Participants

Trial design

The study consisted of a randomised, parallel-group trial in a 2 (group: high ADHD traits vs low ADHD traits) x 2 (intervention: placebo or 5-HTP) x 2 (time: pre- and post-administration) design. Participants were stratified into two groups of high and low levels of ADHD traits using the ASRS, and assignment to either the acute, high-dose 5-HTP administration group or placebo group was randomised in a 1:1 allocation ratio. Both participants and trial investigators were blinded to the allocation. The trial primary outcome was a change in distractibility, measured by recording mean reaction time, standard deviation of reaction time, accuracy, and the percentage of false positives in two distractor tasks (See Fig 1). Analysis centred on observing differences between these values pre- and 90 minutes post-administration. Public involvement extended only to participation in the trial.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Study design (CONSORT) flow chart: Flow diagram outlining the study protocol, including participant recruitment, randomisation, allocation and analysis [58].

https://doi.org/10.1371/journal.pone.0349512.g001

The study protocol and analysis plan were preregistered on the Open Science Framework (OSF) (https://doi.org/10.17605/OSF.IO/H6KET) and later registered with ISRCTN (reference ISRCTN12905294). No changes were made to the preregistered protocol. De-identified participant data and experiment code can be accessed at the University of Sheffield Online Research Data Repository (https://doi.org/10.15131/shef.data.29183702). Reporting aligns with CONSORT 2025 guidelines [58].

Materials and measures

Task- relevant distractor: Flanker task.

Participants completed an adapted version of the Eriksen flanker test (sets of chevrons were presented rather than the letter sets presented in the original task) [61], considered a reliable test of response to task-relevant distractors. There is evidence that ADHD individuals experience increased reaction time and reduced accuracy on such tests compared to a typically developed population, and that ADHD symptom remission is linked to improved performance on the flanker test [62,63].

The flanker task was built and delivered using Psychopy 2023 2.3 [64]. Participants were presented with a series of 5 chevrons and asked to report the direction of the middle chevron using a key press (Z for left, M for right). Chevrons were displayed in either congruent (>>>>>, <<<<<) or incongruent (<<><<,>><>>) displays (see Fig 2). The chevrons were presented for 200 ms, and following participant response, a fixation cross was presented for 1000 ms prior to the next trial. Participants completed a 20-trial practice block prior to the test phase. The testing phase consisted of 10, 40-trial blocks, each presenting each combination of chevrons 10 times in a random order, resulting in 400 trials with 100 presentations of each condition. Participants were given a minimum 30 s break between each block, with the task taking around 15 mins to complete. Participants’ performance was assessed with several measures: mean reaction time, standard deviation of reaction time, and accuracy as a percentage for congruent and incongruent trials.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Stimuli used in the flanker and N-back tasks: A) the 4 stimuli, congruent and incongruent, displayed in the flanker task.

B) Schematic of stimulus presentation in the n-back task.W forms a cue, and a target response, and auditory stimuli are present on non-cue, non-target letters.

https://doi.org/10.1371/journal.pone.0349512.g002

Procedure

Following engagement with recruitment, individuals were required to complete the pre-screening survey, providing information on demographics and ADHD traits. Responses were reviewed, and eligible participants were contacted by email with a link to book a 2.5-h slot at the University of Sheffield for the trial. Participants were contacted 3 days prior to the session with further details and a reminder. Participants were asked at this time to complete a food diary for the 48 h preceding the test, logging food, alcohol and medication consumption. Participants were also asked to refrain from consuming alcohol, nicotine, or caffeine on the day of their appointment, and to refrain from eating for two hours before their appointment.

Upon arrival, participants were given an information sheet,asked to complete a consent form, and reminded that they may withdraw consent at any point during the study. Participants were seated in a quiet room with a monitor and a keyboard, with an investigator present. The investigator explained the tasks to participants and presented them in their previously assigned, randomised order. Baseline tasks took around 35 min to complete.

Following completion of the initial tasks, participants were given an envelope containing either 5-HTP or placebo and a glass of water and were observed swallowing the tablets. Participants were then shown to a separate room where they were allowed to participate in a self-chosen leisure activity whilst they waited. The waiting period was 90 ± 5 min, based on previous pharmacokinetic reports of 5-hydroxytryptophan levels in serum and cerebrospinal fluid [[70].

Once the 90 min wait period was completed, participants returned to the quiet room and completed the two tasks again in the exact manner as in the baseline test phase. Upon completion, participants were debriefed and provided with researcher contact details. The procedure took approximately 2 h and 40 min to complete per participant.

Data preparation and statistical analysis

Individual data was prepared by removal of reaction times outside of the range 200–3000 ms for the flanker task and reaction times under 200 ms for the N-back task, based on mean population values of simple human reaction time, and boundaries suggested by the NIH toolkit for flanker analysis [71,72]. Values that fell outside of ±3 standard deviations of the individual’s mean were also removed, following recommendations for reducing outliers in reaction time data [73]. All participants maintained at least 95% of data for each task. No participants had missing or incomplete data. Mean and standard deviation of reaction time values were calculated, as well as accuracy as a percentage of correct hits. The percentage of false positives was also calculated for the N-back task.

All analyses were completed using IBM SPSS Statistics (Version 29) and, unless otherwise specified, included all participants and were performed with a two-tailed significance value of 5%. Data were checked for normality using histograms, measures of skew and kurtosis and the Kolmogorov-Smirnov normality test. Significant outliers were noted in all measures of the flanker task, and in the standard deviations of reaction time and false positives in the N-back task. As such, these values were winsored prior to analysis. Due to the abnormality of the distribution, Box-Cox transformation was used to normalise all measures of the flanker task and the standard deviation of reaction time on silent trials and of false positives in the N-back task. This normalisation still resulted in a non-normal distribution of congruent accuracy in the flanker task, but given the robustness of ANOVAs, parametric testing was still used in the analysis [74]. Descriptives statistics, including means, standard deviations (SD) and percentages, were used to characterise the data.

Data analysis sought to answer three key questions: did the N-back and Flanker tasks manipulate distractibility, did the two tasks discriminate between individuals with high and low ADHD traits, and finally, did 5-HTP supplementation affect performance on the tasks. This was assessed by looking at differences in reaction time, individual variance of reaction time (standard deviation), and task accuracy. The N-back additionally looked at the rate of false positive hits as a metric of distractibility.

Test efficacy was assessed using paired t-tests between control (silent [N-back] or congruent [flanker] trials) and distractor (auditory stimulus [N-back] or incongruent [flanker] trials) condition measures at the pre-administration timepoint. To compare task performance between ASRS groups, a MANOVA was used to observe all of the variables of the flanker task and the variables of the N-back task, to look for any gross group differences. A 2x2 split-plot ANOVA was used to assess differences between the ASRS group and distractor/no distractor conditions, and independent t-tests to look at differences between groups on each measure. Finally, Intervention outcomes were first assessed with a 2x2 ANOVA (intervention x timepoint) to identify effects of the 5-HTP regardless of ASRS group, and then further assessed using a 2 x (Group: High ADHD traits, low ADHD traits) 2 x (Intervention: 5-HTP, placebo) 2 x (time: pre and 90 min post) ANOVA. ANCOVAs with age and baseline ASRS score were also performed. A final ANOVA was conducted with participants experiencing adverse events removed, to assess the impact these events had on the outcome measures.

Study population

In total, 374 participants completed the pre-screening survey, of which 176 were excluded for not meeting the eligibility criteria. The most common reason for exclusion was a non-eligible ASRS score (N = 83), although a considerable number of participants were excluded due to the use of serotonin-affecting medications (N = 37), smoking status (N = 25), dietary needs (N = 19) and a diagnosis of dyslexia (N = 12). A further 83 participants passively withdrew from the study prior to assignment to a trial group by not following up. 3 participants’ data were excluded from the analysis prior to the completion of data collection; one participant had a change to their medication status, one did not follow the fasting requirements, and one was excluded due to a computer error in the laboratory.

Demographic and clinical characteristics of each group are described in Table 1. The groups did not significantly differ in age, gender, or clinical diagnoses when looking between groups of the same ASRS status. There was a significant difference in ASRS scores in each measure between high and low ADHD trait groups, but additionally, there was a significant difference between high ADHD trait groups on the screener score and the hyperactive component – in both instances, the intervention group had a higher score (Table 2). Given the pronounced differences between low and high groups, this was not a concern.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of participants. No significant difference was observed between groups for age (F = 1.89, p = 0.132) or gender ratio (F = 1.28, p = 0.283).

https://doi.org/10.1371/journal.pone.0349512.t001

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Scores on the ASRS screener, hyperactive and inattentive subsections, and subsequent subtypes of the high ASRS group.

https://doi.org/10.1371/journal.pone.0349512.t002

There was an unequal distribution of ADHD traits, with 62.5% (N = 35) had scores suggesting combined-type traits, 33.9% (N = 19) with scores suggesting predominantly inattentive traits, and 1.8% (N = 1) with predominantly hyperactive traits. A further 1.8% (N = 1) had a screener score within inclusion criteria, but with the extended ASRS did not reach clinical scores in either domain.

Did the N-back and Flanker tasks manipulate distractibility?

Paired t-tests between control and distractor stimuli revealed no significant difference between the distractor and non-distractor conditions for the accuracy, reaction time or standard deviation of the reaction time in the N-back task at time point 1 (see S1 File). However, there was a significant difference between the number of false positives on distractor and non-distractor trials, although surprisingly, the silent trials (M = 42.72%, SD = 16.54%) had a significantly higher number of false positives than the auditory trials (M = 39.01%, SD = 15.84%), t (111) = −3.08, p = 0.003.

All measures of the flanker task differed significantly between the two conditions; incongruent trials (M = 478.66ms, SD = 68.28ms) had a longer reaction time than congruent trials (M = 425.12ms, SD = 57.51ms) (t(111) = 22.35, p < 0.001), and incongruent trials had a lower accuracy (M = 88.95%, SD = 7.84%) than congruent (M = 98.01%, SD = 3.81%) (t(111)= −13.845, p < 0.001). There was also a significant difference in the mean standard deviation of reaction times for incongruent and congruent trials, such that reaction times on incongruent trials (M = 109.08ms, SD = 53.4ms) varied more widely than reaction times on congruent trials (M = 92.41ms, SD = 41.06ms; t(111) = −13.845, p < 0.001). It was thus determined that the presentation of auditory stimulus in the N-back did not provide a significant distractor effect. However, the flanker task provided a significant distraction that impacted reaction time, accuracy and standard deviation of reaction time.

Did the two tasks discriminate between individuals with high and low levels of ADHD traits?

Effect of 5-hydroxytryptophan

Prior to analysing the effect of 5-HTP, analysis was completed to ensure that differences observed in standard deviations of reaction times and false positives between the high and low ADHD trait groups pre-administration in the N-back task remained when split down into the four intervention groups. ANOVA between the four groups highlighted a significant main effect of group on non-distractor standard deviation of the reaction time in the N-back (F(3,108) = 10.03, p < 0.001), and post-hoc analysis with Tukey correction identified significant differences between the high ADHD trait 5-HTP group and all other groups (Table 3). A significant main effect of group was also found on the standard deviation of the reaction time in auditory trials (F(3,108) = 2.837, p = 0.042), although the difference appeared to be only between the high and low 5-HTP groups (Table 3, Fig 3). Significance was also observed in the percentage of false positives in silent trials (F(3,108) = 3.675, p = 0.014) and the percentage of false positives in auditory trials (F(3,108) = 3.745, p = 0.013), but post-hoc analysis identified differences within the ADHD trait groups rather than between them (Table 3).

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Between-group difference results from Tukey post-hoc analysis of one-way ANOVA looking at effect of intervention group on N-back performance measures at time point 1.

https://doi.org/10.1371/journal.pone.0349512.t003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. N-back performance across ASRS groups, conditions and timepoints.

Comparison of mean scores in measures of standard deviation of reaction time across the N-back, split into intervention and ADHD trait groups at both pre- and post-administration (blue being low ADHD traits, pink being high ADHD traits. Darker shades are 5-HTP, lighter placebo). Error bars represent the standard error of the mean. A) Standard deviation of the reaction time in silent N-back trials. B) Standard deviation of the reaction time in auditory N-back trials.

https://doi.org/10.1371/journal.pone.0349512.g003

As ADHD trait group still had a significant effect on standard deviation of reaction time in N-back task measures when split into intervention groups, a 2 x 2 x 2 ANOVA between ADHD trait group, intervention and time point was used to assess if 5-HTP administration impacted the initial differences observed between groups. There was no 2-way interaction between 5-HTP administration and time point on the standard deviation of reaction times in the N-back task (see S4 Table). There was however, a difference between silent trial accuracy in the N-back between 5-HTP and placebo at pre- and post-administration measures F(1,110) = 6.693, p = 0.011, ηp2 = 0.057). Placebo saw an increase in mean from 53.25% to 60.82% of silent targets hit, whereas 5-hydroxytryptophan administration had an initial mean of 54.08%, which was reduced to 53.69% post-administration. No significant 3-way interactions were observed (see S5 Table).

To check that effects, or lack of, were not due to covariance, ANCOVAs were conducted. Age and baseline ASRS score did not provide any different results to the initial ANOVA when controlled for as covariates (see S6 and S7 Table).

Adverse events

Several participants (N = 11) reported adverse events during the experiment, including nausea (N = 2), vomiting (N = 4) and significant fatigue (N = 5). All adverse event reports were from participants on the 5-HTP intervention, and all participants who reported nausea or vomiting were female. Otherwise, there appeared to be no unifying characteristic linked to the experience of adverse events. Some participants (N = 4) also reported significant hunger, which they believed may impact their task performance; all but one of these participants were on the trial’s placebo arm.

To explore the mediation of these effects, a further ANOVA excluding participants who experienced adverse events was completed. Similarly to the initial three-way ANOVA, a difference between silent trial accuracy in the N-back was observed (F(1.93) = 5.636, p = 0.20, ηp2 = 0.057), whereby participants in the placebo group saw an improvement in accuracy (52.5% to 60.5%), and the 5-HTP group saw no change (51.1% to 52.2%). No other two or three-way interactions were observed.

Summary of findings

As initially established, the analysis of this study sought to answer three core questions: do the tasks manipulate distraction, do the tasks discriminate between high and low levels of ADHD traits, and does 5-HTP administration affect distractibility. The flanker task was found to manipulate distraction with significant differences between congruent and incongruent trials. The N-back with auditory stimulus did not provide a distractor effect.

When examining ADHD trait discrimination, the N-back task highlighted differences between high and low ADHD trait groups, in both overall task performance and in the standard deviation of reaction times. No difference in performance was found in the flanker tasks.

Evaluation of the impact of 5-HTP administration found one difference between intervention and placebo groups pre- and post-intervention, where placebo accuracy in the N-back increased between time points but decreased for 5-HTP. No other differences were noted because of 5-HTP administration, and there was no significant interaction related to ADHD trait groups. Analysis of the effect of adverse events on potential outcomes yielded no change to these findings.

Discussion

This study aimed to assess the effects of acute 5-HTP administration on the core ADHD symptom of distractibility in individuals with high levels of ADHD traits. We hypothesised that 5-HTP administration would reduce distractibility in individuals with high levels of ADHD traits, with a group of individuals with low levels of ADHD traits present to act as a control group. The data highlighted no significant improvements due to 5-HTP on any measure of distractibility in either group. Currently, no literature has examined the impact of 5-HTP administration in the management of ADHD traits; thus, the results, despite not indicating an effect of 5-HTP administration, are novel. Given that the tasks did not effectively discriminate between high and low ADHD traits, and a lack of distractor effect from the N-back task, it is difficult to draw conclusions about 5-HTP impact on ADHD traits.

Effectiveness of the flanker task

In the flanker task, congruency had a significant effect on all measures, indicating that the distractor condition was indeed distracting, as has been widely observed previously [75,76]. As such, the flanker task provided results similar to those observed in previous work, whereby incongruent trials provide a distraction that impacts task performance. However, the flanker failed to highlight a difference between groups with high and low levels of ADHD traits.

A previous systematic review of flanker performance found that individuals with ADHD generally had slower reaction times and greater error rates, but, as with much work in ADHD, it focused on a child population [63]. Given that attentional networks mature and develop over time, and that performance in tasks such as the flanker develops and differs between childhood and adulthood, it is perhaps unsurprising that the same results are not mirrored in adult populations [77]. There are fewer instances of the flanker task in ADHD adults, but results are conflicting; one study reports individuals with ADHD as having a longer reaction time and a higher level of accuracy, another found no difference in reaction time and a lower level of accuracy, and another found differences in reaction time and standard deviation of reaction time but no differences in accuracy [78–80]. It is also possible that a lack of observed difference is due to the non-clinical sample used within this study.

It is also of note that there were several methodological differences between the flanker tasks used; Merkt et al. (2013) [78] used a paradigm where participants reported if a number was odd or even, and the distractors were either odd or even numbers, McLoughlin et al (2009) [80] presented two flanker arrows 100 ms before the initial paradigm, and Marquardt et al (2018) [79] presented the flanker for 100 ms, rather than the 200 ms used in this protocol. All these features may make the task more challenging and may have better highlighted differences between groups; the current study had similar accuracy and congruent reaction times to those seen in Marquardt’s (2018) [79] work, but a considerably lower incongruent reaction time, and considerably lower reaction times and higher accuracies than those observed in Merkt et al., (2013) [78]. As such, it is possible that results from the current study skew lower due to the relative ease of the task.

Effectiveness of the N-back task

With regards to the N-back task, the presence of the auditory stimulus did not create a distractor effect. The task did, however, show some discrimination between high and low ADHD traits groups, whereby individuals with higher levels of ADHD traits had a greater standard deviation of reaction time in the silent condition, and saw a decrease in standard deviation in the presence of an auditory stimulus. The low ADHD traits group instead saw an increase in the standard deviation of reaction time with an auditory stimulus present.

There are several possible reasons as to why the N-back task failed to distract. Previous work with unrelated auditory stimuli in a visual task has demonstrated a distraction effect, but there has been a focus on the novelty of auditory stimuli presented in comparison to the presentation of a standardised sound, whereas the current study used only a standardised beep compared to silence [67,68]. However, other work using novel auditory stimuli has also observed alerting or orienting effects, whereby performance improved with the presence of novel auditory stimuli [81,82]. There are several theories for such variance in performance; firstly, that working memory load may modulate response to novel sounds, where lower working memory load results in greater distraction due to sound, thus the cognitive load of the N-back task may have contributed to a lack of distraction effect [81]. Alternatively, the informational value of the sound, i.e., whether it predicts a target response, may also modulate whether a sound is distracting or not in experimental conditions [83]. With regards to the current study, it is possible that the repeated use of the same auditory stimulus reduced its effect as a distractor, although given that there appears to be no unifying trend in performance changes over time, this may not be the case. Furthermore, the volume of the sound may have played a role in the N-back’s failure; distraction by auditory stimuli has been considered to have a similar effect when at different volumes within the range of typical human speech, but there is some suggestion that at loud levels (above 80dB), the sound may present an alerting effect [84]. Again, evidence here is conflicting, and some work has found increased reaction time with increased volume [85]. As such, it is unclear as to why the N-back task failed to distract, but differences observed in the N-back will instead be considered primarily as differences in working memory.

ASRS scores and task performances

The one difference observed between ADHD trait groups was the standard deviation of reaction time in the N-back task, whereby individuals with high levels of ADHD traits scores had a higher standard deviation than those with a low score. Response variability has been observed to differ between ADHD and neurotypical groups in several reports [86,87]. Such differences are postulated to be because of occasional lapses of attention and may often result in a larger magnitude difference than other neuropsychological markers, such as reaction time or accuracy in tasks [88]. Such changes have also been paired with the assessment of stimulant medication, whereby the administration of stimulants reduces response variability [87,88].

Efficacy of 5-hydroxytryptophan

5-HTP administration only resulted in one effect, whereby individuals on placebo saw an improvement in the accuracy on the silent portion of the N-back task, and individuals taking 5-HTP saw no real change to their performance. This lack of other changes could indicate that 5-HTP did not reduce distractibility or affect working memory, but further considerations should also be made. Aside from the significant caveats that the flanker did not provide discrimination between the ADHD trait groups and the N-back failed to work as a distractor, the absence of changes between pre- and post-administration could also be related to the dose of 5-HTP used. A 200 mg dose was used based on previous research where a single 200 mg administration resulted in changes in behaviour on moral judgement tasks [60]. Across the literature, dosage of 5-HTP varies greatly, with some individuals using 5-HTP for myoclonus taking single doses of up to 725 mg, but within the literature for depression, doses from 50–300 mg are standard and result in symptom improvement [70,89]. With regards to 5-HTP and attention, preclinical work in macaques used a dose of 20 mg/kg, significantly higher than the dose used here [51].

Alongside dose, dose frequency should also be considered; it is well accepted that other serotonergic medications, such as selective serotonin reuptake inhibitors, require long periods of use before providing symptomatic relief, potentially because of the inhibitory activity of the 5-HT_1A receptor, which reduces the activity of serotonergic neurons in the initial increase of serotonin [90]. Such a possibility should also be considered with 5-HTP, where a longer administration may be needed to provide benefits. Furthermore, the dose given does not directly reflect central bioavailability; around 95% of serotonin synthesis occurs in the gut, and it is likely that a significant portion of the 5-HTP was indeed metabolised in the periphery [91]. It is possible that a higher dosage than 200 mg is needed to observe changes in performance on cognitive tasks.

Although direct measures of central serotonin were not used, the profile of adverse effects reported provides some indication that central serotonin likely increased. Firstly, the reporting of fatigue in several participants is promising; 5-HTP has been widely observed to improve sleep, and is believed to occur via synthesis of melatonin, one of the end metabolites in serotonin synthesis [54,92]. Additionally, presentation of nausea and vomiting also indicate that the dosage was significant enough to induce adverse events; however, some of this may be linked to activation of 5-HT₃ receptors in the gastrointestinal tract, which would be an indicator of raised peripheral serotonin rather than central changes [91]. From analysis, it appears that the adverse effects did not contribute to the null findings of this study, but as adverse events were dependent on self-report, it is possible that not all participants who experienced fatigue, which could affect reaction times and accuracy, reported it. As ADHD is a biologically and clinically heterogeneous disorder, it is also possible that not every individual with high levels of ADHD traits presents with the same reduction in serotonin, nor the same genetic profile, which would result in reduced synthesis. For example, TPH2 variants, which have been linked to reduced 5-HTP levels, have been linked to ADHD in some cohorts, but not in others [20,34,93]. As such, future work should consider looking at baseline levels of serotonin and its subsequent metabolites, as well as genotyping, to better understand variance within the sample and characterise responses.

One other possible reason for the lack of differences between groups could be the use of a pre-clinical population. ASRS screening has been reported to have a high validity, with a classification accuracy of over 90%; thus, it was deemed appropriate for determining ASRS groups [57]. Comparison of ASRS scores between individuals who reported an ADHD diagnosis and others in the high ADHD traits group revealed a significant difference in hyperactivity, with ADHD diagnosed individuals having a higher mean score (6.625 compared to 4.915), but no differences were found in other ASRS scores (see S8 Table). As such, it appears that symptom severity, as noted by the ASRS screener, is not directly linked to diagnosis within the study cohort. As diagnosis was self-reported and not clinician-assessed, there is also room for error within this measurement. Future work should consider the use of clinical evaluation in determining the ADHD cohort, alongside potentially including collateral reports of trait presentation, as would be done in diagnosis.

Study strengths

The current study, to our knowledge, is the first to assess the impact of 5-HTP supplementation on ADHD traits and has considerable strengths. Firstly, the study employed a randomised, controlled protocol with blinding, widely considered the gold standard for efficacy research [94]. The trial administrator blinding was maintained for 105 participants (93.75%); the last 7 participants were tested without trial administrator blinding as data needed to be prepared for presentation. Furthermore, the study population was overwhelmingly female (79% of the total study population), a group that has been historically underrepresented in ADHD diagnosis, treatment and research, addressing calls for greater female inclusion in ADHD research [95]. The study used well-validated measures of ADHD traits and tasks that had previously established differences between ADHD and typical cohorts [57]. Lastly, the dose of 5-HTP was selected based on evidence from previous research, and medication-naïve individuals were used to best isolate the effects of the intervention [60]. As a preliminary study, there is significant scope to extend and better understand the role and action of 5-HTP supplementation in ADHD, particularly in using tasks which are more sensitive to ADHD traits. One significant caveat with the study lies in the efficacy of the distractor tasks.

Limitations

Although methodologically strong, the study has limitations. Firstly, although sample size recruitment was met and is the largest (and only) study assessing 5-HTP impact, the sample size was determined on power calculations for a medium effect size. As such, smaller effects may not have been possible to elucidate in the current sample. The sample also used a pre-clinical group. This could be considered a benefit, given it is well accepted that ADHD is underdiagnosed in the adult population, particularly in women, meaning the sample includes those typically under-represented in the current literature [59,95,96]. However, the ASRS does not equate to the depth of a clinical assessment. Regarding the intervention, there are several further limitations. Due to the small size of the study team, it was not possible to use central randomisation or outsource statistical analysis to individuals not involved in data collection, which could have resulted in less than optimal randomisation and blinding. There was also not a fully pre-defined statistical analysis plan, due to the anticipation of novel findings which may require exploratory analysis.

The most considerable limitation of the study is the efficacy of the cognitive tasks. The flanker task served as a distractor but did not discriminate between high and low levels of ADHD traits, so no conclusions about 5-HTP efficacy could be drawn from this. The N-back task showed the converse: no distractor effect was observed, but there was some discrimination between high and low ADHD traits, suggesting that conclusions about distractibility cannot be drawn from N-back results. Future work should consider tasks more sensitive to both ADHD traits and distractibility.

The current study did not employ direct measures of 5-HTP, serotonin, or any metabolites, meaning it is unclear as to whether serotonin was indeed increased from administration. Use of a fixed dose and acute administration also limits generalisability, as effects may vary in a dose-dependent manner. Crucially, it is well established that existing serotonergic therapies, such as Selective Serotonin Reuptake Inhibitors, require chronic administration to see clinical effects, even if serotonin itself increases rapidly. As such, a null result may be due to an acute intervention, and longer administration periods should be assessed in future work. Finally, when assessing the impact of supplementation, the current study focused on distractibility, which means that current findings cannot be generalised to other ADHD trait domains of hyperactivity and impulsivity.

Conclusions

In conclusion, this is the first randomised, controlled trial looking at the effect of 5-HTP administration on ADHD traits. This preliminary study assessed the efficacy of acute dosage on distractibility in individuals with high levels of ADHD traits, with a low-level trait group as a control. The study found no significant effect of 5-HTP on performance in measures of distractibility or working memory, although this could be linked to the tasks not being sufficiently sensitive to ADHD traits, and the N-back task failing to work as a distractor. Further work is needed to establish the effect of 5-HTP on the other ADHD trait domains of hyperactivity and impulsivity, alongside observation of a range of doses to better elucidate the relationship between 5-HTP supplementation, changes to serotonin, and ADHD traits.