The number worth sitting with this week, from the Cleo Family Health Index confirmed in March and pulled into a series of follow-up surveys this spring, is sixty-four percent.
That's the share of sandwich-generation working women — women raising children while also caring for an aging parent or in-law — assessed as at high burnout risk on a validated clinical instrument. The follow-up cut by age band is starker: among women 40 to 54, the cohort most likely to be deep in the work, 46 percent score in the most severe burnout category. About a quarter of family caregivers across the broader population put in caregiving hours equivalent to a full-time job, on top of the paid jobs and the raising of children that the survey defines them by.
These are not the numbers of a population that is, in the metaphor that's been doing the rhetorical work for thirty years, calmly sandwiched between two slices of obligation. They are the numbers of a population whose health, by every clinical measure we know how to ask about, is being quietly burned through.
The metaphor was always too tidy
Sandwich generation is a phrase from the early 1980s. It described, originally, a real demographic fact: a generation of adults whose parents lived long enough to need care and whose children stayed in the household long enough to still need parenting. The shape was symmetric. Both sides demanded similar things. The middle person held both. It made sense as a metaphor in a moment when family caregiving was still mostly a story about logistics — driving Dad to the doctor on Tuesday, picking up the kids from soccer on Wednesday, fitting it around a full-time job.
What the metaphor never quite captured was the difference in kind between the two slices. Raising a child is hard work that gets easier, in a curve that bends toward independence; caring for an aging parent is hard work that gets harder, in a curve that bends toward more medical visits, more medications, more difficult conversations, and eventually a more difficult death. The child becomes a teenager who borrows the car. The parent becomes a patient who can no longer take their own. The sandwich is not balanced. One slice is getting lighter while the other is getting heavier, and the person in the middle is doing more work over time, not less.
The other quiet failure of the metaphor is that it never asked which side of the sandwich the person was. In practice, in this country, roughly seventy percent of family caregivers are women, and the daughters in their forties and early fifties are the most likely single demographic to be doing the day-to-day caregiving for an aging parent. The "sandwich" is doing real work — but it's mostly being held by one half of the population, and the data is starting to say so explicitly.
What the burnout actually looks like
The Cleo data is one signal. A U.S. News survey earlier this year added the labor-economics angle: roughly 25.5 percent of family caregivers report putting in caregiving hours equivalent to a full-time job. The AARP / NAC Caregiving in the US 2025 report puts the median weekly time commitment at 27 hours and notes that the population of caregivers has risen by 45 percent over the past decade. The AARP Valuing the Invaluable 2026 update puts the annual economic value of family caregiving at $1.0 trillion — across 49.5 billion hours.
Underneath those numbers, in the qualitative interviews that the surveys build from, the same patterns keep surfacing. Sleep loss that doesn't recover on weekends because the phone might ring. The performance evaluation at work that quietly slips for the second year running. The marriage that becomes a logistics partnership without anyone deciding it should. The hobby that doesn't get picked back up. The friend whose calls don't get returned. The doctor's appointment for the caregiver herself that gets rescheduled four times and then forgotten.
What gets measured in surveys — work output, hours, dollar value — is the part of caregiving that has a clean unit. What gets less measured is the cognitive load: the running mental inventory of which medication ran out, which lab result is still pending, which sibling needs to be looped in about which decision, what the cardiologist said in March about the prednisone, whether the dishwasher needs to be unloaded before the home aide arrives at noon. That load doesn't go away when the caregiver clocks out of her paid job; it follows her into the evening, into the weekend, into bed.
The clinical burnout instruments capture this indirectly — through proxies like exhaustion, cynicism, and reduced efficacy — but they capture it. Sixty-four percent at high risk and forty-six percent at severe-risk in the most exposed cohort are not numbers about a tired demographic that needs a vacation. They are numbers about a structural mismatch between the work being asked of these women and the resources available to do it.
What would actually help
Three categories of intervention show up in the research literature, in roughly the order they're hard to deliver:
1. Make the work less invisible. Most of what exhausts a primary family caregiver is the cognitive load of being the only person in the family who knows the picture. A shared workspace that holds the medication list, the appointment schedule, the visit summaries, the documents, and the audit log — visible to every member of the family in the language they read — removes the brokering work that otherwise eats the caregiver's week. This is what software can do well, and it's a meaningful fraction of the burnout load.
2. Make the financial losses visible and recoverable. The West Virginia Caregiver Tax Credit Act, the Connecticut credit enacted in May, and the broader state-level credit movement tracked by ASPE are slow-moving but real. They don't fix the cognitive load, but they begin to compensate for the financial bleed that compounds it. The Senate Finance "Affordable Home Care Guarantee" push, if it ever clears committee, would shift the largest single financial cost — paid in-home care — onto a federal coverage pathway.
3. Take the work seriously at the workplace. The fastest-moving piece is in employer benefits — the Wellthy + Maven employer partnership, Cleo itself, the consolidation underway in the employer family-care market. Most caregivers don't know what their employer's caregiving benefits cover. The benefits exist, increasingly, because the actuarial math has finally caught up with what HR has been reporting for years: caregivers leave jobs, take FMLA, take unpaid leave, and lose promotions at rates that cost the employer real money. The employer-channel response is the corporate self-interest finally lining up with the obvious humane response.
None of these alone fix the 64 percent. Together, with time, they might begin to bend it.
What the data should change
The shift worth noticing is that the data has gotten better. Five years ago, the burnout numbers were estimates from advocacy organizations using small samples and informal instruments. Today they're from validated clinical instruments administered through major-employer benefits platforms reaching tens of thousands of caregivers — Cleo's Family Health Index is the most public example, but it is not the only one. The numbers were always real. They are just, finally, measured.
The reason that matters is that what gets measured gets argued for. The trillion-dollar AARP figure is the number we now use to push for federal legislation, employer benefits, and tax credits — because it gives a concrete reference point in policy conversations that used to default to "family caregiving is hard, here's a phone number." The 64 percent does the same kind of work for the burnout argument. It moves the conversation from anecdote to instrument.
For the daughters in their forties and fifties who are doing most of this work, that's a slow comfort. The data showing the load is real is not the same as the load being lighter. But the data is also what gets the load taken seriously — by the employer who decides whether to offer caregiver benefits, by the legislator who decides whether to advance the tax credit, by the spouse who finally believes that "I'm exhausted" is a clinical assessment and not a complaint. The number is one part of moving the work from invisible to visible. That is the part our category of software has any business contributing to.