Twice a week is the number most people settle on. A compendium of 146 controlled exercise trials in older adults suggests that for a long list of blood markers, twice a week is not enough.
The researchers, publishing in Ageing Research Reviews, pulled together every controlled training intervention they could find in older adults and asked a blunt question of each one: did this marker improve, or didn't it? Then they looked at what separated the programs that worked from the programs that didn't. The separator that kept showing up was frequency.
The threshold
At least three sessions per week were needed before the following moved in a favorable direction. Several of them are lab-report language, so here is what each one is and what pushes it around:
- Body mass. What the scale says. Diet drives most of it; exercise contributes, and this analysis found it needed three sessions a week to shift at all.
- Insulin and HOMA-IR. How hard the pancreas has to work to keep blood sugar in range. HOMA-IR is a simple calculation from fasting glucose and fasting insulin — the higher it climbs, the less your cells are listening to insulin. That is insulin resistance, and it is the road that leads to type 2 diabetes. It responds to weight loss, to cutting back on refined carbohydrates, and to muscles being used regularly, because working muscle pulls glucose out of the blood without needing much insulin at all.
- Triglycerides and total cholesterol. The fats circulating in your blood. Triglycerides in particular are the fast-moving one: they rise with surplus calories, sugar and refined starch, and with sitting still, and they come down again quickly when those change.
- IL-6, TNF-α and leptin. Signaling molecules, all three released in part by fat tissue itself. IL-6 and TNF-α are inflammatory messengers — useful in short bursts when you are fighting an infection, harmful when they stay mildly elevated for years, which is what tends to happen with excess body fat and with age. Leptin is the hormone fat cells use to tell the brain that energy stores are sufficient; when fat mass is high, leptin is high too, but the brain gradually stops responding to the signal.
- Adiponectin and IGF-1 — these went up, not down. Adiponectin is also made by fat cells, but it works the other way: it improves insulin sensitivity and has anti-inflammatory effects, and unusually, levels fall as fat mass rises. Getting it up is a good sign. IGF-1 is a growth factor tied to muscle maintenance and repair; it declines with age, and in older adults that decline is unwelcome. Worth knowing that IGF-1 is not a simple more-is-better marker in every context, but in aging muscle, more is what you want.
There is one documented way around the frequency threshold. Two sessions per week also worked — but only in programs running longer than 24 weeks. Less often, for longer. That trade is worth knowing about if three days a week is genuinely impossible, though it means roughly six months before the same adaptations show up.
Exercise improved every outcome the team examined except IL-1β — another inflammatory messenger, one of the earliest signals the immune system sends — which did not shift.
Aerobic came out ahead, but not everywhere
Aerobic training produced the largest overall benefits across the panel. Resistance training improved most markers too, with two specific misses: it did not reduce body mass, and it did not reduce IL-6. The one marker where resistance training beat aerobic was IGF-1.
That should not be read as a verdict against lifting. Body mass is a poor measure of what resistance training does — it builds tissue while removing fat, and the scale reports the net. This compendium simply wasn't designed to capture that. If the goal is specifically the metabolic and inflammatory panel above, the evidence here points to aerobic work as the higher-yield choice. Most people are better served doing both, which is also what the trials in this collection largely did.
The short-program result is a caution, not a shortcut
Programs shorter than 12 weeks produced larger increases in adiponectin and IGF-1, and stronger reductions in TNF-α and IL-1β, than longer ones. The authors read this as a transient response — an early adaptive spike that settles as the body adjusts to the new load.
It is an easy finding to misuse. A 10-week block showing dramatic bloodwork changes is not proof that a short program beats a long one; it is more likely a snapshot taken while the system is still reacting. Judge a program by what it holds at six months, not by what it spikes at eight weeks.
Who gained the most
Women improved more than men. People classified as unhealthy improved more than healthy participants. People with overweight or obesity improved more than those without. The pattern is consistent and unsurprising: the further a marker starts from where it should be, the more room there is to move it.
The anti-inflammatory effects were more pronounced when body mass and fat mass came down and glucose and lipid metabolism improved — but they were not dependent on those changes. Inflammatory markers improved in people whose weight did not.
What this analysis cannot tell you
The compendium's method is its own limitation. Classifying each study as "improved" or "not improved" and comparing the groups answers what distinguishes programs that work, but it does not produce a pooled effect size per marker. So this evidence supports "three sessions a week is the point where these markers respond" and does not support any claim about how much your triglycerides will fall.
The comparisons between exercise types, sexes and health states come from subgroup analyses within a body of literature, not from trials that randomized people head-to-head. And the population is older adults throughout. Whether the same frequency threshold applies at 35 is a reasonable guess, not a finding.
The practical version
Three sessions a week, mixing aerobic work with resistance training, sustained past the point where the early changes level off. If two sessions is the realistic ceiling, plan for at least six months before expecting the same effect. And if your starting point is poor — high weight, poor glucose control, elevated inflammatory markers — the evidence says you are the person with the most to gain, not the least.
Medical note. This article is for general education and is not personal medical advice. Talk to your own clinician before starting a new exercise program, particularly if you have a diagnosed heart, metabolic or joint condition.