Participants who used smartphones on the toilet were younger than non-users (mean ages 55.4 vs. 62.1, p = 0.001). Of all respondents, 66% used smartphones while on the toilet.
Participants who used smartphones on the toilet spent significantly more time there than those who did not, with 37.3% of smartphone users spending more than five minutes per visit on the toilet, compared to 7.1% of non-smartphone users (p = 0.006).
Furthermore, in a multivariate logistic regression, smartphone use on the toilet was associated with a 46% increased risk of hemorrhoids (p = 0.044) after adjusting for age, sex, BMI, exercise activity, straining and fiber intake.
The most common activity performed while on the toilet was reading “news” (54.3%), followed by “social media” (44.4%).
The study suggests that prolonged engagement with smartphones while using the toilet may be associated with an increased prevalence of hemorrhoids.
Is this an example of “correlation doesn’t mean causation”?
I mean, those p values would suggest it is statistically significant, after adjusting for other factors. Not really a correlation thing as I don’t think they’re literally suggesting using your phone is what gives you them but more a third variable (time spent on the toilet) problem.
No, not with proper statistical controls. A logistic regression can account for all those other variables at the same time, plus some unaccounted variables that might be heavily associated with those measured (like when I control for upbringing by asking about a parental education).
There’s always a little noise, of course, but that’s why we look at effect sizes. There’s clearly a relationship, even if it’s plus or minus a certain amount from what’s reported.
As someone who would read shampoo bottles to pass time on the toilet, I will say that some of us were doomed to long poops from the get go. Sigh.
Tip, put a timer that beeps every 3 minutes, it helps
