2.2.2. Grip Strength

Grip strength was measured using a Jamar dynamometer (Sammons Preston, USA) while the participant was in a seated position. The size of the grip span was adjusted according to the hand size of the participant, visually evaluated by the experimenter, and tested by the participant before they did the grip test. The experimenter first demonstrated the appropriate sitting position and how to hold and use the dynamometer with the elbow bent at 90 degrees, as recommended by the American Society of Hand Therapy (ASHT) [82]. A poster demonstrating the appropriate posture and arm/hand position was also visible to the participant (Figure 1A). The participant was asked to squeeze the dynamometer to their maximum ability for two seconds. Grip strength was measured (lbs) for the left and right hand and after a rest of approximately one minute, each hand was measured again. An average of both measures for each hand was used in the analysis. All staff and volunteers of the Me, Human project were trained to measure grip strength with the same protocol to limit measurement error.

**Figure 1.** Appropriate posture and arm/hand position for the grip strength on the poster showed to participant ( **A**) and the measurements taken on hand scans (**B**). W = hand width, L = hand length.

## 2.2.3. Hand Size and Shape

We measured hand size and shape from scans using a flatbed scanner (Epson Perfection V39). Participants were asked to place each hand palm side down, lining up their fingers to fit within an outline drawn on a transparent plastic sheet and keeping the fingers and palm flat (Figure 1B). Two differently sized outline transparencies for intermediate and large hands were available to allow a participant to best align their hand in a standardised manner (Figure 2). A 2 cm scale was included in each transparent plastic sheet to facilitate the accurate measurement of hand size from the scans. Hand size and shape were measured from each scan using freeware tpsDig2 software version 2.31 [83]. The hand width (W) was measured from the radial side of the second metacarpal joint to the ulnar side of the fifth metacarpal joint and hand length (L) from midline of the distal wrist crease to the tip of the middle finger (Figure 1B), following [59]. A ratio of hand width to length (W/L) was used as an indicator of hand shape, such as in [59]. We denoted hands with a ratio >0.5 as 'wide' hands and hands with a ratio <0.5 as 'long' hands. To correct the potential effect of hand size on grip strength, W\*L was used as an estimate of hand area to quantify the relative grip strength (i.e., grip strength/hand area). All measurements were taken by one researcher (KT) and to test interobserver error, 20% of both right and left hands (n = 266)

were measured by a second researcher (AB). An intraclass correlation coefficient (ICC) was used to test for interobserver reliability in R (R Core Team, 2020) with the "irr" package [84]. Both measurements of right and left hands were consistent between the two observers (ICC = 0.981, *p* < 0.0001), indicating an excellent reproducibility and repeatability of the measurements. Therefore, we only used the measurements of the first researcher for all subsequent analyses.

**Figure 2.** The two different sizes of outline transparencies (intermediate to the left and for large hands to the right) used on the flatbed scanner to allow the participant to best align their hand in a standardised manner. The scale of 2 cm was placed in the middle of the palm.
