SYNTH-107 · Guide
Choosing gym exercises, training at home with kettlebells or dumbbells, and circuits
Build the session on a few heavy multi-joint lifts and add an extra only for a named reason; with light home weights go single-leg, add reps and end the last set nearer failure; a heavy circuit saves time with no detected strength cost, a light one is a fallback, and neither replaces aerobic sessions
The question it answers
Which exercise ("keep the Pallof press?", "a lying curl for a standing single-leg curl?", "10-12 reps for upper-body endurance on the bars?"), what to do with only kettlebells or dumbbells and a pull-up bar, and what a circuit is for ("why a circuit?", "can one with run bouts replace aerobic volume?"). Sets and sessions are SYNTH-105 and SYNTH-43 / SYNTH-44; the next number is SYNTH-101; HYROX stations are SYNTH-71.
Default (what to prescribe first)
Three or four heavy lower-body lifts carry the session (squat or leg press, hinge, calf raise, and a single-leg lift for runners and burst-event riders), short of failure, at SYNTH-43 / SYNTH-44's dose, or SYNTH-105's first weeks for a beginner; heavier loads raise strength (REF-1476, A, mostly inexperienced trainees).In a model of 295 studies, mostly of untrained or recreational lifters, the number of exercises did not improve prediction of gains, while each extra set per session added a little (REF-1478, B+), so a further exercise needs a named reason (inference).
Up to two extras, each with its reason (practice), chosen from a hamstring exercise beside the hinge, a short trunk block, and an upper-body push and pull. For a runner who sprints the hamstring slot is a Nordic: programmes with it about halved hamstring injuries in athletes of a range of sports (REF-766, A; a transfer to runners).
Trunk: one dynamic or loaded exercise plus one hold, 10-15 min at most (practice). In competitive athletes trunk training lifted local muscular endurance most (SMD 1.29), sport performance moderately (0.64), maximal strength and leg power little (0.39, 0.30) (REF-1632, A-, ages 11-37, no endurance-sport estimate); in adult novice runners a supervised hip-and-core programme cut lower-limb injuries (REF-1384, A, bands and body weight).
Upper body, bar-holding endurance included: 2-3 sets of 8-12 reps, a couple short of failure (SYNTH-105's starting range; practice for endurance).ACSM's 2026 stand set no muscular-endurance prescription (REF-1476, A); reps against a fixed load rise far more reliably after lifting than reps at a share of current maximum (REF-1509, B+); over 22 weeks 10RM sets built more strength than 30RM sets, whose nearly doubled volume load gave no clear edge in fixed-load reps (REF-1510, B+, cadets, sets to failure).
At home with light weights, same movements made harder (practice): single-leg variants (split squat, step-up, single-leg Romanian deadlift), a lowering of about 3 s, reps to the top of a range, the last set 1-2 reps from failure, then a harder variation or the next bell (SYNTH-101).Light loads taken to failure grew muscle like heavy ones while 1RM favoured heavy (REF-677, B+); adding reps is the route its trial's authors name when heavier loads are missing (REF-1479, B+); home, band and circuit training each raised strength against none, in evidence mostly from inexperienced trainees (REF-1476, A).Holds add to the lifting, not replace it (SYNTH-88).
Decision
| Situation | Choice | Why |
|---|---|---|
| Pallof press; best for obliques | Keep it if done cleanly; pair it with a trunk exercise of another kind (side plank, a loaded free-weight move); none is shown best (practice) | No trial in the base compares named trunk exercises on a training outcome; free-weight exercises drew the most external-oblique EMG and stability exercises the most internal-oblique, no exercises named (REF-1633, B-, a marker, healthy adults); in 21 male college athletes of mixed sports a programme with the Pallof press and side plank raised plank time within the group, economy no different from control (REF-1019, B-); planks vs anti-rotation has no head-to-head trial (REF-1018, B+) |
| Lying curl for a standing single-leg curl | Yes, as a swap; a seated curl, if there is one, is the pick for hamstring size | Seated 14% vs lying 9% growth after 12 weeks of 5 × 10 at 70% 1RM twice a week; both legs then about equally protected from an eccentric bout (REF-1634, B+, 20 adults, training status unstated); a standing curl acting like the lying one is that entry's inference; no trial tests it |
| Two curls, no leg extension | One knee-flexion exercise beside the hinge; a runner who sprints swaps the second curl for a Nordic; a leg extension is optional, the squat or leg press loads the front of the thigh (practice); a masters cyclist may add it | More exercises did not predict more gain (REF-1478, B+); Nordic programmes about halved hamstring injuries (REF-766, A), low volume matching high for eccentric strength after 6+ weeks (REF-887, A-); 3 weeks of knee extensions raised the cycling efficiency of 9 masters endurance athletes, sport not stated (REF-860, B, no non-lifting control; SYNTH-43) |
| Upper-body endurance to hold the bars | 2-3 sets of 10-12, short of failure, rows and back extensions included, plus the trunk block; not 25-30-rep sets (practice) | As in the Default (REF-1509, REF-1510, REF-1476); trunk training raised local muscular endurance most (REF-1632, A-); riders with low back pain had weaker back-extensor endurance in small observational studies, no training tested (REF-1266, B); no trial tests training for the bars |
| Only kettlebells | Goblet squat, split squat or step-up, single-leg Romanian deadlift, swing, press, row, twice a week; a road or TT rider keeps one two-legged lift as heavy as the bells allow (practice); for a cyclist, SYNTH-43's no-gym prescription (all-out on-bike efforts against very high resistance, 2×/week) can carry the heavy leg work, with the bells for the rest (practice); a runner hears that the running-economy evidence is for heavy loads (SYNTH-44) | The barbell raised back-squat 1RM more over 6 weeks, jump did not differ, and only that group trained the test lift (REF-1636, B+, 30 young men); kettlebell effects typically trivial to small (REF-1635, B-); two-legged lifting improved economy where alternating single-leg lifting did not (REF-148, B, pilot); low-cadence all-out efforts matched squats in well-trained cyclists (REF-158, B+) |
| Dumbbells and a pull-up bar | Dumbbell versions of the default, single-leg if light; pull-ups (assisted or lowering-only, then reps, then load) and push-ups or a press, by SYNTH-101's bodyweight row (practice) | 6RM band push-ups raised bench 1RM like the bench press over 5 weeks in trained students (REF-671, B+), progressive push-ups over 4 in moderately trained men (REF-673, B+); no entry tests pull-up progressions |
| "Why a circuit?" | Shorter, not stronger: exercises alternated, main lifts heavy; a cyclist after sprint power keeps long rests on the heavy lift (inference) | In 33 trained lifters, 6RM circuits with about 35 s between sets matched 3-min rests for bench and half-squat 1RM and lean mass over 8 weeks; only long rests raised Wingate peak power (REF-1639, B+); supersets cut sessions about 37%, 1RM SMD 0.10 (−0.40 to 0.60, three small trials), RPE higher (REF-1495, A-) |
| Circuit with run bouts for part of the aerobic volume | No: a circuit is added to the running, not swapped for it (SYNTH-44); a runner short of time keeps the runs, shortens the circuit or moves it to another day (practice) | 5-km gains over 12 weeks did not differ detectably between 3 runs with 3 CrossFit sessions and 5 runs, but VO2max rose more on 5 runs (4.3 vs 1.78 mL/kg/min) (REF-1501, B-, 21 recreational runners); circuits: VO2max +6.3%, maximal aerobic speed or power +0.3%, before-after, no endurance athletes named (REF-1638, B+); kettlebells left aerobic fitness unchanged (REF-1637, B+, workers); strength is added to running, not swapped for it (SYNTH-44) |
| "Why the upper body in a circuit?" | Time: pair a push with a pull, or upper with lower body; not two exercises for the same muscle (practice) | Push-pull pairs allowed more reps (SMD 0.68, two studies, low certainty), same-muscle pairs lost volume load (SMD −1.08) (REF-1495, A-, acute sessions in mostly young trained lifters); circuits raised bench 1RM against control in healthy adults (REF-1494, B+) |
| Runner: circuit or heavy lifting | Heavy or explosive lifting; the circuit is the no-load fallback (SYNTH-44) | All three raised 1RM, jump and vVO2max; economy improved with explosive and later maximal work and was not reported to after the circuit (n = 7, no between-group test; REF-1641, B); isometric and plyometric work beat a bodyweight circuit on a 2.4-km time trial (REF-1157, B) |
How to progress / periodise
The next number is [SYNTH-101]'s: one axis per exercise; a bodyweight move goes to the top of its range, then a harder variation or load; a circuit compares only with itself, repeated as a benchmark every 4-6 weeks (practice).
Trunk: longer holds or more reps, then a harder variation (practice); power and sprint effects went with programmes of more than 18 sessions (REF-1632, A-, a between-study subgroup, not a tested dose).
When NOT to use it / red flags
A circuit or kettlebell session logged as aerobic work for a trained athlete: kettlebells left aerobic fitness unchanged (REF-1637); 45-min weights circuits drew 20.4-29.9 mL/kg/min, highest with cardio stations (REF-1640, B, participants not described), a modest share of a trained athlete's VO2max (inference).
Fast, complex lifts in a tired circuit before they are learned: high-intensity functional training ran 4.3 injuries per 1000 h overall and 9.9 in prospective cohorts, 26% each at shoulder and back (REF-1489, B+; rates, not causes); learn swings and single-leg lifts at moderate load first (SYNTH-105, practice).
Sets to failure as routine: short of failure stays the rule (SYNTH-43, SYNTH-44); on a light home load the last set may end 1-2 reps short, a rep nearer than the usual 2-3 to spare (SYNTH-43, SYNTH-101), because the light-load muscle result comes from sets taken to failure (REF-677) (practice and inference).
Pain: pain that lingers after an exercise: swap or drop it, a clinician if it persists; sharp joint pain, bony pain, or pain or a tendon worsening across sessions: stop that exercise and see a clinician (SYNTH-44, SYNTH-101; SYNTH-73; SYNTH-94 running; SYNTH-92 the bike); new, severe or worsening back or neck pain, or pain with leg symptoms, goes to a clinician (practice).
A new or damaging home session (first Nordics, first heavy single-leg work) near a key ride or run: SYNTH-31 sets the gap.
What the evidence does and does not show
Does not show: a ranking of named trunk exercises; standing vs lying curls; a kettlebell- or dumbbell-only programme in endurance athletes; circuits with run bouts replacing a trained athlete's aerobic volume (one recreational comparison, REF-1501); training for holding the bars.
Evidence gaps
No trial compares the Pallof press, side plank or any named trunk exercise on strength, endurance, performance or injury.
No study of upper-body or trunk training for fatigue on long rides.
No trial of circuits against separate strength and endurance sessions in trained endurance athletes; REF-1638 compares only with control and baseline.
How this entry was checked and changed
Why this level
the direction rests on B+ trials and reviews, most read from abstracts (curls, kettlebells, circuits, load and muscular endurance), A- pools on trunk training and supersets (REF-1632, REF-1495) and ACSM's 2026 stand (REF-1476, A); B-, not B, because the default's picks (which trunk move, which curl, the upper-body reps, the home rules of single-leg work, slower lowering and nearness to failure) are the base's practice, and the two runner trials of circuits are B with arms of 7-11 (REF-1641, REF-1157); who was measured, mostly untrained or recreational adults, belongs on the Tier line, not in the level
Content correction 2026-11-07
2026-11-07: written after athletes asked about exercise choice, home sessions with kettlebells or dumbbells, and circuits; the base held the load, core and circuit entries (REF-677, REF-1018, REF-1019, REF-1494, REF-1509, REF-1510) without a guide answering them; ten entries added the same day (REF-1632…REF-1641) carry the exercise, kettlebell and circuit evidence.