Selfitis and behavioural addiction: A plea for terminological and conceptual rigour

In his response to our debate paper (Starcevic et al., 2018), Griffiths (2018) states that we have misrepresented a paper on ‘selfitis’ by Balakrishnan and Griffiths (2018) and then argues that we are obfuscating the concept of behavioural addiction. We disagree with both assertions and demonstrate that many of our differences in opinion are a consequence of terminological and conceptual ambiguity.

A rose by any other name is still a rose: is there a difference between disorder and condition, and between ‘selfitis’ and ‘selfie addiction’?

We fully agree with Griffiths (2018) that selfitis is neither a mental disorder nor a behavioural addiction and that it should not be equated with ‘selfie addiction’ (if selfitis and selfie addiction do exist). We only wish Balakrishnan and Griffiths (2018) were so clear and explicit in the first place. Instead, they refer to selfitis as a ‘condition’, as illustrated here:

Using these data, the SBS was validated and the selfie-taking behaviour was examined in relation to three intensity types (i.e., borderline, acute, and chronic conditions). (p. 732)

According to the English Oxford Living Dictionaries, one of the common meanings of condition is ‘an illness or other medical problem’. In the context of a paper describing an abnormal or unusual behaviour and linking it with aspects of psychopathology such as compulsion, the term condition when referring to that behaviour is more likely than not to denote a mental disorder.

Another link between a condition and disorder was provided by Balakrishnan and Griffiths (2018) when they first seemed to reject or at least be critical of the notion that selfitis meant ‘three levels of the disorder’ (p. 722), that is, borderline, acute and chronic. However, in the discussion of their findings on selfitis, Balakrishnan and Griffiths (2018) then referred to borderline, acute, and chronic conditions, as noted above. It seems that disorder and condition were used interchangeably, therefore suggesting that selfitis is a (mental) disorder.

Moreover, Balakrishnan and Griffiths (2018) implicitly acknowledge that selfitis is a disorder when they write in the final paragraph that ‘present research [on selfitis] is a novel addition to the research literature examining technology-related disorders’ (p. 733).

The lack of clarity about the relationship between selfitis and selfie addiction in the paper by Balakrishnan and Griffiths (2018) is first obvious in the following sentence:

As with internet addiction, the concepts of ‘selfitis’ and ‘selfie addiction’ started as a hoax, but recent research including the present paper has begun to empirically validate its existence. (p. 732)

The empirical validation that the authors mention here may apply to the ‘existence’ of both selfitis and selfie addiction. As selfitis and selfie addiction are referred to in a singular form (‘its existence’), the reader should not be admonished for the impression that there may be no difference between them.

Balakrishnan and Griffiths (2018) also write that ‘the qualitative focus group data from participants strongly implied the presence of “selfie addiction” although the SBS does not specifically assess selfie addiction’ (p. 732). It is again difficult to draw the line between selfitis and selfie addiction here because the authors suggest that study participants with selfitis also had selfie addiction, although the two concepts (or ‘conditions’?) may be assessed by different means. Thus, the sentence begs the question of what the difference between selfitis and selfie addiction is. If there is no difference, then selfitis is, ipso facto, a behavioural addiction.

Had Balakrishnan and Griffiths (2018) avoided a terminological and conceptual vagueness about selfitis and selfie addiction, the present debate might have been superfluous. This calls for a careful use of language and fostering of clarity and precision when attempting to identify and conceptualise emerging mental health problems.

How to set the bar when conceptualising behavioural addiction?

The same lack of semantic rigour and arbitrary approach are present in the domain of behavioural addiction. Griffiths (2018) seems to think that the debate about behavioural addiction is over and perhaps unnecessary because we already have a good definition of it – the one put forward by Griffiths (2005) himself (‘any addiction that does not involve the ingestion of a psychoactive substance’) – and a tool for conceptualising excessive and repetitive behaviour as behavioural addiction that is derived from the ‘components model’ promoted by Griffiths (2005). However, if the situation were so simple, there would be little disagreement and controversy about behavioural addictions. As Griffiths must be aware, the latter is not the case.

If the impact of a model is assessed quantitatively, the components model can be regarded as a ‘success story’ because it has been invoked to lend support to numerous behavioural addictions or ‘problematic behaviours’ related to behavioural addictions: dance addiction, fortune telling addiction, addiction to body image (muscle dysmorphia), problematic TV series watching, study addiction, work addiction, Facebook addiction, problematic Tinder use, tanning addiction and ontological addiction, among others. Griffiths (2018) argues that some of these are ‘subtypes’ of other behavioural addictions, but they remain ‘addictions’ nevertheless. These ‘addictions’ have a very limited clinical relevance and none are recognised as disorders by the official diagnostic and classification systems (i.e. Diagnostic and Statistical Manual of Mental Disorders and International Classification of Diseases) and outside the circle of supporters of the components model. Given that many of these concepts were endorsed by Griffiths himself, it is surprising that Griffiths (2018) states that there are ‘very few’ (p. 731) behavioural addictions.

We are concerned about the proliferation of behavioural addictions and the corresponding diagnostic inflation. Consequently, we believe that it is important to define behavioural addiction in a way that would minimise the risk of conceptualising as behavioural addiction activities that people engage in passionately and that have few or no negative consequences. The way to do it is to set the bar high, that is, use a high diagnostic threshold and clearly stipulate the exclusion criteria. The operational definition of behavioural addiction proposed by Kardefelt-Winther et al. (2017) is one such definition. An important feature of this definition is that it is not a ‘closed statement’ about its subject, but a collaborative work-in-progress that allows modifications as we gain new knowledge, improve our understanding and collect more evidence.

A simplified definition of behavioural addiction based on the proposal by Kardefelt-Winther et al. (2017) as a repeatedly performed behaviour that persists despite its adverse consequences and is associated with significant distress or functional impairment is broad and still requires exclusion criteria. Unlike Griffiths (2018), we believe that the latter are important, although we acknowledge that they may need to be refined. For example, the circumstances under which an underlying or co-occurring disorder obviates the diagnosis of behavioural addiction might include both of the following: (1) such a disorder (e.g. depression) was unequivocally present prior to the onset of problematic behaviour and (2) the current severity of the disorder is comparable to the severity of problematic behaviour or exceeds it.

In conclusion, our fundamental disagreement with Griffiths appears to stem from an attitude towards a trend to medicalise or pathologise excessive and repetitive behaviours, which inspired our previous paper (Starcevic et al., 2018). While we criticise this trend as misleading, Griffiths (2018) seems to both celebrate it and deny taking any part in it.