Diabetic foot represents a severe condition characterized by infection, ulceration, or tissue destruction in the foot of individuals with diabetes, often concomitant with lower extremity neuropathy, peripheral artery disease (particularly lower extremity arterial disease) (Jodheea-Jutton et al., 2022), or Charcot foot. According to the comprehensive analysis conducted by Zhang et al. (2017), the aggregated global prevalence of diabetic foot ulceration stands at 6.3% (95% confidence intervals: 5.4%–7.3%). This prevalence is on a continual rise, mirroring the escalating numbers of individuals afflicted by diabetes and those at risk, thereby establishing the diabetic foot as a burgeoning global public health challenge (van Netten et al., 2020; Edmonds et al., 2021).

The surge in diabetic foot cases has intensified the demand for effective prevention and treatment strategies. Diabetic foot ulcers (DFU), infections, or soft tissue damage often evade early detection, particularly in neuropathic feet, as the typical indicators of local infection, such as redness, pain, or tenderness, may be absent (Edmonds et al., 2021). Complicating matters, diabetic patients with deep foot infections may not exhibit traditional signs like an elevated white blood cell count and temperature, leading to delayed diagnosis and treatment (Apelqvist, 2012; Yu et al., 2022). Common imaging tests prove effective primarily in diagnosing diabetic foot conditions in the presence of bone alterations, abscess formation, and soft tissue destruction, missing the optimal window for intervention (Senneville et al., 2020; Yufeng and Yanwu, 2021).

Treatment and care pose challenges, including intricate cures, high recurrence rates, and elevated treatment costs (Kerr et al., 2019; Edmonds et al., 2021). The 5-year mortality and direct care costs for patients with diabetic foot complications rival those of cancer (Armstrong et al., 2020). Lack of awareness about prevention further compounds the issue, making treatment more challenging once diagnosed. In the initial stages of diabetic foot, lower limb pain during walking may be alleviated or disappear after resting, leading high-risk groups to neglect prevention services, increasing the likelihood of severe ulcers or amputations (Chatwin et al., 2020). Therefore, early prevention and intervention during the initial stages of disease development are crucial for diabetic foot patients (Lim et al., 2017). The utilization of specialized therapeutic insoles emerges as a commendable preventive measure (Schaper et al., 2020). These insoles are widely employed to mitigate plantar pressure peaks or optimize pressure distribution, effectively reducing the likelihood of complications and amputations (Korada et al., 2020). Nevertheless, challenges persist in further optimizing pressure offloading performance, enhancing comfort, and streamlining the design and manufacturing cycle and cost of customized insoles (Collings et al., 2021).

In the realm of diabetic foot insoles, notable advancements in design approaches and manufacturing techniques have brought about significant transformations, leading to a marked improvement in their performance, as illustrated in Figure 1. The triad of design, manufacturing, and performance evaluation stands as a critical nexus within the domain of diabetic foot insoles, with each facet exerting a profound influence on the others. Advanced manufacturing technologies now facilitate the creation of diabetic foot insoles boasting intricate structures, including honeycomb porous configurations and triply periodic minimal surface (TPMS) structures. These elaborate structures encounter challenges in precision and production costs when subjected to conventional manufacturing methods such as hand forming or numerical control techniques. Concurrently, the evaluation of diabetic foot insole performance serves as a driving force behind advancements in manufacturing and design. Thus, the current study offers a comprehensive review of the design, manufacturing, and performance evaluation of diabetic foot insoles, aiming to propel further breakthroughs in this dynamic field.

The development of diabetic foot is not solely linked to neuropathy and changes in blood glucose but is also influenced by biomechanics (Asghar and Naaz, 2022). The plantar pressure distribution in healthy individuals versus diabetic foot patients differs significantly, as depicted in Figure 3A (Deschamps et al., 2013; Fernando et al., 2016; Bus et al., 2021). Moreover, variations exist in pressure conditions across different plantar areas, as illustrated in Figure 3B (Perren et al., 2021). Ground pressure, excessive plantar shear stress (Jones et al., 2022), and compressive stress from shoes or adjacent toes collectively impact diabetic foot development (Allan et al., 2016). Designing specialized insoles for diabetic foot patients becomes imperative due to these complexities. Diabetic foot insoles can be broadly categorized into two groups: pressure offloading insoles and smart monitoring insoles.

Traditional customized foot orthotics have historically been crafted through subtractive manufacturing processes, involving the carving or milling of materials to achieve the desired orthotic shape (Newman et al., 2015). This method, however, tends to be time-consuming, labor-intensive, and generates waste through excess material removal (Paris et al., 2016). In contrast, additively manufactured foot orthoses offer advantages such as high precision, customization, rapid production, and environmental sustainability. These attributes contribute to enhanced biomechanical control, comfort, risk reduction for pressure points, and shortened delivery times, ultimately expanding patient access to customized orthotics (Jin et al., 2015; Chen et al., 2016; Paris et al., 2016).

As a burgeoning technology, additive manufacturing (AM) is gradually finding application in the production of diabetic foot insoles (Srivastava and Gaur, 2020). AM constructs three-dimensional objects by incrementally adding thin layers of materials guided by digital models (Wong and Hernandez, 2012). Its adoption streamlines production processes with short development cycles, and it is extensively utilized in the biomedical field (DebRoy et al., 2018). There are various additive manufacturing (AM) technologies, encompassing Selective Laser Sintering (SLS), Selective Laser Melting (SLM), Stereolithography (SLA), Fused Deposition Modeling (FDM), and 3D Printing (3DP) (Chen et al., 2016; Abdulhameed et al., 2019). Notably, FDM and SLS predominantly manufacture customized insoles for diabetic foot patients, as depicted in Figures 8A, B. FDM is the most commonly used technology for crafting customized insoles for diabetic feet, creating three-dimensional structures by depositing thermoplastic material onto a substrate in layers through a temperature-controlled printhead (Winarso et al., 2022). Research by Davia-Aracil et al. (2018) comparing traditional subtractive techniques to FDM for manufacturing anatomical insoles found FDM to be 43% more cost-efficient and time-effective. Ma et al. (2019) employed material extrusion techniques, including FDM and fused filament fabrication (FFF), to print adjustable modulus porous structures for diabetic foot insoles, demonstrating their mechanical effectiveness. Peker et al. (2020) implemented an upgraded FDM system to manufacture diabetic foot insoles with reduced production time.

While FDM is the most economical technique for creating entirely geometric customized insoles (Vyavahare et al., 2019), it has limitations. Interlayer bonding characteristics in FDM result in printed parts with highly anisotropic mechanical properties, potentially leading to early failure (Rankouhi et al., 2016; Lim et al., 2021). Additionally, materials used in FDM may exhibit suboptimal durability and mechanical properties (Hudak et al., 2022).

Another AM technology used for customized diabetic foot insoles is SLS, an industrial 3D printing method utilizing a powder bed to construct 3D objects (Fina et al., 2017). SLS involves selectively melting a thin layer of powdered material on a build platform according to a 3D model using a laser. This process is repeated layer by layer until the final object is formed (Mazzoli, 2013), allowing the production of complex structures with superior dimensional accuracy and strength (Awad et al., 2020). Li et al. (2022) highlighted SLS’s ability to successfully fabricate the unit array design beneath the insole’s lower surface. In comparison to FDM, SLS exhibits greater potential for manufacturing insoles with intricate structures. Table 5 provides a comparative overview of AM methods for diabetic foot insoles.

The evaluation of diabetic foot insoles encompasses both objective and subjective assessments. Objectively, evaluations focus on plantar pressure distribution analysis, gait analysis, and electromyography, primarily examining the pressure offloading performance of diabetic foot insoles (Bellomo et al., 2022). Prior to clinical application, a crucial step involves the performance evaluation of customized insoles for diabetic feet, ensuring accurate and efficient feedback to enhance the design (Anderson et al., 2020; Song et al., 2022).

The in silico method, also known as the numerical method, employs computational models to numerically evaluate a product’s performance (Anderson et al., 2021). Once developed and validated, it efficiently assesses the pressure-offloading performance of customized insoles (Peng et al., 2021). Unlike conventional experiments that involve human participants, in silico approaches, such as FE modeling, offer a more precise alternative by isolating the effects of specific insole design variables without the need for physical experiments on human subjects or models (Behforootan et al., 2017). However, the current challenge lies in developing valid computational models. Musculoskeletal models of the foot and foot-insole FE models have become common computational tools for evaluating diabetic foot insoles (Keenan et al., 2022). The musculoskeletal model is crafted to emulate human movement by integrating rigid bones, muscle fibers, tendons, ligaments, and other connective tissues crucial for understanding and predicting human activities (Falisse et al., 2019). Utilizing these models allows for an assessment of how insole design parameters impact everyday gait. Chen et al. (2015) introduced a 3D FE musculoskeletal model of the foot, providing a tool to evaluate the pressure-offloading performance of insoles made from different materials and thicknesses. Similarly, Jafarzadeh et al. (2021) employed a foot-insole FE model to scrutinize the influence of insole stiffness on shape optimization outcomes. Despite their utility, developing musculoskeletal models remains challenging in the current stage of research. These models are simplified representations, and as such, the boundary conditions derived from them may lack accuracy and appropriateness. The complexity of accurately capturing the intricacies of human biomechanics poses ongoing challenges in refining musculoskeletal models for robust and reliable evaluations of insole designs.

The in vitro evaluation method employs laboratory testing to assess the performance of diabetic foot insoles (Zhang et al., 2019). This testing utilizes a pressure measurement system to gauge the plantar pressure of subjects during both gait and standing (Burns et al., 2019; Anzai et al., 2022). The experimental results are then compared with FE simulation results to evaluate the pressure-offloading performance of the insoles. In its early stages, the F-Scan system served as the primary pressure measurement system; however, it encountered issues of poor accuracy and repeatability (Lorkowski et al., 2021). Subsequently, with advancements in science and technology, the Pedar system, renowned for its superior performance, has become widely adopted for pressure measurement (Yamamoto et al., 2020). Numerous researchers have conducted pressure tests on diabetic foot insoles using the Pedar system. Nouman et al. (2021), for instance, utilized the Pedar system to measure peak plantar pressure in diabetic foot patients while maintaining balance on level ground. The experimental average peak plantar pressure was then compared with the peak contact pressure obtained from FEA for validation. Tang et al. (2019) employed the Pedar-x insole pressure measurement system (Novel Pedar System, Germany) to measure contact pressure between feet and insoles during static standing with arms akimbo, assessing the insoles’ performance. D’Amico et al. (2021) utilized the Pedar-X system to measure in-shoe dynamic plantar pressure during walking in multiple trials along a 15 m walkway in a laboratory setting. Shi et al. (2022) obtained pressure values using the Pedar insole measurement system (Novel GmbH, Munich, Germany) to evaluate four different insoles during walking. Participants self-selected walking speeds to exhibit their typical gait characteristics on a 10 m straight sidewalk while barefoot and wearing insoles. A summary of in vitro performance evaluations of diabetic foot insoles is presented in Table 6.

Subjective evaluation primarily involves collecting patient feedback on comfort, compliance, and overall satisfaction, typically obtained through randomized trials and questionnaires. Lavery et al. (2012) conducted a randomized trial with 299 patients, assessing standard care versus reduced insoles over 18 months. Patient questionnaires revealed insights into compliance and durability, impacting the effectiveness of shear-reducing insoles in preventing diabetic foot ulcers. In another study, Ming et al. (2019) conducted a 24-month trial evaluating sensor-equipped insoles for measuring foot temperature in 300 participants with diabetes and severe peripheral neuropathy. While objective measurements offer quantitative data, they may require specialized equipment or expertise. In contrast, subjective measurements provide qualitative data and may be more manageable. Combining both approaches could yield a more comprehensive assessment of diabetic insole effectiveness. A comparison of the advantages and disadvantages of each evaluation method is detailed in Table 7.

In summary, customized insoles for diabetic individuals exhibit significant potential in effectively offloading plantar pressure, contributing substantially to early prevention, pain reduction, and economic burden alleviation. The continuous evolution of smart insoles for diabetic foot care is paving the way for an improved remote digital healthcare system (Ahmed et al., 2020). The ongoing advancement of smart insoles for diabetic foot care is paving the path toward an enhanced remote digital healthcare system. Nonetheless, despite the relatively comprehensive research on diabetic foot insoles, numerous limitations still require further exploration. Consequently, existing research encounters several challenges, and the following issues are underscored, along with corresponding suggestions for the design and implementation of diabetic foot insoles.

(1) Biomechanical Correspondence: Future studies must comprehensively consider the biomechanical correspondence between various tissues in the human foot, including bones, muscles, and fascia (Allan et al., 2016; Asghar and Naaz, 2022). Evaluating joint angles, muscle activity, and skeletal stress and strain is crucial for effective studies on diabetic foot, offering insights into foot motion, stability, muscle coordination, functionality, and predicting disease risks.

Based on the challenges highlighted earlier, several future research directions for the advancement of diabetic foot insoles can be outlined. These include:

(1) Reverse Design with Machine Learning: Utilizing machine learning for reverse design of insoles tailored to different levels of Diabetic Foot Ulcers (DFU). This involves exploring data characteristics, model training for parameter optimization, and selecting optimal insole designs, significantly reducing the overall design cycle.

This review has provided a comprehensive overview of the design, manufacturing, and performance evaluation of diabetic foot insoles. The design phase often employs iterative optimization methods rooted in biomechanics and FEA, allowing for the prediction of design parameters associated with insoles that offer superior pressure offloading effects. The discourse also delves into the evolving landscape of smart insoles designed specifically for diabetic foot care. In the realm of manufacturing, while Fused Deposition Modeling (FDM) stands out for its cost-effectiveness and efficiency, its limitations in printing complex structures necessitate the use of Selective Laser Sintering (SLS) for intricate insole designs. Insoles featuring total contact customization, softer and absorbent materials, elliptical porous structures, or TPMS structures are deemed more fitting for diabetic foot prevention, taking into account both pressure offloading performance and comfort. Prior to clinical application, a thorough evaluation of insole performance through in silico and in vitro testing approaches is crucial. Despite the strides made in exploring various facets of diabetic foot insoles, there are avenues for further research. Future investigations could explore inverse design strategies leveraging machine learning, use topology optimization techniques for lightweight insole development, integrate flexible sensors for enhanced functionality, and delve into the development of novel skin-like materials tailored specifically for diabetic foot insoles. These advancements hold promise for refining the efficacy, personalization, and innovation of diabetic foot care technologies.