Training load

The physical-activity input — the adaptive load that provokes muscle and the cardiovascular system to grow more capable.

Joint wear reduces it.

Joint pain reduces the exercise a person can sustain. This link is drawn from the reviews of this outcome rather than from a dedicated trial, and no effect size specific to this link has been reported.

Memory & focus increases it.

PMC4034407 — 10.3389/fnhum.2014.01044 — low - executive function governs adherence to any behavioral intervention, and exercise touches 13 outcomes here. doi 10.3389/fnhum.2014.01044 · PMC4034407

Joint comfort increases it.

10.1177/2325967119878136 — 10.1177/1759720X241273009 — This link rests on the reviews of this outcome as a whole. No effect size specific to this arm has been reported, so it is shown at reduced weight. (from the reviews of this outcome). doi 10.1177/2325967119878136

It increases Oxidative stress.

Exercise generates reactive oxygen/nitrogen species in working muscle, and these RONS are not merely damage — they are the signal that drives the adaptive response, including Nrf2 activation (PMC6617290; human skeletal-muscle Nrf2/Keap1 during exercise to exhaustion, PMC7358388). The proof is that blocking it blocks the benefit: exercise-induced ROS regulate AMPK, MAPK, NRF2 and PGC-1-alpha, driving increased glucose uptake, mitochondrial biogenesis and hypertrophy — and most of these adaptations ARE blunted by antioxidants, which the authors state reveals the crucial role ROS play during and after exercise. Sources are named: NADPH OXIDASES, xanthine oxidase and mitochondria. And a paradox that kills the simple reading: ROS generation also rises in skeletal muscle after prolonged disuse (immobilization). ROS is not a marker of activity — both loading and unloading raise it, and what differs is the context in which it is read. PMC6617290

It increases Supports BDNF signaling.

10.3390/biology12010126. doi 10.3390/biology12010126

It increases Insulin sensitivity.

Regular aerobic and resistance training improve insulin sensitivity and glycemia in type 2 diabetes — meta-analyses show 0.5-0.7% HbA1c reductions, with resistance training giving 10-15% improvements in insulin sensitivity. PMC8802999

It increases Bone health.

Resistance training raises bone mineral density (10-15% improvement range in older adults alongside strength and lean mass; ACSM consensus statement, PMC8802999); structured exercise preserves or improves BMD in postmenopausal women. The direction holds: pooled RCTs in premenopausal women show significant gains at femoral neck (7 effect sizes, 466 participants, g=0.342, 95% CI 0.132-0.553, P=0.001, NNT=5) and lumbar spine (g=0.201, CI 0.009-0.394, P=0.04, NNT=9) (PMC3563173). But the magnitude and the certainty ARE both low, and the sources say so themselves: a 2025 systematic review and meta-analysis of RCTs concludes physical exercise has shown low efficacy and moderate evidence for gaining total bone mineral density in the elderly (Muscles Ligaments Tendons J 2025;15(3)); and a 2025 meta-analysis of bone metabolism in postmenopausal women rated certainty of evidence low for CTX and osteocalcin and very low for all remaining outcomes under GRADE, downgraded for risk of bias and imprecision (doi 10.3389/fendo.2025.1597046). And it is POPULATION- and modality-specific: resistance, impact, whole-body vibration, endurance and combined training give different site-specific results, and premenopausal, postmenopausal and elderly cohorts do not behave the same. Effect sizes are small (g~0.2-0.34). doi 10.3389/fendo.2025.1597046 · PMC8802999

It reduces Blood pressure.

Physical-activity interventions lower blood pressure in randomized trials — systematic review/meta-analysis across 18 low- and middle-income countries (PMC11293006); mechanisms include improved insulin sensitivity and reduced vascular resistance. (1) Edwards 2023, Br J Sports Med 57(20):1317-1326, doi 10.1136/bjsports-2022-106503 — 270 randomized controlled trials, 15,827 participants. Resting BP fell with every modality (p<0.001): aerobic -4.49/-2.53 mmHg, dynamic resistance -4.55/-3.04, combined -6.04/-2.54, HIIT -4.08/-2.50, and isometric -8.24/-4.00 mmHg. (2) JAHA 2025, doi 10.1161/JAHA.125.044003 — 105 RCTs: combined -12.05 mmHg SBP, HIIT -10.97. (3) Schneider 2023, J Hypertens 41(7):1068-1076 — 37 studies, 1,942 participants: -6.4/-3.7 mmHg. (4) PMC12357484 — isometric -7.31 mmHg SBP at moderate GRADE certainty (DBP only low). (5) A 92-study review in older hypertensives confirming all modalities. The finding that matters here — two measurement methods give different verdicts on the same question: on resting blood pressure, isometric training ranks first by SUCRA and an expert review states it produces reductions greater than current exercise guidelines achieve and possibly comparable to standard antihypertensive monotherapy (Sports Med 2024;54:1459-1497). But on 24-hour ambulatory blood pressure, a 2026 network meta-analysis of 31 trials / 67 arms concludes evidence for dynamic and isometric resistance training remains uncertain, with only aerobic, combined and HIIT showing consistent 24-hour reductions. A cuff reading in a clinic and a 24-hour monitor do not agree about which exercise is best. GRADE certainty is moderate for systolic and low for diastolic. Exercise-versus-exercise superiority remains inconclusive. doi 10.1136/bjsports-2022-106503 · PMC11293006

It increases Mood.

Randomized trials pooled by systematic review/meta-analysis show physical exercise (aerobic, resistance and group formats) improves depressive symptoms in older adults (PMC11427357). Mechanistically consistent with the exercise→BDNF→mood route on the map. (1) BJSM 2026, PMID 41667154: 63 studies, 81 meta-analyses, 1,079 component studies, 79,551 participants — exercise reduced depression SMD -0.61 (95% CI -0.69 to -0.54) and anxiety SMD -0.47 (-0.59 to -0.36), aerobic exercise most substantial. (2) J Affect Disord 2024: 27 reviews, 190 experiments — SMD -0.67, number needed to treat 2.78. (3) JAACAP 2025: 21 reviews, 375 RCTs, >38,000 children and adolescents — SMD 0.45. Where the effect is largest, as stated: emerging adults 18-30 and postnatal women; greater reductions with group and supervised settings. For anxiety the pattern inverts — shorter duration and lower intensity worked best. Heterogeneity is substantial (I-squared 68-71 percent). PMID 41667154 · PMC11427357

It increases Memory & focus.

Meta-analyses of RCTs show exercise improves cognitive function in older adults (PMC9858649); in mild cognitive impairment, 26 RCTs (2,085 participants) gave a large effect on global cognition (SMD 0.81, 95% CI 0.58-1.05) (PMC12753406). Exercise->mood rests on SMD -0.61 across 79,551 participants. EXERCISE->cognition does not come close: an umbrella review re-estimating 24 meta-analyses (109 primary studies meeting criteria) found a small median effect size d = 0.17 (95% CI 0.13-0.19) with substantial heterogeneity (I-squared 43.4 percent) — and reported that all but two primary studies used underpowered designs to detect an effect that small. That re-estimation is a preprint (biorxiv 2022.02.15.480508), so it is not relied on as the verifying source; it is shown because a null-leaning finding should not be suppressed for being inconvenient. The peer-reviewed support is J Affect Disord 2024 (umbrella review of exercise on depressive symptoms and cognitive function). The honest statement: the direction holds; the magnitude is small and the primary literature is underpowered for it. PMC9858649

It increases Sleep.

Systematic review/meta-analysis of RCTs finds exercise improves sleep quality and insomnia severity in adults (PMC8215288), replicated on the Pittsburgh Sleep Quality Index in older adults (PMC12528139). Honest caveat: not universal — the same MCI meta-analysis that showed a large cognitive gain found NO significant sleep improvement (SMD 0.07, p=0.94, PMC12753406), so the effect is population-dependent. Corroborated in older adults by PMC7559727 (37 RCTs, 3,224 participants; Hedges g 0.25-0.71 across outcomes including sleep quality). Scope, and it is narrow: the strongest sleep evidence sits in clinical populations — patients with depression, cognitive impairment, breast-cancer survivors, older adults — not healthy sleepers. Effect sizes are small to moderate. Subjective sleep improves more readily than objective sleep does. PMC8215288

It increases Motivation.

Exercise raises brain dopamine: habitual physical activity of ~1-2 h/day increases brain dopamine levels (systematic review, DOI 10.1186/s12970-016-0155-6), and high-intensity interval training raises dopamine D2 levels and modulates dopamine signaling. Mechanistically, dopamine rises in the ventral striatum during exercise. doi 10.1186/s12970-016-0155-6

It increases Ketone fuel.

10.7554/eLife.15092 — mouse hippocampus, 30 days voluntary wheel running. doi 10.7554/eLife.15092

It increases Energy sensing.

Energy-sensor review 2018 (PMID 31314464) — muscle contraction during exercise raises AMP:ATP and activates AMPK, the hub coordinating exercise metabolic adaptation. PMID 31314464. Loss-of-function: SIRT1-mediated adaptations to exercise and caloric restriction are blunted in AMPK-deficient skeletal muscle, and the same work used mice overexpressing a kinase-dead alpha-2 AMPK subunit to isolate the requirement. Human arm: skeletal-muscle biopsies from young men after 15 sessions of one-legged knee-extensor endurance training over three weeks — the untrained leg serves as a within-subject control, which is a stronger design than between-group comparison. PMID 31314464

It increases Builds mitochondria.

Pilegaard 2003, J Physiol 546:851 — acute exercise transiently activates transcription of PGC-1α in human skeletal muscle, initiating mitochondrial biogenesis. PMID 12563009. Intensity selects the isoform, and by different receptors: a single bout of low-intensity exercise raises PGC-1-alpha-b and -c mRNA via the beta-2 adrenergic receptor, whereas raising PGC-1-alpha-a requires high-intensity exercise and is independent of beta-2 AR. PGC-1-alpha-b is the predominant exercise-responsive isoform and shows the largest rise. 'Exercise raises PGC-1-alpha' is therefore three different arrows depending on how hard you go. Nitric oxide is not the mediator: neither pharmacological nos inhibition in rats nor genetic deletion of eNOS or nNOS in mice prevents exercise-induced PGC-1-alpha expression — NO supports PGC-1-alpha and fiber-type shift generally, but is dispensable for the exercise response itself. PMID 12563009

It increases Growth hormone.

Exercise is one of the named physiological regulators of GH secretion; peak release occurs during slow-wave sleep and in response to exercise, fasting and physiological stress (Endotext NBK279056).

It reduces Calm under stress.

anxiety is the most heavily swept outcome here: four independent enumerations, roughly twenty searches, twenty-one sibling relationships worked. It did not have one of the best-evidenced anxiolytic interventions in existence. The rule that found it uses no words: exercise drives mood and sleep, both of which share drivers with anxiety, so exercise was implied by the edge list alone. Twenty-seven randomized controlled trials, 1447 participants, in people diagnosed with an anxiety disorder or identified as high-anxiety on validated scales: exercise demonstrated a significant anxiolytic effect, Cohen d = -0.42, p < 0.01, using a three-level model that accounts for effect-size dependency. And the prescription is specified: optimal outcomes with programs under 12 weeks, sessions of 45-60 minutes, more than 3 times per week, and total weekly duration of at least 180 minutes. In youth (27 RCTs, 2022 participants) the overall effect was SMD -0.36, and adherence decided it: high adherence to the exercise prescription gave -0.67 while low or uncertain adherence gave -0.13 and was not significant (p = 0.21). Dose is the whole difference between a real effect and nothing. Publication bias is present and the authors found it themselves: visual inspection showed funnel plot asymmetry and a multilevel Egger regression detected small-study effects (estimate -2.29, SE 1.02, z = -2.25, p = 0.024) — smaller trials reported larger effects. The true effect is very likely below d = -0.42. And comorbidity blunts it: comorbid chronic conditions attenuated efficacy to d = -0.29. doi 10.1016/j.ctim.2026.103328 · PMC12846938

Gainz (Muscle Protein Synthesis) →