Cognitive load in midlife, the invisible work that empties capacity

A woman finishes a work call, opens a message about a school form due tomorrow, remembers mid-sentence that the dishwasher needs unloading before anyone can use a clean mug, and returns to her laptop to find she has lost the thread of what she was about to write. Nothing in that sequence looks like hard work from the outside. There was no lifting, no long hours, no obvious crisis. And yet by mid-afternoon her capacity for the day is measurably lower than it was at nine, and the research on why is more specific than "she is busy."

Cognitive load is a real, measurable cost, not a metaphor

The phrase "mental load" started as a way of naming something that felt obvious to the people carrying it and largely invisible to everyone else: the ongoing, background work of noticing what needs to happen, deciding when, and remembering it until it is done. A 2023 systematic review pulled together thirty-one peer-reviewed studies on this specific dimension of household and childcare work and found a consistent pattern across them. Women are more likely to be held responsible for, and to actually perform, the anticipating, planning and organising work of a household, distinct from the physical tasks it produces, and this cognitive dimension is associated with higher stress, lower life and relationship satisfaction, and measurable career disadvantage (Reich-Stiebert, Froehlich & Voltmer, 2023). The review's authors note this is unlikely to be fully explained by differences in paid work: research on the physical side of unpaid labour has found women do a larger share of it even when both partners earn similarly, and they suggest the same pattern likely extends to the planning and organising behind it.

A 2025 study from researchers at the University of Bath and the University of Melbourne, surveying 2,133 partnered heterosexual parents in the United States, put a number on that persistence. Mothers earning more than $100,000 a year reported thirty per cent less time on childcare and seventeen per cent less on housework than lower-earning mothers, but performed the same volume of mental labour regardless of income, averaging 13.72 mental tasks tracked in the study compared with 8.2 for fathers, a gap the researchers termed "gendered cognitive stickiness," describing how organisational tasks stick to mothers and resist redistribution in a way that outsourceable physical chores do not (Weeks et al., 2025). Even when fathers took on more physical and cognitive work, the mothers' overall load did not fall by a corresponding amount.

Interruption has a price that shows up as speed, not just as feeling

None of this load operates in isolation from the demands of paid work, and the two interact badly. A 2008 lab study by Gloria Mark and colleagues, using a simulated office email task with 48 participants, examined what happens when a task is interrupted mid-stream. The researchers found that people did not necessarily take longer overall to finish an interrupted task; instead, they compensated by working faster once they returned to it. That compensation carried a cost captured in participants' own workload ratings: those who had been interrupted reported significantly more stress, higher frustration, greater time pressure and more effort than those who worked without interruption (Mark, Gudith & Klocke, 2008). Read alongside the mental load research, the implication is straightforward. A day built from constant small interruptions, a message here, a remembered errand there, is not simply less pleasant than an uninterrupted one; it appears to demand measurably more psychological effort to produce the same output. The 2008 study did not examine household or caregiving interruptions specifically, since its participants were completing a simulated office task in a lab, but there is no obvious reason the same underlying mechanism, a task resumed at a faster, more effortful pace after a break, would behave differently when the interruption is a child needing something rather than a colleague at the door. Extending laboratory findings of this kind across settings they were not designed to cover is exactly the sort of gap the Observatory tries to be explicit about rather than paper over.

The planning and remembering work of a household falls unevenly

The imbalance in unpaid work more broadly provides the wider context for why this particular kind of load falls so unevenly. Analysis of English and Welsh data found that 57.7 per cent of unpaid carers were women, that women in paid employment provided unpaid care at a noticeably higher rate than men in the same position, and that the overall gender gap in unpaid care provision was widest of all in the fifty-to-sixty-four age band (Office for National Statistics, 2013). Unpaid care of this kind is rarely just physical; it carries its own planning and remembering burden, appointments to track, medication schedules, forms to complete, layered on top of whatever mental load a household already generates. None of the major national statistics on workplace stress currently separate out how much of the burden they measure originates inside paid work and how much arrives from outside it. Great Britain's occupational statistics record work-related stress, depression or anxiety as the largest single driver of work-related ill health, responsible for 22.1 million lost working days in 2024/25 (Health and Safety Executive, 2025), but that figure, like most of its kind, cannot distinguish a day lost to an unmanageable job from a day lost to a job made unmanageable by everything sitting alongside it.

Perimenopause adds a second, temporary load on top of the first

For women in their forties and fifties, a further and more direct layer sits on top of the mental load research. The Study of Women's Health Across the Nation followed 2,362 women for four years and found a specific, time-limited dip in cognitive performance during the transition itself: women in premenopause, early perimenopause and postmenopause improved their scores on repeated tests of processing speed and verbal memory, as most people do with practice, but women in late perimenopause did not show that same improvement, a pattern the researchers read as a genuine, if temporary, cognitive disadvantage concentrated in that specific window (Greendale et al., 2009). A separate clinic-based survey of 978 women found memory difficulties reported by 93 per cent of respondents and concentration problems by 91 per cent, the second and third most common complaints after fatigue itself (Reisel et al., 2024).

Read together, this means a woman in late perimenopause may be managing a documented, biologically grounded dip in processing speed and working memory at precisely the point in her life when the unpaid mental load literature shows the planning and remembering demands on her are least likely to have eased. Two separate bodies of research, one about hormones and one about household labour, are describing two halves of the same week without ever citing each other. Neither, on its own, tells a woman why a task that used to feel automatic now takes visible effort; only read side by side do they start to suggest an answer, and even then the Study of Women's Health Across the Nation is explicit that the dip it measured was transient, resolving by postmenopause rather than persisting indefinitely.

What the Observatory observes

Cognitive load of this kind is difficult to see because it does not announce itself the way a missed deadline or a raised voice does. It shows up instead as a slightly shorter temper, a lost thread of thought, a task that used to take twenty minutes now taking closer to forty. It also tends to be invisible to the person carrying it in the moment, precisely because noticing and naming a rising load takes the same limited attentional resource the load itself is consuming. By the time it is obvious enough to mention out loud, several hours or several days of it have usually already passed unremarked.

The Observatory's approach starts from regarding these small, easily dismissed signals as real data rather than as complaints to be managed away or evidence of some personal shortfall. Members are invited to record simple, everyday observations, what the day involved, how attention and energy moved through it, over weeks rather than in a single self-assessment, because a pattern read across a month tends to say more than any single difficult afternoon. A day that looks, from the outside, identical to the one before it, the same job, the same household, the same broad routine, can carry a very different cognitive load depending on what else converged on it: a sick child, a work deadline, a night of disturbed sleep, an unplanned conversation that needed managing. Observing several such days side by side, rather than judging any one of them in isolation, is often what reveals which factors are actually driving the load and which are merely present alongside it.

This is not a clinical assessment, and Obsera does not diagnose, treat or monitor any condition; it is a way of building a more honest, personal record of where a day's capacity is actually being spent, alongside small, optional practices a member can try and observe rather than commit to indefinitely. Contribution of that record to the Observatory's wider, anonymised, consented dataset is a separate and optional choice, and nothing about it is presented as a finding or a promised outcome. For a closer look at how overload tends to present before it becomes exhaustion, see what cognitive overload looks like in real life; for the Observatory's working definition of the word at the centre of all this, see what we mean by human capacity.

The dishwasher, the school form, the lost sentence: none of it looks like much on its own. Observed together, across enough days to see the shape of it, it starts to look like exactly what the research describes, a real and measurable cost, carried disproportionately, and rarely counted anywhere it might actually be seen.