Physiological reserve

Why can an older person be entirely fine day to day and yet be felled by a minor infection?

Two people catch the same urinary infection. One is thirty and mildly inconvenienced. One is eighty-five, was doing her own shopping last week with unremarkable blood pressure and blood tests, and within two days is confused, immobile and in hospital.

The tempting explanation is that the older person was already unwell in some undetected way. But the resting measurements really were normal, and they were normal because at rest she was genuinely fine. So the explanation must account for something stranger: how a person can be simultaneously well and precarious, with no resting measurement able to tell the two apart.

Start with a design observation. Organs are not built for the demand of a quiet afternoon. They are built with peak capacity enormously in excess of resting requirement, and the ratio between the two is the interesting quantity.

Some rough figures, flagged as recalled orders of magnitude rather than precise values. Cardiac output at rest sits around five litres a minute; a fit young adult's maximum is roughly four to five times that. Resting ventilation is about six litres a minute, while maximum voluntary ventilation runs into the low hundreds — more than a tenfold margin. And the kidneys are redundant enough that a healthy person can donate one and live normally, so that margin is at least twofold.

Call the ratio of peak capacity to resting requirement the reserve multiple. Everyday life samples only the bottom of that range. Walking to the kitchen asks the heart for something close to its resting output — a test the body passes by an enormous margin, and whose result carries almost no information about the margin itself.

Now age the system. Peak capacities decline with adult age; resting requirements largely do not. Suppose peak output falls at about one per cent a year from age thirty — a crude but not unreasonable stand-in for several organ systems. Starting from a fivefold reserve, at eighty the multiple is 5 x 0.99 to the fiftieth power, which is 5 x 0.605, or about three.

Look carefully at that result, because it is the answer to the question. A threefold margin is still a large one. The eighty-year-old with a reserve multiple of three is not ill, is not compensating, and will produce entirely normal resting measurements — three times the required output is comfortably more than the required output. Half a lifetime of decline has been invisible, and invisible for a good reason rather than through any failure of measurement.

So what does the infection do? It does not attack the ceiling. It raises the floor. Fever lifts metabolic rate and oxygen demand; faster breathing raises the work of breathing; vasodilation means a higher cardiac output is needed merely to hold blood pressure where it was; and the immune response itself is expensive. The demand line rises toward a ceiling that has been quietly descending for fifty years.

That is why the same insult produces opposite outcomes. If illness demands three times resting output, the thirty-year-old with a fivefold ceiling has margin left over. The eighty-year-old with a threefold ceiling has none — and the tipping point is reached not by a bigger insult but by an identical one meeting a smaller gap. Outcome is flat in insult size everywhere below the ceiling and steeply nonlinear at it.

There is a second element, and it is where the word "reserve" gets subtle. Organ reserves are not independent. A kidney working near its ceiling clears drugs slowly, so a drug accumulates and reaches the brain; delirium follows; the delirious person stops moving; immobility costs muscle within days; weak muscle means falls. Each system might have survived alone. They do not fail alone, because they share a bloodstream, a drug load, and a body that either moves or does not.

How the account could be wrong. The claim is that reserve, not resting state, determines what an insult does — and it is testable, because reserve can only be seen by demanding it. If it is right, two people with indistinguishable resting physiology but different exercise capacity should fare differently after a comparable stress. That is what is found: measures making a demand, such as gait speed, grip strength, or exercise testing, predict outcomes after surgery and acute illness better than resting vital signs and better than chronological age. I name the direction rather than an effect size, since those vary by population and instrument. The refuting observation is clean: if capacity measures added nothing once resting measurements and age were known, the construct would be doing no work.

An honest caveat. Reserve is inferred, not measured; no instrument reads out how much margin a person has. Frailty indices are proxies built on different principles — a phenotype of weakness, slowness, exhaustion and weight loss on one hand, a count of accumulated deficits on the other — and they disagree about who is frail. The mechanism is well supported; the measurement is not settled.

Two refinements connect the pieces.

First, this is redundancy, but not the engineered kind. An engineered redundant system carries identical spare channels, switched in when the primary fails. Physiological reserve is headroom within a single channel — one heart with more output available, not two hearts. And the channels are coupled: they share circulation, drug load, and the consequences of immobility, so the failure mode is common-mode — exactly the mode duplicated channels are famously bad at surviving. Spare capacity and spare parts behave very differently under a shock that touches everything at once.

Second, a misconception worth naming. "She was fine until the infection" is usually said as though something was missed. Nothing was. She was fine, in the only sense resting observation can establish, and the reserve shrinking for decades was never visible in those observations and could not have been. The information was not hidden by inattention; it sat in a variable nobody was sampling, because sampling it means making a demand. None of which says what should be done for any particular person — what weight to give a slow gait belongs to a clinician who can examine them.