Grip Dynamometer Position Is Skewing Your Readings — Here's the Standardized Fix

A physio retests an athlete's grip strength on a Monday morning and gets 38kg on the right hand, down from 46kg logged three weeks earlier. Nothing in the training log explains an 18% drop: no injury, no soreness, no missed sessions. The explanation shows up when someone pulls the intake photo from the first session and compares it against Monday's setup. The dynamometer handle sat two notches wider the first time, and the athlete's elbow was propped on the plinth instead of hanging free at 90 degrees. Nothing about the hand changed between those two sessions. The test did.

That gap is not really a measurement error in the sense of a broken instrument. The dynamometer reported exactly what it was handed: a hand producing less force from a different, weaker point on its own length-tension curve, stabilized by a shoulder and elbow not doing the same job twice. What follows is what handle span and joint position actually do to a reading, the studies that put numbers on it, and a standardization protocol that turns a suspicious grip strength change back into a claim worth standing behind.

How Handle Width and Joint Angle Actually Change the Number

Grip strength is not a fixed property of a hand the way height is a fixed property of a body. It is the output of a lever system, the finger flexor tendons pulling across a handle set to whatever span it happens to be sitting at, and every one of those muscles has a length at which it generates the most tension, with force falling off on either side of that optimum. Squeeze a handle set narrower than a hand's own optimal span and the flexors start the contraction already shortened, producing less force than the same hand could generate at its best span. Set it too wide and the fingers cannot close around it far enough to load the tendons efficiently either. A Jamar-style dynamometer offers five handle positions, running roughly from about 3.5cm of span at position I to roughly 7.6cm at position V, and which one produces a given hand's true maximum depends on hand size and finger length, a detail most retest protocols never write down.

Elbow and shoulder position add a second variable that has nothing to do with the fingers at all. The commonly used standard position, seated with the shoulder adducted, elbow flexed to 90 degrees, forearm neutral, wrist in slight extension, was adopted for reproducibility across clinics, not because it pulls a hand's single highest reading. An extended elbow, or a shoulder braced against a table edge instead of hanging free, changes how much the trunk and shoulder girdle can contribute to stabilizing the arm during a maximal squeeze, and that shows up as a different dial reading even though nothing in the hand itself moved.

What the Research Actually Shows

Mathiowetz, Kashman, Volland, Weber, Dowe, and Rogers (1985), in Archives of Physical Medicine and Rehabilitation, built the normative grip-strength tables most clinics still test against: 0-30 degree wrist extension, 90-degree elbow, adducted shoulder, handle position II on a Jamar dynamometer, across several hundred adults. They chose that position not because it maximizes output, but because it could be reproduced consistently across examiners and sites. Their norm tables are valid only for that exact position at that exact handle setting; test an athlete at position III instead, or with the elbow extended, and the number is not wrong, just no longer comparable to those published norms. It is a different test wearing the same units.

Firrell and Crain (1996), in the Journal of Hand Surgery, tested grip strength across all five Jamar handle positions in a sample of adult hands and found no single position produced the maximum reading for every hand. The position that maximized output varied with hand size, and for many hands, testing at a non-optimal position rather than their own best span cost a meaningful chunk of recorded force, with within-subject swings on the order of 10-20% or more between a hand's best and worst position. The finding argues for holding handle position constant between a test and its retest even more than it argues for one universal best setting. Their sample came from a hand-clinic population on a Jamar hydraulic device only, so exact percentages will not transfer identically to a digital dynamometer or a young athletic population.

Su, Lin, Chien, Cheng, and Sung (1994), in Archives of Physical Medicine and Rehabilitation, tested grip strength across combinations of elbow position, extended versus flexed to 90 degrees, and shoulder position, adducted, flexed forward, or abducted, in healthy adults. Grip strength came out measurably higher with the elbow extended than with the ASHT-standard 90-degree flexion, and higher again with the shoulder adducted than abducted or raised in front, with differences in the high single digits to low double-digit percentage range between the strongest and weakest combinations tested. Their subjects were healthy young adults tested seated, so the exact magnitude may not hold for a standing protocol, an older population, or a shoulder that cannot fully adduct after injury.

Equipment and Test Setup

None of this needs new equipment, just a way to confirm and record the same setup every time, since the problem above comes from drift between sessions, not an inaccurate dynamometer.

Keep the chair constant too: armrests let an athlete brace the forearm mid-squeeze, adding leverage the standard protocol assumes is not available.

Step-by-Step Standardization Protocol

1. Seat the athlete in an armless chair with back support, feet flat, hips and knees at roughly 90 degrees.
2. Set shoulder position adducted and neutrally rotated, arm hanging at the side, not braced against the torso or a table edge.
3. Set elbow position flexed to 90 degrees, confirmed with a goniometer rather than by eye; an elbow drifting toward extension is the single most common source of an inflated reading.
4. Set forearm and wrist forearm neutral, thumb up, wrist held between 0 and 30 degrees of extension and 0 to 15 degrees of ulnar deviation.
5. Fix the handle span at position II for comparison against published norms, or at the athlete's own previously logged optimal position for tracking that individual over time. Either choice is defensible; changing it silently between sessions is not.
6. Deliver the same cue word for word, at the same volume and timing, every trial. Coaching mid-squeeze inflates later trials relative to earlier ones tested without it.
7. Run three maximal trials per hand, alternating hands, 15 to 30 seconds of rest between efforts on the same hand. Record the peak or the mean, and keep that choice consistent across every future retest.
8. Log the full setup, not just the number: handle position, elbow angle, chair, cue, and time of day, saved alongside the score so a retest six weeks later starts by matching that record instead of guessing.

The whole sequence runs 6 to 8 minutes per athlete including setup, most of it spent confirming position before the first squeeze rather than in the squeezing itself.

Reading a Change: Real Strength Shift or Setup Drift?

Grip strength has a known day-to-day technical error of measurement, commonly cited around 4 to 5kg or roughly 8-10% for a well-controlled test, even with position held constant. Anything smaller than that is normal test noise. What matters is checking the logged setup before assuming a bigger swing is real.

Mistakes That Undo the Standardization

Making This Part of Standard Practice

Pick one standard, write it down, and post it where every examiner testing that athlete can see it: shoulder adducted, elbow at 90 degrees confirmed with a goniometer, forearm neutral, wrist in slight extension, handle at a fixed and logged position. Which exact position you choose matters less than whether every examiner uses the same one and records it every time.

Treat a photo or short video of the setup as part of the test record, not an optional extra, the same way you would log the surface and cleat choice for a sprint test. When a number looks off, check that record before checking the athlete. A change inside the roughly 8-10% technical error of measurement this article's cited research points to is normal noise; a change tied to a documented shift in handle position or elbow angle is a setup problem with a five-minute fix, not a training or injury conversation waiting to happen.