percent (n=103) of the participants guessed correctly. Among those, there were no group and gender differences as indicated by χ² analysis (p=0.382). Similarly, there were no differences among the remainder who guessed incorrectly (p=0.793).

In order to assess the possible influence of adverse side-effects on our results, we also asked participants to record, on a visual analogue scale, how much they ‘liked’ or ‘disliked’ being on the drug, where the mid-point of the line indicated ‘no preference’. A group×gender ANOVA indicated no main effects or interactions on this variable. Moreover, the mean score was almost exactly in the middle of the line with approximately equal numbers in each half of the distribution.

Discussion

In the present study, the effects of a moderate dose of MP (0.5 mg/kg) on appetite ratings, cravings, and snack-food consumption were compared among healthy adult men and women, classified either as normal weight or obese. Results indicated a highly significant gender×weight×drug day interaction for all three dependent variables. Contrary to expectation, we did not find any gender differences in the normal-weight group, although both men and women showed significant decreases from placebo to MP. While the expected food-related decreases to MP were also seen in obese women, the obese men showed small (although not statistically significant) increases from placebo to MP on all three variables, implying a crucial role of adipose tissue in addition to, or in combination with, sex/hormonal effects in response to MP.

We are aware of only two prior studies that have assessed the influence of MP on eating and appetite in adults. The first (Goldfield et al.2007) included both normal-weight and overweight/obese adults and found reduced energy intake following MP, but did not provide any gender comparisons. The second (Leddy et al.2004) only assessed obese men, and found – differently from our results – that participants ate significantly less pizza at a laboratory test meal after a 0.5 mg/kg and a 1.0 mg/kg dose of MP, compared to placebo. However, both these studies had very small samples (14 and 9 adults, respectively), which tends to compromise the reliability of their findings. Because of the gender-specific hormonal changes that occur with puberty, the few studies investigating appetite and eating responses to MP in prepubescent children (e.g. Leddy et al.2009; Sonuga-Barke et al.2007) are also of limited utility in understanding the gender and obesity effects found in our sample of adult participants. To our knowledge, no pre-clinical studies using obese strains of animals (or clinical studies of obesity) have examined sex differences in response to stimulant drugs. Therefore, in the absence of any directly related previous research, we can only speculate on the possible mechanisms underlying our results.

The finding that the women in our study, irrespective of weight category, experienced pronounced MP effects could be ascribed to the hormonal influences of oestrogen on brain neurotransmission (Becker, 1999). Indeed, female sex hormones are known to modulate the expression of components of dopamine (e.g. DAT) and serotonin (e.g. 5-HT_2C receptor) systems in mesolimbic nuclei, which are important in the response to psychostimulants (Zhou et al.2002). For example, animal studies have found that DAT binding-site density and DAT mRNA levels in dopamine cell bodies, as well as DAT activity in striatal regions, are significantly lower in males than in females (Dluzen & McDermott, 2008; Rivest et al.1995).

Of interest is the evidence that BMI is inversely associated with striatal DAT availability in healthy volunteers (Chen et al.2008). Because of this – and the fact that a substantial concentration of MP may also be sequestered in fat tissue due to its lipophilicity – one might predict that obese subjects would be less responsive to the drug challenge. However, the presence of the dopamine stimulating effects of oestrogen, which increases DAT density and activity (Chavez et al.2010) and serotonin 5-HT_2C receptor levels (Zhou et al.2002) in brain areas that regulate the hedonic aspects of food, appears to render obese women still able to experience the effects of MP, regardless of its reduced availability by fat tissue.

It is important to note that because menstrual-cycle phase was not assessed in our study, the stage of cycle during which MP was administered to the women could be a potential confounding factor in our research. However, there is no reason to expect a systematic bias between the normal-weight and obese women in this regard, and it is probably safe to assume roughly equal representation of each menstrual stage across the two female groups. We must also acknowledge that confirmation of smoking and caffeine abstinence on the drug-challenge testing days was only obtained by self-report, not by more objective means such as carbon monoxide assessed by expired breath analysis in the case of smoking. Finally, although no participants reported any adverse events during the drug-challenge assessment period, the study is also limited because side-effects were not formally assessed.

In summary, these data show, for the first time, the existence of gender differences in food-related responses to MP, which appear to be moderated by excess adiposity. One strength of our research is the large adult sample we employed compared to previous related studies. In addition, our findings were consistent across the three separate measures of appetite and food intake we employed. However, the reliability of these drug-induced responses can only be confirmed with replication, and further research is needed to assess these effects in the broader context of food consumption, and in more ecologically valid settings.

Our findings are especially timely given the strong links that have been established between obesity and ADHD symptoms in the past decade (see Davis, 2010 for a review), and the recent evidence that stimulant treatment for obese individuals with ADHD symptomatology was highly successful in achieving sustainable weight loss (Levy et al.2009). However, 92% of Levy and colleagues' sample was female, so there was no opportunity to assess gender differences in the efficacy of the stimulant medication. Our results suggest a strong need to re-assess treatment outcomes for stimulant-induced weight loss in obese adults, taking account of possible gender differences. It is also important to acknowledge that the use of MP for weight loss – even in those with a diagnosis of ADHD – should be used cautiously because of some concern for its abuse potential.

Acknowledgements

This research was funded by a government grant from the Canadian Institute of Health Research (MOP-84257).

Statement of Interest

None.

References