Digitally acquired recreational computer programming activity and its association with bipolar disorder mood episodes Article Swipe

MAKE IMPACT

Adam Bayes · YOU? · · 2023 · Open Access · · DOI: https://doi.org/10.1111/bdi.13354

K.T. is a single, 43-year-old computer engineer who initially presented for psychiatric assessment because of depressive symptoms which had worsened over the previous year. He reported becoming more hopeless, amotivated, experienced difficulty getting out of bed in the mornings, with reduced appetite (but weight gain), initial insomnia, early morning wakening, poor concentration, difficulty making decisions, suicidal thoughts (without plans), anergia, diurnal mood variation and greater social isolation, which were consistent with a diagnosis of a melancholic depressive episode. There was a past psychiatric history of a depressive episode at age 23 associated with an unsuccessful attempt at carbon monoxide poisoning. Since that time he had been treated with venlafaxine 75 mg mane and had been 'getting along ok'. This had more recently been increased to 225 mg mane but had been ineffective. There was a history of social anxiety since a teen. He denied a past history of mania/hypomania, psychosis or deliberate self-harm. His medical history was unremarkable and he was otherwise in good health. There was concurrent alcohol abuse of up to six standard drinks 5 days/week which he says assisted with insomnia—he denied physical withdrawal symptoms. He was a smoker (five cigarettes/day). K.T.'s premorbid personality was somewhat 'obsessional'. Developmentally he reported his younger brother experienced a near-drowning as a toddler and remained severely disabled with a brain injury until his death at age 19. There was no family history of mental illness. In addition to working, K.T. was very dedicated to recreational computer programming in his free time, which he considered 'non-work' work. He had for many years used a program to automatically log his daily programming activity via an online platform. The day's activity was graphically represented on a public website with five increments from 'less' through to 'more' via a heatmap (Figure 1). This heatmap logs activity based on the number, n, of contributions from zero upwards. A grey square indicates '0' contributions and increasingly darker green squares indicate more contributions. The total annual number of contributions is also provided in Figure 1. Additionally, Figure 1 details mood state (depression, partially treated depression, mania/hypomania and euthymia) as well as formal employment status across the time period 2018–2022. In 2018, K.T. was predominantly depressed—he was treated with increasing doses of venlafaxine and with mirtazapine augmentation (and later with the addition of olanzapine) with only a partial response, associated with many periods of zero activity as shown on the heatmap (Figure 1, 2018). Early in 2019, K.T. was hospitalised and his antidepressant was switched to amitriptyline 100 mg daily, with mood improvement—coinciding with an increase in programming activity in March and April (Figure 1, 2019). As his mood declined in May (associated with little heatmap activity) lithium augmentation was trialled which improved productivity in June but the combination lost effectiveness and was associated with very little programming for the rest of the year despite escalating amitriptyline doses. Methylphenidate was added as an augmenting agent in January 2020 and there was a marked increase in heatmap activity in March (Figure 1, 2020). In April, it became evident that K.T. was hypomanic (with features of talkativeness, increased energy and confidence, distractibility and spontaneous laughing). Methylphenidate was ceased, amitriptyline was reduced and olanzapine was temporarily added. His heatmap activity continued to be elevated in May with increased programming activity at which time amitriptyline was withdrawn and in June hypomania resolved (Figure 1). It was unclear if this episode of mood elevation was indicative of an underlying bipolar disorder (BP) or whether it was the result of overstimulation secondary to the addition of the psychostimulant. K.T.'s mood soon worsened, and clomipramine was trialled owing to the presence of both depressive symptoms and marked obsessional features. The heatmap shows little activity associated with his mood worsening and in August, he was admitted to the hospital for electroconvulsive therapy (ECT). He was initially treated with right unilateral 0.5 ms pulse width (later increased to 1.0 ms) before converting to bitemporal electrode placement. He underwent a total of 25 treatments with ketamine eventually added as an anaesthetic to augment the ECT. His mood improved and he was discharged from the hospital. Soon after discharge, K.T. developed an elevated mood (he was accelerated, talkative, with reduced sleep and poor concentration). This was associated with auditory hallucinations (e.g. a child screaming, 'electrical buzzing') and functional hallucinations (e.g. 'when the shower is running, I can hear the neighbour's music playing very loudly'). He was again treated with olanzapine which settled the mania and perceptual disturbances. Given the clear manic psychotic episode, he was re-diagnosed as having a bipolar I disorder. He again became depressed and lamotrigine was added to lithium and clomipramine. Lurasidone was trialled but ceased owing to akathisia. Repetitive transcranial magnetic stimulation and intermittent theta burst stimulation were trialled but both were ineffective. Overall, the annual number of contributions was markedly higher at 1073, compared to the previous 2 years, where it was 281 and 321 in 2019 and 2018, respectively. Tranylcypromine (Parnate) was added in August 2021. From that time on, heatmap data (Figure 1, 2021) indicate greater levels of programming as well as consistency in activity into 2022 with the total number of contributions increasing to 2061. This was associated with 'inspiration coming back' and being able to work 4 h/day. By October 2021, he felt 'fully productive', working 8 h/day across multiple projects. In association with the introduction of tranylcypromine he was undergoing structured psychological therapy which involved a graded increase in work activity. K.T. remains euthymic on a combination of lithium, lamotrigine and tranylcipramine and is working productively across the week (Figure 1, 2022). BP impacts functioning and productivity with a systematic review of economic analyses of US data finding that about half (50.3%) of the non-medical costs of BP I related to unemployment with a further 6.4% due to productivity loss.1 This case gives a clear clinical example of bipolar mood episodes markedly impacting productivity related to recreational computer programming. The current case study examines recreational functioning in relation to mood episodes and highlights the potential usefulness of daily digital data acquisition of recreational or work activity in patients with BP. Productivity may be a proxy for affective state—being markedly reduced or absent during depressive periods, increased during manic/hypomanic periods and more stable and consistent during euthymia. A cross-sectional survey of BP I patients found that compared to age- and gender-matched healthy individuals, the former group reported lower levels of work productivity (as measured by the Endicott Workplace Productivity Scale), were more likely to miss work, have worked reduced hours due to mental health or medical issues, receive disability payments, or have been fired from their job.2 In another study, Simon et al.3 examined 441 outpatients with BP and found depression was strongly and consistently associated with decreased probability of employment and more days missed from work due to illness. However, they found symptoms of mania/hypomania demonstrated more variable effects on work productivity. Long-term clinical monitoring of BP is a useful method to track therapeutic stability. There is growing interest in digital data acquisition from technologies such as computers or smartphones including via 'apps' (i.e. applications)—with monitoring of domains such as sleep and activity levels that are relevant to BP (see systematic review by Ortiz et al.4). For example, Faurholt-Jepsen et al.5 sought to examine automatically generated objective data relating to patient phone calls and text messages (i.e. number and duration) as well as self-monitored data (e.g. mood, sleep duration, activity levels) collected using smartphones. The authors found correlations between this data and clinically rated depressive and manic symptoms which also differed between affective states. In the current case, the ineffectiveness of medications can be observed where productivity dips and a stable response to medications is indicated by greater consistency and stability of output (and overall the highest annual programming contributions) once the patient was commenced on the combination of tranylcypromine, lithium and lamotrigine. The current case used passive data acquisition which has advantages over active means, as it is likely more reliable as well as being more convenient for the individual. Data acquisition in this case was specific to the patient's recreational computer programming, though with technological advances, it is possible to envisage digital data acquisition which is personalised to the individual's specific work or recreational context. While the current case study was retrospective, there is potential to develop an approach whereby an individual with BP could be provided objective feedback relating to a change in their activity (either increasing or decreasing) and which may serve as an early warning signal of a switch into a manic/hypomanic or depressive episode. Implications of such monitoring in the future could include integrating a broader set of patient data into artificial intelligence systems which could then inform proactive clinical decision-making with respect to patient management. For example, the earlier initiation of antidepressants, adjustment of doses of anti-manic agents, or use of hypnotics. Future studies could examine recreational or work productivity combined with other real-time data to explore whether this information could be utilised as an electronic biomarker of bipolar illness activity. Thank you to the subject of this case report who kindly consented to this report being written and who also checked the accuracy of the manuscript. Open access publishing facilitated by University of New South Wales, as part of the Wiley - University of New South Wales agreement via the Council of Australian University Librarians. The author reports no conflicts of interest. The data that support the findings of this study are available from the corresponding author upon reasonable request.

Type

article

Language

en

Landing Page

https://doi.org/10.1111/bdi.13354

PDF

https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/bdi.13354

OA Status

hybrid

Cited By

1

References

5

Related Works

10

OpenAlex ID

https://openalex.org/W4380368803

Raw OpenAlex JSON

OpenAlex ID

https://openalex.org/W4380368803

Canonical identifier for this work in OpenAlex

DOI

https://doi.org/10.1111/bdi.13354

Digital Object Identifier

Title

Digitally acquired recreational computer programming activity and its association with bipolar disorder mood episodes

Work title

Type

article

OpenAlex work type

Language

en

Primary language

Publication year

2023

Year of publication

Publication date

2023-06-12

Full publication date if available

Authors

Adam Bayes

List of authors in order

Landing page

https://doi.org/10.1111/bdi.13354

Publisher landing page

PDF URL

https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/bdi.13354

Direct link to full text PDF

Open access

Yes

Whether a free full text is available

OA status

hybrid

Open access status per OpenAlex

OA URL

https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/bdi.13354

Direct OA link when available

Concepts

Mood, Bipolar disorder, Association (psychology), Psychology, Recreation, Clinical psychology, Psychiatry, Neuroscience, Psychotherapist, Biology, Ecology

Top concepts (fields/topics) attached by OpenAlex

Cited by

1

Total citation count in OpenAlex

Citations by year (recent)

2024: 1

Per-year citation counts (last 5 years)

References (count)

5

Number of works referenced by this work

Related works (count)

10

Other works algorithmically related by OpenAlex

Full payload