I suspect we may be in danger of leaping a little too readily to particular conclusions in respect of today's analysis of Scotland's suicide rate.
In advance, a couple of caveats. I have not yet read the full report prepared by a team at Manchester University: it is not available until later. However, I have studied the summary issued by the university.
This declares that "alcohol and drug misuse mean Scots are almost twice as likely to kill or take their own life compared with people in England and Wales".
These horrific figures deserve serious scrutiny, by politicians and public alike. I am not sure, however, that we would be entirely wise to follow the conclusion of today's report which is that support services should be refocused to prioritise action on dependence rather than the present spotlight upon mental health.
To be fair, the report suggests a number of other reforms to clinical care. The author, Professor Louis Appleby, set out his case sensibly and moderately on BBC Radio Scotland this morning.
But is there not, at root, a logical fallacy here? Are we perhaps at risk of confusing cause and coterminosity?
The researchers scrutinised cases of suicide and homicide. They discerned that, in a substantial proportion of cases, there was an antecedent history of alcohol and/or drug misuse.
From this, they surmise that the alcohol and drug dependence is the cause of the subsequent homicide or suicide.
Logically, however, might not the two factors point to another pre-existing cause? Might not the same individual, perhaps under stress from poverty or personal inadequacy, turn first to drink and then to violence or self-harm?
Alcohol and violence may indeed both be present in the same case. But can we truly assume that one causes the other?
My doubts increase when I consider the statistics in more depth. Scotland has a higher suicide rate than in England. But Scotland's rate has declined sharply.
How does that square with evidence that alcohol misuse is increasing - if alcohol is to named as the causal influence on suicide?
How can the causal problem increase - while the claimed result declines, albeit still remaining above the level in England?