From the “Results” section:

Relative to those in nonremotable jobs, workers in remotable jobs spent approximately one additional hour alone per workday after the pandemic. Those in remotable jobs also differentially increased days spent entirely alone and decreased after-work socializing. The rise in isolation was sharpest for those living alone, whose likelihood of spending the whole day without social contact rose by 7 percentage points (83%).

Mental distress simultaneously increased: Scores on the Kessler (K-6) measure of generalized psychological distress rose by 0.1 standard deviations for those in remotable jobs relative to those in nonremotable jobs. The increase in distress was roughly twice as large for those living alone compared with those living with family. Alternative measures of mental distress—such as the frequency of depression, mental health care utilization, and antidepressant prescriptions—show similar trends. In contrast, workers in remotable jobs did not differentially increase visits to non–mental health care providers or non–mental health prescriptions (statins, for example), suggesting that the change was not merely driven by increased flexibility for doctor visits.

That is from a recently published paper by Natalia Emanuel, Emma Harrington, and Amanda Pallais.