1 Erratum

Upon publication of the original article [1], it was noticed that Figure 2 contained an error. The horizontal bars for the ‘likes’ row were incorrectly shown as blue. The horizontal bars for the ‘likes’ row should be orange. This has now been acknowledged and corrected in this erratum. The correct Figure 2 is shown below.

Figure 2
figure 1

Magnitude and direction of regression coefficients in All-data ( \(\pmb{N=24\text{,}713}\) ) and Pre-diagnosis ( \(\pmb{N=18\text{,}513}\) ) models. X-axis values represent the adjustment in odds of an observation belonging to the target class, per unit increase of each predictive variable. Odds were generated by exponentiating logistic regression log-odds coefficients.

In the section Method, subsection Improving data quality, the sentence ‘We also excluded participants with CES-D scores of 22 or higher.’ should read as ‘We also excluded participants with CES-D scores of 21 or lower’. This has now been acknowledged and corrected in this erratum.