Response to Vanyukov: Why causality is a valid question for the gateway hypothesis

We are grateful for the comments by Vanyukov [1] on our article [2], which focus upon aspects of the ‘gateway hypothesis’ and methods for causal inference. However, we disagree with the conclusions reached in Vanyokov's letter and welcome the opportunity to clarify certain points in our original article. First, we maintain that the question around the gateway hypothesis is clearly a causal one. Vanyukov argues that we misstate the gateway hypothesis. However, we are not testing the specific definition used in the original formulation [3] but how it is commonly applied now; in other words, whether there is a causal pathway from use of one drug (or one form of a drug) to another (typically more harmful form). It is important to be clear about when we are testing a causal question, even in cases where the strength of our causal inferences may be limited by the nature of the data available [4]. In our paper we sought to examine whether this relationship (between smoking or alcohol use and the use of other illicit substances), and the temporal sequence of these, reflects causal pathways. Vanyukov seems to take the position that a causal pathway is likely to be mechanistically biological (e.g. via desensitization of nicotinic pathways). In our view there are many possible causal pathways, which may be biological but may also be social or behavioural. For example, one pathway could be increased social accessibility. Having social groups that use one substance might make it more likely for further substances to be offered/acceptable. Another potential pathway is increased tolerance of riskier behaviours. Using a substance perceived as ‘risky’, but not experiencing any adverse consequences, may make someone more likely to try another ‘riskier’ substance. Just because the order of substance use initiation could be opportunistic does not mean that causal pathways are not involved. Secondly, Vanyukov raises the important point that pleiotropy may be present in the genetic instruments used in our analyses. This is correct, but we consider this when we interpret and discuss our results. We also conducted sensitivity analyses that account for potential pleiotropy, and found that the direction of effect observed in these sensitivity analyses was generally in the same direction as the main analyses (although we also discuss some further limitations of these analyses). Furthermore, our interpretation of our results includes the possibility of an underlying shared risk factor, which Vanyukov claims that we ignore. We note this possibility throughout, and mention in our concluding paragraph that ‘while our findings support the gateway hypothesis to some extent, they also point to a potential underlying common risk factor and with better powered GWAS or those with more precise instruments and additional research we may be able to interrogate this further’. We reiterate here that there is a strong possibility of an underlying risk factor, and this needs further exploration. Ultimately, triangulation of evidence is needed to arrive at any robust causal inferences [5, 6], and our results, obtained using genetically informed causal inference methods, contribute to this. Future studies should consider other approaches that have different strengths and weaknesses and different potential sources of biases. Triangulation across multiple study designs will extend and refine our understanding of the gateway hypothesis, which is crucial given the clear public health importance of this question. None. This work was supported in part by Public Health England, the UK Medical Research Council Integrative Epidemiology Unit at the University of Bristol (Grant ref: MC_UU_00011/7), and the National Institute for Health Research (NIHR) Biomedical Research Centre at the University Hospitals Bristol National Health Service Foundation Trust and the University of Bristol. The views expressed in this publication are those of the authors and not necessarily those of the National Health Service, the National Institute for Health Research or the Department of Health. R.E.W. was supported by a postdoctoral fellowship from the South-Eastern Regional Health Authority (2020024). Zoe Reed: Data curation, formal analysis, investigation, methodology, resources, software, visualization. Robyn Wootton: Methodology, resources, visualization. Marcus Munafo: Conceptualization, funding acquisition, methodology, project administration; supervision.

Récupéré en direct depuis OpenAlex et désinversé. Les résumés ne sont pas conservés dans cette base de données : les index inversés représentent 8,6 Go des 9,3 Go de texte de la base, et le serveur dispose de 13 Go libres.