What Indoor Rowing Trains: A Research-Grade Synthesis

Abstract

Indoor rowing is one of the few mass-produced modalities that simultaneously stresses the aerobic energy system, the anaerobic glycolytic system, the phosphocreatine system, the legs, the trunk, the upper body, and the grip — all in a single coordinated stroke pattern that is repeatable on a machine in any climate ([7] World Rowing indoor discipline, Level 5; [11] Hagerman 1984, Level 5; [13] Kleshnev 2003, Level 5; [14] Kleshnev online, Level 5).

The leg drive — quadriceps, glutes, and hamstrings — supplies the majority of the work per stroke; the trunk is the force transfer, not a primary producer; the arms finish, they do not lift ([16] Parkhouse 1998, Level 2b; [1] Concept2 technique guide, Level 5; [13] Kleshnev 2003, Level 5). The aerobic system supplies most of the energy at every pace, but the anaerobic glycolytic contribution rises sharply as the piece shortens — from under 10% on a 60-minute row to roughly 30-40% at 2K pace ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5).

The cardiovascular adaptation is large: VO2max in trained indoor rowers sits at 60-80 ml/kg/min, the upper end of the endurance-athlete range, and the consistent adaptation is the endurance-cardiac phenotype — large stroke volume, low resting heart rate, and high capillary density in the working muscles ([17] Volianitis 2006, Level 2b; [20] Steinacker 1992, Level 5).

The muscular and metabolic adaptations depend on training distribution: polarized work develops the aerobic base and the endurance-cardiac phenotype, while 2K-heavy work develops the lactate-tolerant peak-power phenotype — both are real adaptations, neither is uniquely "what indoor rowing trains" ([12] Seiler & Kjerland 2006, Level 4; [11] Hagerman 1984, Level 5).

The bottom line: indoor rowing is a generalist trainer that recruits the largest muscle mass on the body, recruits every energy system at every pace, and produces whichever adaptation the rower's training distribution selects — and the MyNextRow AI coach reads the distribution, the pace, and the stroke efficiency to choose what to develop next ([10] British Rowing online coaching, Level 5; [12] Seiler & Kjerland 2006, Level 4; [5] Concept2 stroke efficiency, Level 5; [6] Concept2 force curve, Level 5).

What indoor rowing is, and why it recruits so much

The Concept2 RowErg (and its water-, air-, and magnetic-rower siblings) is a sliding-seat, chain-drive, flywheel-resisted machine that converts a rower's leg drive, trunk swing, and arm pull into rotational work on a flywheel ([1] Concept2 technique guide, Level 5; [7] World Rowing indoor discipline, Level 5). Every stroke has four phases — the catch (the start, where the legs are compressed and the body is forward), the drive (the legs push first, then the trunk opens, then the arms pull), the finish (the legs are flat, the body leans slightly back, the handle is in the chest), and the recovery (the slide returns in the reverse order: arms, body, legs) ([1] Concept2 technique guide, Level 5; [8] World Rowing biomechanics, Level 5; [13] Kleshnev 2003, Level 5). The flywheel spins at a speed proportional to the work done — Concept2's damper setting changes the load curve but not the work, so the same split time at different damper settings represents the same aerobic work with different force-per-stroke demands ([3] Concept2 damper setting, Level 5).

The reason indoor rowing recruits so much of the body is that the stroke uses the largest muscle groups in a coordinated, sequenced pattern. The legs are the prime movers — they supply roughly 60% of net work per stroke in trained rowers ([16] Parkhouse 1998, Level 2b; [13] Kleshnev 2003, Level 5). The trunk — paraspinals, abdominals, obliques, glutes — supplies another 20-25% as the rower opens and closes the body angle. The arms, shoulders, and grip supply the remaining 15-20%, mostly in the final third of the drive. The pattern is leg-led across the stroke, with the upper body progressively joining as the legs finish their work, and the reverse pattern in the recovery ([1] Concept2 technique guide, Level 5; [8] World Rowing biomechanics, Level 5).

The first practical implication is that the indoor rower is not primarily an "arm exercise" — it is a leg-and-trunk exercise that uses the arms. The second is that the rower's stroke rate and pace are not coupled — the same pace can be held at a lower rate (higher force per stroke, more muscular endurance) or at a higher rate (lower force per stroke, more cardiovascular stress), and the AI coach picks the target based on what the rower is trying to develop ([4] Concept2 stroke rate, Level 5; [12] Seiler & Kjerland 2006, Level 4).

Energy systems: aerobic and anaerobic work on the same machine

The energy the rower needs to power the flywheel comes from three systems — the phosphocreatine (PCr) system, the glycolytic (anaerobic) system, and the aerobic (oxidative) system ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5). The three systems scale with duration: PCr dominates the first 10-15 seconds, glycolysis dominates roughly 15 seconds to two minutes, and aerobic dominates everything longer ([11] Hagerman 1984, Level 5).

For indoor rowing, the practical implications are direct. A 30-second all-out sprint on the rower — the kind the AI coach might prescribe as a one-of-the-day session — is almost entirely PCr and glycolytic; the aerobic system contributes relatively little ([11] Hagerman 1984, Level 5; [16] Parkhouse 1998, Level 2b). A 2K piece (~7 minutes) is roughly 60-70% aerobic and 30-40% glycolytic, with PCr negligible ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5). A 5K piece (~18-22 minutes) is roughly 80% aerobic and 20% glycolytic. A 10K piece (~35-45 minutes) is roughly 90% aerobic and 10% glycolytic. A 60-minute piece is overwhelmingly aerobic — the glycolytic and PCr systems have run out of contribution long before the rower crosses the finish ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5).

The energy-system contribution matters because each system adapts differently. The aerobic system adapts by increasing stroke volume (more blood per beat), capillary density (more delivery to the working muscle), mitochondrial density (more oxidative capacity per cell), and to a lesser extent VO2max ceiling ([17] Volianitis 2006, Level 2b; [20] Steinacker 1992, Level 5). The glycolytic system adapts by increasing buffering capacity (tolerating more lactate), glycolytic enzyme activity, and the capacity to tolerate high-power efforts longer ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5). The PCr system adapts by increasing the alactic reserve and the rate at which PCr can be resynthesised during recovery intervals.

The consequence for the rower's training programme is direct: if the rower does only easy steady-state, they adapt the aerobic system and leave the glycolytic and PCr systems at baseline. If the rower does only 2K-style work, they adapt the glycolytic and PCr systems but leave the aerobic system under-developed and produce the lactate-tolerant, peak-power phenotype without the endurance base. The polarized training distribution the literature supports — roughly 75% easy, 7-8% moderate, 17-22% hard — develops all three systems together ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5; [22] ACSM guidelines, Level 5).

Cardiovascular adaptations: the endurance-cardiac phenotype

Indoor rowing is, in the cardiovascular literature, an endurance exercise. VO2max in trained indoor rowers sits in the same range as trained distance runners and trained cyclists — typically 60-80 ml/kg/min for trained male rowers, slightly lower for trained female rowers, and well above the sedentary baseline ([17] Volianitis 2006, Level 2b; [11] Hagerman 1984, Level 5). The cardiovascular adaptation to consistent indoor training follows the standard endurance-exercise pattern: increased left-ventricular stroke volume, lower resting heart rate, lower submaximal exercise heart rate at the same aerobic power output, increased capillarisation of the working muscles, increased blood plasma volume, and improved endothelial function ([17] Volianitis 2006, Level 2b; [20] Steinacker 1992, Level 5; [11] Hagerman 1984, Level 5).

The cardiac adaptation is large because the rower uses so much muscle mass. Heart rate and stroke volume during an indoor row climb together until roughly 40-50% of VO2max, after which stroke volume plateaus and heart rate carries the load alone — the same pattern as other whole-body endurance exercise ([17] Volianitis 2006, Level 2b; [20] Steinacker 1992, Level 5). The maximum sustainable cardiac output in trained indoor rowers is among the largest recorded in any athletic population, and the resting bradycardia that develops with consistent training is a direct consequence of the chronic stroke-volume expansion and the parasympathetic shift that goes with it ([17] Volianitis 2006, Level 2b; [11] Hagerman 1984, Level 5).

The VO2max range that matters for health and mortality is population-level, not athlete-specific. The 2009 [27] Kodama meta-analysis established the canonical dose-response: every 1-MET (~3.5 ml/kg/min) increase in CRF is associated with roughly 13-15% lower all-cause mortality. The 2015 [25] Bouchard, Blair, and Katzmarzyk paper extended this with the cardiorespiratory-fitness-as-mortality-marker framing: low CRF (VO2max below ~18 ml/kg/min) is associated with substantially elevated all-cause mortality. Indoor rowing is one of the modal modalities that produces those CRF gains — at the population level, the 1998 [21] Pollock Position Stand recommends 150-300 min/week of moderate or 75-150 min/week of vigorous aerobic exercise for general health, and indoor rowing comfortably reaches either target at moderate doses ([21] Pollock 1998, Level 5; [22] ACSM guidelines, Level 5; [23] NHS physical activity, Level 5; [25] Bouchard et al. 2015, Level 5; [27] Kodama et al. 2009, Level 2a).

The practical read: a rower who does three indoor sessions per week at moderate intensity for 30-45 minutes each reaches the vigorous-exercise dose and develops a measurable CRF improvement within 8-12 weeks. The rower who adds intervals — even one interval session per week — produces a larger CRF improvement, particularly if the interval work reaches 85-95% of maximum heart rate ([21] Pollock 1998, Level 5; [22] ACSM guidelines, Level 5; [17] Volianitis 2006, Level 2b).

Muscular recruitment: legs first, trunk second, arms last

The EMG literature on indoor rowing is unambiguous about which muscles contribute what. The [16] Parkhouse 1998 study — the most-cited indoor-rowing EMG paper — found that per-stroke muscle activation order is legs first (quadriceps, glutes, hamstrings), trunk second (paraspinals, abdominals, obliques), arms third (lats, deltoids, biceps), and grip last (forearm flexors). The legs supply roughly 60% of net work, the trunk roughly 20-25%, the arms roughly 15-20%, and the grip a residual [16] Parkhouse 1998 (Level 2b; [13] Kleshnev 2003, Level 5). The pattern is robust across speeds and stroke rates — changing pace or rate scales the activation magnitude but not the order ([16] Parkhouse 1998, Level 2b).

The muscular adaptation follows the recruitment. The leg drive develops quadriceps, gluteus maximus, and hamstrings as the prime movers; the paraspinals and abdominals develop as stabilisers and force-transfer; the lats, deltoids, and biceps develop as the second-half-of-drive contributors; the forearm flexors develop as the grip support. The force-curve shape the rower produces reflects this ordering — a smooth ramp-up from the leg push, a maintained peak through the trunk swing, a tail-off through the arm pull ([13] Kleshnev 2003, Level 5; [5] Concept2 stroke efficiency, Level 5; [6] Concept2 force curve, Level 5).

The practical implication for the AI coach is that stroke efficiency is partly technique and partly muscular endurance. A rower whose stroke is jerky at the catch — sudden peak force as the legs engage — is recruiting too much quadriceps relative to the trunk and arms, and the metabolic cost of the same pace is higher ([18] Hofmijster et al. 2009, Level 2b; [5] Concept2 stroke efficiency, Level 5). A rower whose stroke is smooth-ramping, with a maintained mid-drive plateau and a controlled finish, distributes the work across the muscle groups and reaches a lower oxygen cost per stroke at the same pace ([18] Hofmijster et al. 2009, Level 2b; [13] Kleshnev 2003, Level 5).

The trap is to read indoor rowing as an "arm and shoulder" exercise. It is not. The arms and shoulders are the lagging muscle groups, the ones whose work begins after the legs are mostly done. A rower who rows with arms-and-back but lets the legs lag is overloading the trunk and arms, underloading the legs, and producing a stroke whose metabolic cost is higher and whose power is lower ([16] Parkhouse 1998, Level 2b; [1] Concept2 technique guide, Level 5). The first thing a coach corrects on technique is sequencing — legs first, trunk second, arms third — because sequencing is what makes indoor rowing the cardiovascular stimulus it can be, instead of the strength-endurance stimulus it accidentally becomes ([1] Concept2 technique guide, Level 5; [8] World Rowing biomechanics, Level 5).

The trunk and the force-transfer problem

The trunk is the special case among the muscle groups, because it is doing two jobs at once. It has to be a stable platform through which leg power transfers to the handle, and it has to be an active producer through the trunk-swing phase of the drive. The dual role is what makes the trunk the limiting factor in many rowers, not the prime mover ([13] Kleshnev 2003, Level 5; [8] World Rowing biomechanics, Level 5).

When the trunk fails as a stable platform, the rower's stroke shows up as a poor force curve — the catch is jerky, the peak force is high and brief, the per-stroke work is lower than it should be for the metabolic cost. The fix is the sequencing already covered: legs first, then the trunk opens, then the arms pull. When the trunk fails as an active producer, the rower's stroke shows up as a soft finish — the drive ends before the rower has reached full extension, the handle arrives at the chest low rather than high, and the per-stroke work is lower because the second half of the drive under-contributes ([13] Kleshnev 2003, Level 5; [5] Concept2 stroke efficiency, Level 5).

The trunk training that supports both jobs is a mix of anti-rotation (plank variants, Pallof press, suitcase carry), anti-extension (dead bug, hollow hold), and dynamic trunk swing (Russian twists with light load, med-ball rotations). The McGill core-training review is the evidence base for this prescription — plank variants, Pallof press, dead bug, and Russian twists are the spine-stability exercises the literature defends ([26] McGill 2010, Level 4). The Concept2 guidance and the federation biomechanics both reinforce that trunk strength is non-optional for indoor rowing — it is the only way the legs can deliver power to the handle effectively ([1] Concept2 technique guide, Level 5; [8] World Rowing biomechanics, Level 5; [13] Kleshnev 2003, Level 5).

The cross-domain pattern is also worth noting: the recovery sequence (the slide back to the catch) is the same set of muscles performing their second job, and a rower who has not trained the trunk to recover actively tends to collapse the recovery into a short, rushed slide ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5). The force transfer and the recovery sequence are two faces of the same trunk capacity, and the AI coach reads both through the per-stroke force curve and the PM5's drive-ratio metric.

Upper body and grip: the smaller contributors that still matter

The upper body — lats, deltoids, biceps, forearm flexors — supplies roughly 15-20% of net work in the second half of the drive ([16] Parkhouse 1998, Level 2b). The contribution is small but not negligible: a rower with strong legs and weak lats produces a stroke whose drive peaks too early and falls off before the finish, and the per-stroke work is limited by the lat contribution that never arrives ([1] Concept2 technique guide, Level 5; [13] Kleshnev 2003, Level 5).

Grip is the special case within the upper body. The handle has to be held with sufficient force to transmit the work, but the rower does not want a grip so tight that the forearm flexors fatigue before the legs do. The practical anchor is to hold the handle firmly enough to control it, not to crush it — the work is done by the legs and trunk; the grip is the link to the handle, not the prime mover ([1] Concept2 technique guide, Level 5). On long rows, forearm fatigue can precede leg fatigue in rowers who wrap the thumb under the handle rather than resting it on top, and a relaxed grip from catch to mid-drive is the simple mitigation ([1] Concept2 technique guide, Level 5).

The upper-body training that supports indoor rowing is straightforward: vertical pulls (pull-ups, lat pulldowns), horizontal pulls (seated cable row, dumbbell row), and grip work (farmer's carry, dead hangs, wrist curls). The load should be moderate and the volume should not crowd out time on the rower — the indoor stroke itself does most of the upper-body endurance work, and the supplementary strength work is for capacity rather than for the rower-specific stimulus ([22] ACSM guidelines, Level 5; [21] Pollock 1998, Level 5).

Energy contribution by duration: pacing as energy-system selection

The duration of an indoor row determines which energy system dominates, and the duration plus the intensity determines which adaptations the rower is producing ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5).

• 30-second all-out is roughly 90% PCr and glycolytic, 10% aerobic. Develops PCr reserve, glycolytic capacity, and lactate tolerance ([11] Hagerman 1984, Level 5).
• 500m race piece (~90 seconds) is roughly 50% glycolytic, 40% aerobic, 10% PCr. Develops glycolytic capacity and aerobic ceiling together ([11] Hagerman 1984, Level 5).
• 1000m race piece (~3-4 minutes) is roughly 35% glycolytic, 60% aerobic, 5% PCr. Develops aerobic ceiling with a meaningful glycolytic contribution ([11] Hagerman 1984, Level 5).
• 2K race piece (~7 minutes) is roughly 30% glycolytic, 65% aerobic, 5% PCr. Develops aerobic ceiling with a substantial lactate-tolerance contribution ([11] Hagerman 1984, Level 5).
• 5K race piece (~18-22 minutes) is roughly 20% glycolytic, 80% aerobic. Develops aerobic ceiling and lactate clearance ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5).
• 10K race piece (~35-45 minutes) is roughly 10% glycolytic, 90% aerobic. Develops the aerobic base without the lactate-tolerance contribution ([11] Hagerman 1984, Level 5).
• 30-60 minute steady-state is essentially 100% aerobic. Develops the aerobic base and the cardiovascular adaptations without the glycolytic system joining ([11] Hagerman 1984, Level 5).

The practical read for the AI coach is direct: the rower's recent history of durations is the rower's recent energy-system demand. A rower who has done mostly 5K+ pieces is building the aerobic base, and the AI coach can prescribe more interval work to develop the glycolytic system. A rower who has done mostly 2K-style work is building the lactate-tolerant peak-power phenotype, and the AI coach can prescribe longer steady-state to develop the aerobic base ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5).

The de Koning 1999 variational model anchors the pacing decision within each piece: trained rowers settle into a conservative-start, end-spurt shape because that pacing maximises mean power while staying within the metabolic ceiling ([19] de Koning 1999, Level 5). The AI coach's interval prescription uses the same logic at the per-piece level, with workout design and per-piece pacing aligned ([19] de Koning 1999, Level 5; [4] Concept2 stroke rate, Level 5).

Skill and the technique ceiling: an inefficient stroke costs more energy

Indoor rowing has a skill ceiling that, for many rowers, sits higher than the fitness ceiling. A rower with limited technique can hit roughly 70-80% of what a same-fitness trained rower produces on the same pace, simply because the inefficient stroke recruits too much of the wrong muscles, breaks the force curve, and demands a higher VO2 for the same work ([18] Hofmijster et al. 2009, Level 2b; [13] Kleshnev 2003, Level 5).

The mechanism is the force-curve shape. A rower who lifts the handle from the catch with the arms and lets the legs lag produces a force curve that peaks too early — the legs are doing less than they could, the arms are doing more than they should, and the per-stroke work is lower than the metabolic cost ([13] Kleshnev 2003, Level 5; [5] Concept2 stroke efficiency, Level 5). A rower whose sequencing is legs-first, trunk-second, arms-third produces a smooth force curve that peaks in the mid-drive and tails off through the finish, with per-stroke work higher and metabolic cost lower ([18] Hofmijster et al. 2009, Level 2b; [13] Kleshnev 2003, Level 5).

The difference is not small. The Hofmijster 2009 paper found that gross mechanical efficiency on the indoor rower is roughly 18-20%, and changes in sequencing that look subtle can shift the metabolic cost of the same pace by several percentage points — enough to be the difference between a 7-minute 2K and a 7:10 2K without any change in physiology ([18] Hofmijster et al. 2009, Level 2b; [13] Kleshnev 2003, Level 5). The AI coach reads stroke efficiency from the PM5's per-stroke force-curve shape and the drive-ratio metric, and uses both to identify when the rower's technique has slipped and to flag technique-focused drills as a session insertion ([5] Concept2 stroke efficiency, Level 5; [6] Concept2 force curve, Level 5; [2] Concept2 PM5 monitoring, Level 5).

The takeaway for the rower: developing the skill dimension of indoor rowing is not just "rowing more." It is deliberate technique work — pick drill, pause drills, reverse-pick drill, race-pace starts — that lets the rower practice the sequencing and the force-curve shape the indoor stroke demands ([1] Concept2 technique guide, Level 5; [9] British Rowing Go Row Indoor, Level 5).

Acute effects vs chronic adaptations: what one session does, what months of training do

A single indoor session produces acute effects that are quite different from the chronic adaptations of months of consistent training. Acute effects of one session include elevated heart rate, elevated stroke volume during the session, transient rise in lactate at higher intensities, transient rise in catecholamines, glycogen depletion in the working muscles at longer durations, mild dehydration at moderate doses, and an acute inflammatory response that resolves within 24-48 hours ([20] Steinacker 1992, Level 5; [17] Volianitis 2006, Level 2b).

The chronic adaptations of consistent indoor training are different and slower. The cardiovascular adaptation — increased stroke volume, decreased resting heart rate, increased capillary density, increased blood plasma volume — typically takes 6-12 weeks to become measurable, with larger gains at 12-24 weeks ([17] Volianitis 2006, Level 2b; [11] Hagerman 1984, Level 5; [21] Pollock 1998, Level 5). The metabolic adaptations — increased mitochondrial density, increased oxidative enzyme activity, increased glycogen storage capacity — follow a similar timeline but with larger individual variability. The muscular adaptations — increased quadriceps and gluteal muscle mass in trained rowers, increased tendon and ligament robustness, increased bone density in the lower-body — are typically measurable at 12-24 weeks and continue for years ([22] ACSM guidelines, Level 5).

The implication for the AI coach is that one session produces a stimulus, and the stimulus converts to adaptation only if the next several sessions are consistent enough to cumulate. A rower who does one good session and then skips two weeks and does one good session again is producing repeated acute stimuli with no chronic adaptation ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5). A rower who does three sessions per week for twelve weeks produces the chronic adaptations ([21] Pollock 1998, Level 5; [22] ACSM guidelines, Level 5).

The hydration and glycogen-replacement piece is part of the acute-effect story. The 2007 [24] Sawka et al. ACSM Position Stand established that dehydration above 2% body-mass loss measurably reduces endurance performance, and the carbohydrate-replacement recommendation (30-60 g/hour during exercise longer than ~60-90 minutes) keeps the rower's glycogen supply from running out before the piece does ([24] Sawka et al. 2007, Level 5). Indoor sessions under 60 minutes typically need only water; sessions over 60 minutes benefit from a carbohydrate-electrolyte mix and a pre-session carbohydrate load ([24] Sawka et al. 2007, Level 5; [23] NHS physical activity, Level 5).

Overtraining and injury risk: the dose-response curve, not a free lunch

Indoor rowing is a high-load, repetitive, low-impact modality — three properties that together produce a useful exercise dose but also produce a real injury profile if the load outpaces the rower's recovery capacity. The dominant overuse injury sites are the lower back, the knee extensor mechanism (patellar tendon, quadriceps tendon), the shoulder (especially the rotator cuff), and the wrist/forearm (especially for rowers who hook-grip) ([1] Concept2 technique guide, Level 5; [10] British Rowing online coaching, Level 5).

The dose-response curve is real: a rower who progresses volume by less than ~10% per week tolerates the load; a rower who doubles volume in a single week has substantially elevated injury risk. The AI coach reads training load across the recent sessions and flags when the rower's weekly volume has outpaced the previous week's by more than the safe progression envelope, recommending a step-back week before the next build ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5).

The monotony factor matters too. Indoor rowing uses the same muscle groups in the same pattern every session, and a rower who does only indoor rowing without complementary strength work, mobility work, or cross-training accumulates a load that the same-tissue recovery system cannot keep up with. The mixed-modality recommendation — at least one cross-training session per week, plus targeted strength and mobility work — produces a more balanced dose and a lower overuse-injury rate ([12] Seiler & Kjerland 2006, Level 4; [22] ACSM guidelines, Level 5).

The signalling for overtraining — chronic fatigue, declining performance at the same session intensity, elevated resting heart rate, poor sleep, loss of appetite — is the same across endurance modalities. The AI coach reads the rower's recent pace trends, heart-rate trends where available, and self-reported readiness to flag early warning signs before the overtraining syndrome fully develops ([10] British Rowing online coaching, Level 5; [21] Pollock 1998, Level 5).

What indoor rowing is not, and where the framing falls over

Indoor rowing is not a substitute for on-water rowing. The on-water rower is dealing with balance, water resistance, and crew timing, none of which the indoor rower has — an indoor-trained rower who transitions to the water has to relearn the balance and the timing, and the indoor time alone does not produce the water-specific adaptations ([7] World Rowing indoor discipline, Level 5; [8] World Rowing biomechanics, Level 5). The indoor rower is a physical conditioning tool, not a technical rowing simulator.

Indoor rowing is also not a strength training modality in the barbell sense. The leg drive is a hip-and-knee extension under load, but the load is the rower's body weight and the flywheel resistance — not external resistance that the rower can progressively overload in the strength-training sense. Strength gains on the indoor rower plateau at a point determined by the rower's body mass and the flywheel's load curve, which is one reason strong indoor rowers cross-train with barbell or dumbbell strength work ([22] ACSM guidelines, Level 5; [1] Concept2 technique guide, Level 5).

Indoor rowing is also not a low-injury-risk modality in absolute terms. The overuse-injury sites — lower back, knee extensor, shoulder, wrist — are real and common enough that any indoor-rowing programme should include complementary strength, mobility, and technique-drill work to keep the load balanced ([10] British Rowing online coaching, Level 5; [12] Seiler & Kjerland 2006, Level 4). The injury profile is lower than high-impact running, but it is not lower than swimming or cycling, both of which distribute the load across different tissues.

What the MyNextRow AI coach does with this knowledge

The MyNextRow AI coach reads the same physiology, biomechanics, and energy-system framework this article describes, and uses it to choose what the rower's next session should develop. Three signals matter.

The first is training distribution across the recent sessions. A rower whose last seven sessions have been all 5K-plus pieces is aerobic-loaded and can tolerate interval work; a rower whose last seven sessions have been mostly 2K-style intervals is glycolytic-loaded and would benefit from longer steady-state ([12] Seiler & Kjerland 2006, Level 4; [11] Hagerman 1984, Level 5). The polarized training distribution the literature defends — roughly 75% easy, 7-8% moderate, 17-22% hard — gives the coach a reference distribution to compare against ([12] Seiler & Kjerland 2006, Level 4).

The second is recent pace trend at a given distance. A rower whose 5K pace is improving on the recent history is producing a real adaptation; a rower whose 5K pace is flat or regressing is either under-recovering, training distribution has drifted, or the rower's technique has slipped. The AI coach flags each interpretation differently — step-back week for under-recovery, redistribution for distribution drift, technique drills for technique drift ([6] Concept2 force curve, Level 5; [5] Concept2 stroke efficiency, Level 5; [10] British Rowing online coaching, Level 5).

The third is the per-stroke force-curve shape and the drive-ratio metric the PM5 reports. A rower whose force curve has flattened or whose drive ratio has changed is signalling a technique or muscular-fatigue shift, and the AI coach selects the session type accordingly ([2] Concept2 PM5 monitoring, Level 5; [5] Concept2 stroke efficiency, Level 5).

The combined read — distribution, pace trend, per-stroke mechanics — is what the AI coach uses to choose among an easy aerobic row, an interval session, a 2K test, a 5K test, a strength-focused session, a technique-focused session, or a recovery row. The framing is that indoor rowing produces a multi-dimensional adaptation, and the coach's job is to develop the right dimension at the right time ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5).

Practical takeaways

Indoor rowing is a whole-body cardiovascular stimulus that recruits the largest muscle groups in a coordinated, sequenced pattern, stresses every energy system at every pace, and produces whichever adaptation the rower's training distribution selects. The adaptation that follows is the adaptation the rower's recent sessions have stressed, not a generic "indoor rowing" phenotype.

The cleanest practical reads for the rower:

1. Volume before intensity. A rower who is building fitness should progress weekly volume by less than 10% per week, and add intensity only after the volume plateau is established. The cardiovascular adaptations take 6-12 weeks to become measurable; the rower who stops after three weeks is producing acute stimuli with no chronic adaptation ([17] Volianitis 2006, Level 2b; [21] Pollock 1998, Level 5; [22] ACSM guidelines, Level 5).

2. Sequence legs first. The stroke is leg-led; the rower who lets the upper body lead is producing a stroke that recruits the wrong muscle groups and demands a higher VO2 for the same pace. The sequencing correction is the highest-leverage technique intervention the AI coach can recommend ([1] Concept2 technique guide, Level 5; [16] Parkhouse 1998, Level 2b; [8] World Rowing biomechanics, Level 5).

3. Match the piece to the system. A rower who wants to develop the aerobic base should do 60+ minute steady-state or 5K-10K paced rows; a rower who wants to develop lactate tolerance should do 2K-style interval work; a rower who wants to develop the PCr system should do 30-second all-out sprints ([11] Hagerman 1984, Level 5; [15] Mahler 1984, Level 5; [12] Seiler & Kjerland 2006, Level 4).

4. Hydrate before, during, and after. Sessions under 60 minutes typically need only water; sessions over 60 minutes benefit from a carbohydrate-electrolyte mix. Dehydration above 2% body-mass loss measurably reduces endurance performance ([24] Sawka et al. 2007, Level 5; [23] NHS physical activity, Level 5).

5. Cross-train at least once per week. The indoor rower's load is concentrated in the same muscle groups in the same pattern; complementary swimming, cycling, strength, or mobility work distributes the load and reduces overuse-injury risk ([12] Seiler & Kjerland 2006, Level 4; [22] ACSM guidelines, Level 5).

6. Trust the distribution, not the single session. The adaptation comes from the cumulative pattern across weeks, not from one good session. The polarized-distribution evidence is consistent enough that the rower who follows it produces a balanced adaptation; the rower who ignores it produces the imbalance the literature describes ([12] Seiler & Kjerland 2006, Level 4; [10] British Rowing online coaching, Level 5).

7. Read the force curve. The PM5's per-stroke force curve and drive-ratio metric are the technique signal the AI coach reads. A rower whose force curve has flattened or whose drive ratio has shifted is signalling a technique change the recovery or drill session can address ([5] Concept2 stroke efficiency, Level 5; [6] Concept2 force curve, Level 5; [2] Concept2 PM5 monitoring, Level 5).

Open questions and where the evidence is still soft

The indoor-rower-specific physiology literature is denser than it used to be, but several questions remain open or under-studied. The acute dosing question — how much indoor rowing per week is the dose-response curve showing measurable adaptation at the population level — is largely answered by the broader endurance-exercise Position Stand literature, but the population-specific data is thin ([21] Pollock 1998, Level 5; [22] ACSM guidelines, Level 5; [25] Bouchard et al. 2015, Level 5). The female-athlete-specific cardiovascular and metabolic response is improving but still thin compared to male-athlete data; the AI coach reads the available data and treats female athletes on the same framework but with a slightly lower VO2max baseline expectation ([17] Volianitis 2006, Level 2b; [11] Hagerman 1984, Level 5).

The adaptive-aging picture is also improving but sparse. Older indoor rowers reach lower absolute VO2max values but produce the same percentage-improvement response to consistent training as younger rowers, and the AI coach's reference pace adjusts accordingly. The para-athlete literature is improving but still thin enough that the AI coach treats indoor rowing as a useful modality for para-rowers without strong evidence on modality-specific dose-response ([7] World Rowing indoor discipline, Level 5; [25] Bouchard et al. 2015, Level 5).

The honest read of the literature: indoor rowing is a well-validated generalist trainer with enough modality-specific research that the rower can plan a programme with confidence, and enough general-endurance-exercise research that the programme can be adapted to age, sex, para-athlete category, or training goal without re-inventing the dose-response curve from scratch.

What to do with this article

The rower reading this article should treat it as a framework for the AI coach's choices, not a prescription. The AI coach reads the rower's recent training history, the rower's recent pace trend, the rower's stroke mechanics, and the rower's stated goal, and selects the next session to develop whichever dimension is currently the binding constraint. The framework is: cardiovascular and metabolic generalist, leg-led sequencing, energy-system selection by duration, technique ceiling that compounds over months, and adaptive response driven by training distribution rather than by any single session.

The bottom line: indoor rowing is a versatile, evidence-supported whole-body cardiovascular and muscular stimulus that produces a real adaptation in any population the dose-response curve reaches; the AI coach's job is to balance the dose across the right dimensions and to detect the technique, recovery, or distribution drift that would otherwise compound into an imbalance. The rower who lets the AI coach read the framework and applies it consistently produces the kind of balanced adaptation the literature defends.