Pain on the bike
Ask about load, earlier episodes and what changed, and make one change at a time, before blaming the fit: no bike or body measure has been reliably linked to overuse pain, a professional fit improved short-term comfort but has never been shown to prevent injury, and every set-up change is the base's practice
The question it answers
An athlete reports pain, numbness or a sore on or after riding — the front, inside or outside of the knee, the lower back, the neck or shoulders, the hands, the saddle area (men or women) or the feet. Do they keep riding, change the training, change something on the bike, or see someone, and is a professional fit worth paying for?
Short answer: red flags go to a clinician; otherwise look for what changed (the riding first, then the kit) and for earlier episodes, take that change out, then change one thing in small steps and give it about two weeks. The assistant repeats what the athlete reports and never names a diagnosis (SYNTH-73).
What this framework owns and what it leaves
| Owns (decided here) | Leaves to |
|---|---|
| The cycling-specific first questions, the one-change rule, what to try first by pain site and what not to promise | Red flags, the pain-monitoring rule (≤3/10), the two-week rule, substitution and symptom-gated return, and the no-diagnosis rule → SYNTH-73 |
| How saddle height is measured and how much a few millimetres matter | Cadence, and knee pain during low-cadence or torque work → SYNTH-55, SYNTH-82 |
| Saddle-area numbness, sores and the sexual and urinary questions cyclists ask, in men and women | Weekly ramp and deload; for cyclists every ramp number is practice → SYNTH-29; coming back after a break → SYNTH-36 |
| Whether a professional fit is worth it | Time-trial position against power and drag → SYNTH-62, SYNTH-48; position and efficiency → SYNTH-17; a growing rider's fit → SYNTH-81; contact points in ultras → SYNTH-63 |
Default (what to prescribe first)
Not from a study; the base's own practice for the order of questions, the step sizes and the waiting time; the sources say what the order rests on:
Red flags first. SYNTH-73 owns the list; any one means stop the painful riding and see a clinician, with no "easy version". The cycling-specific signs this framework adds are below, each "see a clinician".
Ask what changed, riding first, then kit, and whether it has happened before. The riding: volume, long rides, climbing, low-cadence or torque work, a new indoor block, a race. The kit: bike, saddle, shoes, cleats, cranks, bars or stem, aero bars, shorts.
Across 18 observational studies load had moderate evidence of a link to overuse symptoms (whether any other factor did is not in the abstract), many measures traditionally used in bike fitting had moderate evidence of no link, saddle height was conflicting and no measure was strongly linked (REF-1260, B+, abstract; which load and fit measures were tested is not stated).
A back history was the strongest marker of back pain in amateurs (OR 4.20, REF-1265, B+, cross-sectional), as a previous episode was in rowers (REF-641, B+).
Take the change out, not all riding. Undo the last kit change or cut the load that preceded the pain (SYNTH-73 step 4); keep riding what does not provoke it, inside SYNTH-73's pain rule (≤3/10 and settled by the next morning, the base's practice; the only tested pain-monitoring model allowed up to 5/10, in Achilles tendinopathy, REF-628, B+, so any cycling complaint is transfer).
If nothing changed, one adjustment at a time, in small steps, from the table below. Saddle height is checked with the knee angle measured while the rider pedals, which a 41-study review grades as the strong method (REF-1261, B+, abstract; its abstract gives no target angle); changes under about 4% of leg length (about 3-4 cm for most adults, the base's arithmetic) changed lower-limb loads only trivially to slightly (REF-1261), so move the saddle only when it is clearly off, and do not count on a few millimetres to settle a symptom on their own (they barely change joint loads; their effect on symptoms is untested).
Promise nothing about prevention or cure. No trial of any set-up change has a pain-incidence or injury outcome (Evidence gaps).
When to stop riding and see a clinician
Decision table by pain site
"(practice)" marks a call that is not from a study; the base's own practice. "Try first" means one change at a time, judged after about two weeks.
| Site, what the athlete says | What the base shows (refId, level, key number) | Try first | Do not promise |
|---|---|---|---|
| Front of the knee, or "my knee" | REF-1264 B+: knee 32.1% and anterior knee pain 17.2% of 1,879 overuse injuries reported in 62,758 entries to a 109-km amateur road event (self-reported diagnoses); REF-1260 B+ (abstract): saddle height conflicting, load moderate; REF-1261 B+ (abstract): set from the knee angle measured while pedalling (strong), changes under 4% of leg length trivial to small for loads (moderate), injury evidence limited; REF-882 B+: a lower saddle seems to raise patellofemoral compressive force (mechanics); REF-1076 C+: modelled patellofemoral force rose with workload and a lower saddle, not significantly with pedalling rate (6 men); REF-1267 B (abstract, uncontrolled): in 76 riders with knee pain, worst pain fell 2.52 points of 10 (95% CI −3.04 to −2) at 12 weeks after a physiotherapist-led fit plus cadence and progression advice plus exercises | Take out the recent load jump and any low-cadence or torque work (SYNTH-82); then check saddle height with the knee angle measured while pedalling and raise it in small steps only if it is clearly low (practice) | That a saddle change resolves knee pain (untested; SYNTH-73); a target angle (the abstract of REF-882 gives 25-30° with no measuring method and its full text was not read; REF-1261's abstract gives none); that the 2.5-point fall in REF-1267 was the fit's doing |
| Inside, outside or back of the knee | REF-1264 B+: ITB friction syndrome 10.3% of reported injuries (self-reported); no entry tests cleat rotation, float, stance width, wedges, insoles or saddle setback against knee pain; outer-knee pain a clinician has called ITB syndrome: SYNTH-73's row (REF-919, runners) | Restore the cleat, shoe, pedal or saddle position used before the pain began; for pain behind the knee check that the saddle is not too high or too far back (practice) | Any effect of cleats, float, wedges, insoles or setback (untested) |
| Crank length ("would shorter cranks help my knee or hip?") | REF-1026 B+: maximal power about 4% apart across 120-220 mm; REF-1028 B+: ±5 mm left gross efficiency unchanged and a longer crank added 1.8-3.4° of hip and knee flexion; REF-1029 B: 165, 170 and 175 mm gave identical economy and sprint power; no pain outcome in any; a joint-mechanics study quoted in REF-1028 does not support "shorter cranks unload the knee" | A shorter crank is a reasonable choice for a closed hip at the top of the stroke (practice; joint-angle grounds, efficiency unchanged) | Less knee pain, more power or fewer injuries from a crank change |
| Lower back | REF-1265 B+ (1,274 Italian amateurs, cross-sectional): 26.5% had back pain in the past 12 months; back pain before taking up cycling OR 4.20 (95% CI 3.21-5.40), coached or scheduled training OR 0.53 (0.37-0.74), a past bike-fit visit dropped out of the model; REF-1266 B (8 small studies, 255 men): lower bars flexed the lumbar spine more (posture studies that did not compare riders with and without pain), riders with back pain sat more flexed and had lower back-extensor endurance (one study each); REF-1264 B+: lower back 7.8% of reported injuries, the most often lasting over a year (58.8%) and severe (52.6%); REF-921 B+: 45% of overuse injuries in 109 professionals | Ask about earlier episodes and the load (long rides, long climbs, low-cadence work); raise or shorten the front end in small steps; break long seated spells by standing and moving the hands (inference from REF-1266's time-in-flexion point); back-extensor and trunk endurance work beside riding; low-cadence and torque work back last (SYNTH-82) (practice) | That higher bars or trunk training relieve back pain (neither tested in riders); a fix from a fit (a past fit did not separate riders with and without back pain, REF-1265) |
| Neck and shoulders | REF-1264 B+: shoulder 10.6% and neck 2.8% of reported injuries (some shoulder reports may be old crash injuries, the entry's inference); no entry tests reach, bar height or aero bars against neck or shoulder pain; REF-618 is about overhead athletes, not riders | Raise or shorten the front end, vary hand and head position, build time on aero bars gradually (practice) | Anything; no study. Numbness, tingling or weakness in an arm that persists after the ride or keeps returning is a SYNTH-73 red flag; arm pain spreading from the neck → clinician (practice) |
| Hands (numb or weak fingers) | REF-890 B: after a 6-day, 420-mile tour the deep motor branch of the ulnar nerve conducted more slowly (14 riders, no control) and carpal tunnel syndrome worsened in 3 hands; REF-1039 B-: 23 of 25 riders had hand symptoms after 600 km, motor symptoms no different by experience or handlebar type; REF-1040 B: pressure on the little-finger side of the palm highest in the drops, padded gloves −10-28% (pressure, not symptoms); REF-1264 B+: hand or finger numbness 3.7% of reported injuries | Rotate tops, hoods and drops; padded gloves (thin foam was enough); wrist near neutral; less weight on the hands (practice, direction from REF-1040) | That gloves or hand positions remove symptoms (pressure only); that bar height, reach or setback changes hand load (no verified study, REF-1039) |
| Saddle area, men (numbness, pain) | REF-1268 B+ (abstract; 3,932 self-selected men): standing more than 20% of the time OR 0.4, and a bar level with or above the saddle OR 0.8, for genital numbness; SHIM 19.9 and 20.7 in cyclists vs 19.5 in swimmers and runners; urethral stricture OR 2.5; REF-1270 B+ (22 interventional studies, mostly laboratory, 601 men): sitting cut penile oxygen pressure by 72.58% and standing restored it (6 studies, SMD 5.58, 95% CI 3.18-7.98); no-nose saddles cut perineal pressure by about 63% but felt less stable and loaded the sit bones more; urogenital numbness 73% → 18% after 6 months on no-nose saddles in one uncontrolled study of police cyclists; a central cut-out did not lower total pressure; REF-1271 B+ (abstract): erectile dysfunction no different unadjusted, OR 2.00 (1.57-2.55) after adjustment for age and comorbidities | Stand up regularly; set the bar level with the saddle; if numbness persists, try a no-nose or wider saddle (practice) | That any of these removes numbness (associations, acute markers and one uncontrolled cohort); a "stand every 10 minutes" rule (the review's authors call 10 min an arbitrary number); that cycling is harmless, or harmful, for erections (the two pooled readings disagree) |
| Saddle area, women (numbness, pain, swelling) | REF-1269 B+ (3,118 self-selected women): 39% of 2,065 cyclists reported genital numbness and 39% saddle sores; adjusted odds of numbness 6.5 and 9.1 and of sores 6.3 and 22.7 (low and high intensity) vs swimmers and runners; sexual-function and urinary scores no worse; in its exploratory model a bar level with or above the saddle aOR 0.8 (p = .10), no saddle type significant, and rare or no padded-shorts use went with lower odds of both; REF-1272 B- (41 competitive women): a bar below the saddle +3.47 kPa mean perineal pressure (95% CI 0.24-6.70) and less sensation at one site, but numbness in the last month 63.6% vs 63.2%; REF-1273 B- (the same women, pressure in 42): a cut-out went with 4.41 kPa higher mean and 11.54 kPa higher peak perineal pressure than a solid nosed saddle, and a wider saddle with lower peak perineal and mean total pressure | A bar level with the saddle; a saddle wide enough for the sit bones; test any saddle over real rides; a cut-out is not automatically better, and if one feels better that is her answer (practice) | That a bar, saddle or shorts change reduces numbness (pressure markers only, and REF-1269's own data support no set-up choice); padded shorts as protection (the survey went the other way, exploratory and cross-sectional) |
| Saddle sores | REF-1269 B+: sores in 39% of women cyclists, rising with lifetime miles, off-road riding aOR 1.9; REF-1268 B+ (abstract): a bar level with or above the saddle OR 0.6 for saddle sores in men; REF-1264 B+: 0.8% of reported injuries (likely under-counted on a form about muscles, tendons and joints, the entry's inference) | Clean, dry, well-fitting shorts; shorter unbroken seated spells; the bar level with the saddle (practice; hygiene and creams are practice) | Prevention from any product or set-up (no trial). A spreading, hot or feverish sore or a lump → clinician |
| Feet (burning, numbness, "hot foot") | No entry (REF-1033 is forefoot pressure in runners after a marathon) | Loosen straps; ask whether shoes, cleats or insoles changed and restore them; a cleat moved slightly back (practice) | Anything. Numbness that persists off the bike → clinician |
| "It started after a new bike, saddle or cleats" | REF-1265 B+: 31.3% of the surveyed riders said their back pain on the bike began after a change of set-up or a component (their belief, not a tested cause); REF-1260 B+: load, not measured geometry, had the moderate evidence | Restore the old set-up, or measure the old bike and copy it to the new one, then change from there one step at a time (practice) | That the new part is the cause; check what else changed in the riding |
| "Should I get a professional bike fit?" | REF-1262 B+ (abstract): 162 amateurs randomised to a 3D motion-capture fit or a posture handout; the authors report every outcome differed between groups after the intervention (p < 0.05); within the fit group comfort changed 3.12 points (FEEL scale), fatigue 3.95 (OMNI), groin pain 1.68 and back pain 1.35 (numeric scale); no handout-group change or between-group difference is given; unblinded self-ratings at 15 days; REF-1263 B (849 surveyed cyclists): a self-fit and a professional fit each went with OR 2.35 for riding without pain vs no fit, neither with fewer injuries; REF-1267 B: a fit packaged with advice and exercises, uncontrolled | Reasonable for comfort, or for a complaint that persists after the load change and one adjustment; a careful self-adjustment first is reasonable (inference from REF-1263); for pain, a clinician or a clinically trained fitter (practice) | Injury prevention; performance; that a 3D system adds anything over a careful static fit or a self-fit (never tested) |
Return to riding after an overuse problem
No study in the base; not from a study, the base's own practice: restart at roughly half the previous volume of the riding that hurt, easy first and intensity last, each step held while SYNTH-73's pain rule is met (SYNTH-73, SYNTH-36); flat, seated, at a self-chosen cadence, indoors where the load can be held steady; low-cadence and torque work back last (SYNTH-82); one equipment change at a time; the slowest ramp at a previously painful site (SYNTH-73).
The running cohorts' ramp numbers are not a rider's rule (SYNTH-29).
Key Findings — what the evidence shows
No measure of the bike or the body has been shown to predict overuse pain; load has moderate support. A systematic review found 18 observational studies (3 at low risk of bias) among 3,596 records: moderate evidence of a relationship between load and symptoms, moderate evidence of no relationship with many measures traditionally used in bike fitting, conflicting evidence for saddle height, and no strong evidence for any bike, body or load measure (REF-1260, B+, abstract; which measures and pain sites were tested is not in the abstract).
Associations measured at one time can run both ways: pain changes how much a rider rides (the entry's inference).
A fit makes riding feel better in the short term; nothing shows it prevents injury. In the one randomised trial, 162 amateurs had a 3D motion-capture fit or a posture handout; the authors report a between-group difference on every outcome after the intervention, but the mean differences they give (comfort 3.12, fatigue 3.95, groin pain 1.68, back pain 1.35) read as changes within the fit group, so the size of the fit's advantage is unknown; ratings were unblinded, follow-up 15 days, and a leaflet is not a sham fit (REF-1262, B+, abstract).
In 849 surveyed cyclists a self-fit and a professional fit went with the same higher odds of riding without pain (OR 2.35 each) and neither with fewer injuries; riders with pain may be the ones who seek a fitter (REF-1263, B, abstract).
Among amateurs surveyed about back pain, 96.5% of the 258 who had had a fit for pain or discomfort recalled some improvement, with no comparison group (REF-1265, the entry's arithmetic). After a physiotherapist-led package of fit, cadence and progression advice and exercises, worst knee pain fell 2.52 of 10 points at 12 weeks, again uncontrolled (REF-1267, B, abstract).
Saddle height: measure the knee angle while pedalling; small changes do little. A 41-study review (29,398 records screened) grades as strong that saddle height should be set from dynamic knee-angle measurement, as moderate that changes under 4% of leg length change lower-limb loads trivially to slightly and leave oxygen uptake and efficiency unchanged, and the injury evidence as limited (REF-1261, B+, abstract; no target angle in the abstract).
It updates a 2011 review by the same first author whose abstract recommends 25-30° of knee flexion without saying how it was measured (the full text is closed and was not read) and found a lower saddle seemed to raise patellofemoral force (REF-882, B+; REF-1076, C+, 6 men).
What amateurs report. In 62,758 consented entries to the Cape Town Cycle Tour (2016-2020) 1,879 gradual-onset injuries were reported (period prevalence 2.95%): knee 32.1%, shoulder 10.6%, lower back 7.8%, hip and buttock muscles 5.2%; 40.4% were severe enough to interfere with riding and 43.9% had lasted over a year; rest and physiotherapy were the commonest treatments and a bike set-up change was used for 15.1% (REF-1264, B+, self-reported, 78.9% of entries from men, 35.1% over 50).
In professionals the lower back was 45% of overuse injuries and the knee 57% of time-loss ones (REF-921, B+).
Low back: the back the rider brings comes first. 74.2% of the 581 riders who had had back pain on the bike had had it before they started cycling; back pain before cycling OR 4.20, coached or scheduled training OR 0.53, and a past bike-fit visit, weekly sessions and stretching dropped out of the model (REF-1265, B+, cross-sectional; two further odds ratios at p = 0.02 missed the authors' own threshold).
Lower bars flex the lumbar spine more, and riders with back pain sat more flexed with weaker back-extensor endurance — posture and muscle findings from 1-3 small studies each, all men, none testing a change of bar height or trunk training as treatment (REF-1266, B).
Hands: long rides change the nerve; padding and hand position lower the pressure. A deep ulnar motor branch slowed after a 6-day tour (REF-890, B); 23 of 25 riders had hand symptoms after 600 km (REF-1039, B-); the drops loaded the palm most and padded gloves cut that pressure by 10-28% (REF-1040, B).
Saddle area, men: off the front of the saddle protects perfusion; the erection question is unsettled. Standing restored penile oxygen pressure in six pooled laboratory studies (five sharing authors), no-nose saddles lowered perineal pressure, and a central cut-out did not lower total pressure (REF-1270, B+, healthy men, acute markers).
In a survey, standing more than 20% of the time and a bar level with the saddle went with less genital numbness, and cyclists' SHIM scores were no worse than swimmers' and runners' (REF-1268, B+, abstract); a meta-analysis found no difference in erectile dysfunction until age and comorbidities were adjusted for (OR 2.00) (REF-1271, B+, abstract, observational).
Saddle area, women: numbness and sores are common; set-up evidence is pressure only. Numbness and sores were each reported by 39% of women cyclists, with sexual-function and urinary scores no worse than in swimmers and runners; no handlebar or saddle-type item was significantly associated with numbness or sores, a level rather than nose-down saddle only at the borderline for numbness (aOR 0.7, p = .05) (REF-1269, B+).
In one group of 48 competitive women, a bar below the saddle went with higher perineal pressure but not with more numbness (REF-1272, B-), and a cut-out went with higher, not lower, perineal pressure (REF-1273, B-).
It does not show that any single set-up change prevents or relieves pain (a whole 3D fit lowered unblinded self-rated riding pain over 15 days against a handout, REF-1262), that a fit prevents injury, which saddle-height angle to aim for when measured while pedalling, anything about cleats, float, wedges, insoles, setback, neck pain or hot feet, or what women's saddle choices do to symptoms; those rows are practice or gaps (Evidence gaps).
Conflicts and how they are resolved
Practical Application — how the assistant says it
Evidence gaps
Limitations
How this entry was checked and changed
(the direction — load and the athlete's history before geometry, saddle height measured while pedalling with small changes doing little, a fit for comfort rather than prevention, and for saddle numbness time off the front of the saddle — rests on B+ systematic reviews read from their abstracts REF-1260 REF-1261, a B+ randomised trial read from its abstract with unblinded 15-day self-ratings REF-1262, B+ cross-sectional surveys REF-1264 REF-1265 REF-1268 REF-1269, a B+ review of laboratory markers in men REF-1270, a B+ observational meta-analysis REF-1271, B studies REF-1263 REF-1266 REF-1267 REF-890 REF-1040 and B- laboratory pressure studies in one group of 48 women REF-1272 REF-1273. Above C+ because that direction is what the B+ evidence says; below B because no trial tests any set-up change against pain incidence or injury, the one randomised trial measured comfort for 15 days against a leaflet, and every adjustment, step size, the order of questions and the two-week re-check are practice)