The Science of Aging Well: How Your Beliefs Affect Your Health
Editor’s Note: In this fourth installment in The Council for Disability Awareness’ research mini-series, Fred Schott reviews the state of American adults’ mental health during the last quarter of 2020.
I’ve always been interested in the topic of mental health, but never has it been more interesting to follow than during the pandemic. The following chart comes from the Household Pulse Survey and looks at U.S. adults reporting that they are anxious or depressed.
The chart maps the trend of two key mental health metrics that the Census Bureau (based on input from the National Center for Health Statistics) is tracking:
- The percentage of respondents reporting symptoms of anxiety or depression in the last seven days. This is based on answers to a series of questions drawn from two survey instruments used by mental health clinicians to screen for symptoms of depressive disorder (PHQ-2) and anxiety disorder (GAD-2). You can check out the methodology on this page; just make sure to scroll down to the Technical Notes section.
- The percentage of respondents answering “Yes” to either one of these questions:
- At any time in the last 4 weeks, did you take prescription medication to help you with any emotions or with your concentration, behavior or mental health?
- At any time in the last 4 weeks, did you receive counseling or therapy from a mental health professional such as a psychiatrist, psychologist, psychiatric nurse, or clinical social worker? Include counseling or therapy online or by phone.
Taken together, these two metrics (I’ll refer to them as the “symptoms” and “treatment” metrics, respectively) provide a good snapshot of how the U.S. adult population is faring in the midst of the Covid-19 pandemic. And as the chart above indicates, so far in the fourth quarter of 2020, the metrics are trending upward, which means people’s mental and emotional wellbeing is taking a hit.
[Note: The Household Pulse Survey has been tracking the “treatment”– or “tx,” to use clinical shorthand– metric only since mid-August, but it’s been tracking the “symptoms”– or “sx”– metric since April. I wrote a piece over the summer about early findings regarding the “sx” metric; the link is here. It’s worth noting that 4Q 2020 levels of this metric are at or above earlier levels– and 2020 levels in general are way higher than pre-pandemic baseline, which according to NCHS was 11%.]
As compelling as the story told by the above chart is, it’s still flat and one-dimensional. Why? Because it’s an “all-in” national average. Embedded within the bottom-line numbers are several additional stories about how different sub-populations are faring during these challenging times. Adding these additional layers of detail give the overall picture more depth and accuracy.
I did a deeper dive on the “sx” metric last summer and again in the fall. One of the things that was clear in my earlier analysis was that the “sx” metric was higher– i.e., mental health was worse– for people who said they weren’t working than for people who said they still were. And that’s still the case with later data. But what I didn’t take into account in my earlier analysis was the composition of the “non-working” population.
When HPS respondents indicated they’re not working (or rather, more accurately, answer “no” to the question In the last 7 days, did you do ANY work for either pay or profit?), they receive a follow-up question asking them to select one of a baker’s dozen of reasons why. Here’s an overview of how people responded over the past three months (size of “non-working” adult population, based on Census Bureau estimates, has ranged from 102 to 106 million over the period):
[Note: While the first chart in this article presents data through November 23, 2020, I based the rest of this article on data through November 9, 2020. The reason for the difference: Every two weeks, the Census Bureau drops two sets of data: (a) high-level summaries of the most-recently completed survey wave from the previous week, and (b) the detailed micro data files for the wave preceding the most recent one. I base my deep-dive analyses on the micro data files.]
Three out of every eight people in the “non-working” group are retirees. And a look at the “sx” and “tx” metrics for this sub-group reveals a very different story from other “non-working” subgroups:
The “not working, not retired” subgroup – which includes people not working because they’re sick/disabled, providing care for someone else, or out of work because of layoff or business closure – has a much more unfavorable set of “sx” and “tx” metrics than retirees or those who are still working.
I work for an organization that’s focused on working adults, so most of my analysis of HPS data concentrates on the “working” group. That said, there is in fact a subset of the “not working” group that’s of particular interest to us: those people who have been out of work for an extended period of time because of a disabling illness or injury. And there’s a way we can carve out a HPS sub-population that’s a proxy for that category.
During August 2020, the Census Bureau began asking respondents a series of questions about their experience with social-insurance benefits such as Social Security (Retirement, Disability, Survivors), Supplemental Security Income, or Medicare. For each of these benefits– including Social Security Disability Insurance– respondents were asked (a) if they had applied or attempted to apply for it since March 13 and if not, (b) how likely they were to apply for it in the next twelve months. Based on responses with respect to SSDI– yes, they had applied or attempted to apply since March 23; or if not, they were extremely, very, or somewhat likely to apply for it in the next twelve months– we can spike out a group that roughly corresponds to people on long-term disability.
And here’s what our “by work status” charts look like if we do that:
I’m not one given to hyperbole, but let me indulge in a bit of it here. The difference between the “LTD proxy” group and all others in terms of the “sx” and “tx” metrics is mind-boggling.
Rather than offering up my own commentary on this finding, let me ask you, dear readers:
- Does this surprise you? Why or why not?
- What explanations do you have for this huge difference in metrics between the “LTD proxy” population and everyone else?
- What kinds of practices or products and services do you think can help this population (the long-term disabled from work) improve their mental/emotional wellbeing?
Related posts
