To meet the requirements for high measurement accuracy, personalized fit, and high wearing comfort in hand biomechanics assessment, this study proposes a high-precision six-channel hand force measurement system. It integrates nine independently adjustable degrees of freedom, including 7 translational and 2 rotational. Through slide rail and rotary hinge mechanisms, it accommodates physiological variations in palm length, palm width, and finger length across subject populations. To alleviate tenderness under high-intensity pressing, the contact interface employs flexible contact units 3D-printed with thermoplastic polyurethane. Finite element simulation using a second-order Mooney-Rivlin hyperelastic model shows 1.422 mm deformation under 100 N axial load, demonstrating excellent cushioning performance. The system integrates six strain-gauge pressure sensors with a 0~200 N range, supporting 100 Hz sampling, and maintains a maximum absolute error of less than 0.5 N within the 0~100 N range. Based on this platform, 53 healthy volunteers and 5 stroke patients are recruited for maximum voluntary contraction （MVC) and 30% MVC constant force holding experiments. This study employs a normalized coupling matrix of relative finger output forces to quantitatively characterize inter-finger enslaving effects, and uses normalized force standard deviation to evaluate steady force control capability. Results show: In healthy individuals, the thumb exhibits highest independence, while the three ulnar fingers demonstrate strong coupling; with aging, the inter-finger coupling matrix becomes denser and the normalized force standard deviation increases; in the stroke group, MVC decreased by 60%~70%, presenting a globalized co-activation pattern, with normalized force standard deviation approximately three times that of the young and middle-aged groups. Results show that the device has significant potential in hand strength monitoring and neuromuscular function evaluation.