Footwear comfort: a systematic search and narrative synthesis of the literature

Most studies were laboratory-based, repeated measures designs where comfort was measured under different footwear and/or insole conditions [14, 15, 18, 20‐23, 25‐30, 33, 35‐39, 41, 43‐48, 50‐55, 57, 58, 60‐71, 73‐79, 81‐87, 89, 90, 93, 97, 98, 100‐105], but there were also 13 surveys [1, 3, 9‐11, 16, 19, 24, 40, 56, 88, 92, 96], eight clinical trials [12, 34, 42, 49, 59, 94, 95, 99], three qualitative studies [13, 17, 91] and three reviews [5, 7, 72]. Sample size ranged from 5 to 1524, and primarily included healthy young adults [10, 14, 15, 17, 18, 22, 28, 31, 32, 35, 37, 43‐46, 48, 50, 53, 54, 56, 58, 60, 64, 67, 74‐76, 78, 79, 83, 88, 90, 101‐104], but also children [71, 91], older people [1, 52, 63, 76], participants with medical conditions (such as diabetes [16, 27, 68, 105], rheumatoid arthritis [29], patellofemoral pain [73], plantar fasciitis [99], hallux valgus [93] and non-specific musculoskeletal disorders/symptoms [25, 95]), specific occupational groups (such as military personnel [42, 69, 85, 89], factory workers [9, 12], school teachers [11], kitchen staff [13], hospital staff [13], coal miners [40, 41] and police officers [94]) and sportspeople (such as runners [26, 30, 36, 39, 47, 49, 65, 66, 70, 77, 82, 84, 86, 98], basketball players [61, 62, 92, 100], soccer players [38, 87, 97], cyclists [20, 21], aerobic dancers [34], skiers [51], rugby players [59], people attending gymnasiums [3], badminton players [55] and tennis players [96]).

No studies provided a specific definition of comfort, although four studies were designed to explore how comfort is conceptualised. Alcantara et al. [10] developed a list of 74 adjectives related to footwear design and manufacture and asked 67 people to evaluate 36 shoes using these adjectives on a 5-point scale. Principal components analysis demonstrated that perception of casual footwear could be described on the basis of 20 independent concepts, two of which pertained to comfort. The first was characterised as ‘pure comfort’ and included the positive adjectives good fitting, soft, comfortable, flexible, light, relaxing, smooth, and the negative adjectives rough, hard, strong, heavy, rigid, and robust. The second was characterised as ‘thermal comfort’ and included the positive adjectives fresh, light, breathable, and the negative adjectives hot, heavy, and safe. Similar findings were reported in a qualitative study of footwear comfort perceptions of standing workers by Anderson et al. [13], who found that positive adjectives used were cushioning, arch support/contour, breathability/ventilation and negative adjectives used were hardness, heaviness and heat. A study of younger women’s perceptions of dress shoes identified ten criteria which differentiated between comfortable and uncomfortable shoes, the strongest being absence of pain, feeling, sound, and texture [17]. Finally, in a qualitative study of children, the adjectives soft and padding were most frequently used to describe comfortable shoes, while hard, tight, loose and heavy were used to describe uncomfortable shoes [91].

A wide range of measurement tools have been used to quantify comfort, including simple dichotomous responses [11, 12, 29, 94], ranking footwear conditions in order of preference [14, 31‐33, 36, 47, 48, 52, 60, 63‐65, 76, 77, 79, 81], 4-point [42, 80], 5-point [37, 40, 45, 58, 70, 76, 97], 6-point [59], 7-point [47, 64, 79, 87], 9-point [57] and 12-point [41] Likert scales, 10-point numerical rating scales [43, 78], and 100 mm [15, 16, 18‐21, 27, 28, 30, 35, 39, 48, 49, 53, 56, 63‐65, 67, 68, 71, 73, 74, 77, 79, 82, 85, 89, 90, 93, 99, 104, 105], 150 mm [22, 25, 26, 36, 46, 50, 55, 61, 62, 66, 69, 75, 83, 84, 98, 100] and 170 mm [44] visual analog scales. The anchor statements indicating the lowest possible comfort score included ‘not comfortable at all’ [19‐22, 25, 26, 30, 39, 49, 50, 53, 55, 56, 61‐65, 73, 79, 82‐85, 89, 93], ‘very uncomfortable’ [27, 28, 40, 41, 45, 46, 58, 68, 74, 78, 79, 98], ‘least comfortable’ [44, 76, 80, 90], ‘extremely uncomfortable’ [48, 59, 89], ‘not comfortable’ [71, 75], ‘not at all comfortable’ [15, 18], ‘not acceptable’ [37, 70], ‘totally disagree’ [10], ‘least comfortable imaginable’ [35], ‘not satisfactory’ [42], ‘very bad comfort’ [43], ‘very, very low’ [47], ‘minimum comfort’ [69], ‘maximal pain/discomfort’ [99], ‘not very comfortable’ [100], ‘completely uncomfortable’ [104], ‘extremely bad’ [57] and ‘unbearable discomfort’ [87]. The anchor statements indicating the highest possible comfort score included ‘most comfortable imaginable’ [19‐22, 25, 26, 30, 35, 39, 46, 50, 53, 56, 61, 63, 65, 73, 79, 82‐84], ‘very comfortable’ [18, 27, 28, 40, 41, 49, 58, 68, 71, 74, 75, 78, 79, 98, 100], ‘most comfortable’ [44, 55, 62, 76, 80, 85, 90], ‘extremely comfortable’ [48, 59, 87], ‘just right’ [37, 70], ‘totally agree’ [10], ‘very much’ [15], ‘excellent’ [42], ‘very good comfort’ [43], ‘not at all uncomfortable’ [45], ‘very, very high’ [47], ‘maximum comfort’ [69], ‘maximal comfortable’ [64], ‘no pain/discomfort’ [99], ‘completely comfortable’ [104], ‘most conceivable comfort’ [93] and ‘extremely good’ [57]. Most studies documented an overall comfort score for the whole foot/shoe, while others reported separate comfort scores for specific regions of the foot/shoe [19, 21, 22, 36, 57, 59, 64, 66, 71, 73, 75, 83, 87, 104]. The vast majority of tools considered comfort to be a unidimensional construct, although some incorporated additional perceptual components including in-shoe ‘climate’ [55], thermal comfort [51, 67, 103], dampness [67] and air permeability [67].

Seven studies specifically addressed the psychometric properties of comfort scales. Mündermann et al. [83] assessed the reliability of 150 mm visual analog scales documenting comfort pertaining to forefoot cushioning, heel cushioning, arch height, heel cup fit, shoe heel width, shoe forefoot width, and shoe length in runners wearing standardised running footwear with four inserts of differing hardness. Overall, intra-test repeatability was high (intraclass correlation coefficient [ICC] 0.80 and improved with four to six repeated sessions, although it was noted that some participants reported highly variable comfort ratings. A subsequent study by these authors demonstrated consistency of scores using this scale with repeated sessions over 3 weeks [84], and Lam et al. [62] confirmed the reliability of the same 150 mm visual analog scale in basketball players (ICCs from 0.61 to 0.80). Mills et al. [79] compared the reliability of documenting comfort (overall, cushioning of the forefoot, arch and heel, and support of the arch and heel) in participants wearing their usual footwear and four different inserts using 100 mm visual analog scales, 7-point Likert scales and ranking, and found that ranking was the most reliable measure, followed by the visual analog scales and Likert scales. Similarly, Lindorfer et al. [64] assessed 30 runners over six repeated sessions, and found that ranking provided the highest reliability (Pearson’s r = 0.07), followed by a 100 mm visual analog scale (r = 0.67) and 7-point Likert scale (r = 0.63). More recently, Bishop et al. [19] reported a detailed psychometric evaluation of a new running shoe comfort assessment tool incorporating four components measured with a 100 mm visual analog scale (heel cushioning, forefoot cushioning, shoe stability, forefoot flexibility and an overall comfort score). Reliability of the overall score was excellent (ICC 0.88) and good for each of the component scores (ICC > 0.70).

In contrast to these positive findings, Hoerzer et al. [50] examined intra-rater reliability of 150 mm visual analog scales and dichotomous (yes/no) ratings of insole comfort, and found that less than a third of participants provided reliable scores across the two sessions. These findings suggest that psychological factors, such as mood, may influence the perception of comfort and that documenting a mean score across multiple sessions may be necessary to obtain acceptable reliability. Furthermore, a recent systematic review of comfort scales by Matthias et al. [72] demonstrated that few studies explicitly evaluated validity, and many exhibited methodological bias, such as lack of participant and assessor blinding.

Eight studies evaluated perceived comfort while wearing footwear that varied according to midsole cushioning, including running shoes [39, 60, 78, 90, 98], basketball shoes [61], casual shoes [63] and military boots [85]. All but one study [63] reported that participants found the footwear with softer midsole materials to be more comfortable, although Sterzing et al. [95] also demonstrated that the use of harder materials under the forefoot did not negatively affect comfort provided that the material under the rearfoot was soft. However, documentation of midsole cushioning across these studies was inconsistent, with some studies using no objective measures [60, 61, 78] and others reporting either density (which ranged from 0.15 to 0.24 g/cm³) [39, 90] or hardness (which ranged from Shore A 25 to 66) [63, 85, 90]. Furthermore, only three of these studies [61, 63, 98] specifically manipulated midsole cushioning while controlling for other shoe characteristics, making it difficult to attribute differences in comfort to cushioning alone.

Nine studies examined the effects of heel elevation on comfort perception while wearing high-heels [28, 53, 54, 56, 101], casual shoes [75, 76] and running shoes [23, 82]. In the two running shoe studies, heel elevations (i.e., heel-toe drop) ranged from 0 to 15 mm, but no differences in comfort were found during treadmill walking [23] or running [82]. In contrast, significant reductions in comfort were consistently reported in each of the studies examining high-heels [28, 53, 54, 56] and casual shoes [75, 76] when heel height was increased by between 45 and 76 mm. However, comfort perception was higher in participants who were habituated to wearing high heels [28], and three studies reported that the discomfort associated with heel elevation could be partly ameliorated by the use of heel cups [56] and arch supports [53, 54, 56]. An interaction between heel height and the sagittal plane angle of the heel seat (‘wedge angle’) was also reported by Witana et al. [101], who found that comfort while wearing high heels could be optimised by selecting the most appropriate wedge angle for a corresponding heel height.

Five studies evaluated the influence of shoe weight on perceived comfort in military boots [85, 89], safety footwear [18], running shoes [60] and basketball shoes [62]. However, only three studies provided objective measurements of shoe weight (ranging from 335 to 800 g) [62, 85, 89] and none controlled for other shoe characteristics. Nevertheless, all reported that the lightest shoe was perceived to be the most comfortable. In addition to the absolute weight of the shoe, the distribution of mass may also be important. Chiu et al. [33] added weights to different locations of casual canvas shoes while keeping the total weight constant, and found that most participants preferred rear-weighted shoes and perceived them to be lighter than when the weights were added distally.

Four studies examined the effect of sole flexibility on comfort: two in running shoes [37, 78], one in coal mining boots [41], and one in people with diabetes [105]. In running, comfort related to sole flexibility may depend on running speed. Miller et al. [78] compared comfort while standing, walking and running in participants wearing three different commercially-available running shoes, and found that although comfort ratings differed according to the activity, the most comfortable shoe on average was the most flexible. However, Day et al. [37] tested participants in shoes with and without carbon fibre plates and found that participants preferred the standard shoe when running at 14 km/h, but the stiffer shoe at 17 km/h. The coal mining boot study assessed comfort when coal miners wore two standard boots (one with a stiff shaft and one with a flexible shaft), and each boot was then modified to create a more flexible sole by cutting slits across the sole at the level of the metatarsophalangeal joints. Although most participants preferred the boot with a flexible shaft combined with a stiff (unmodified) sole, there was large variability in the comfort scores and no significant effect among the different boot types [41]. Finally, Zwaferink et al. [105] found that adding a 3 mm carbon-fibre stiffening insert to extra-depth shoes in people at risk of foot ulceration resulted in lower plantar pressures, but had no detrimental effect on comfort.

Seven studies evaluated the influence of various aspects of midsole geometry on comfort, including offloading/rocker-sole footwear (in healthy individuals [43, 66], older people [52], people with diabetes [27, 68] and people with rheumatoid arthritis [29]), and the application of lateral wedges to footwear for the treatment of knee osteoarthritis [44]. In the offloading footwear studies, rocker-sole shoes were perceived to be more comfortable than standard footwear [29] or forefoot offloading shoes (i.e., shoes with no ground contact at the forefoot designed to avoid toe-off) [27, 43], and rigid rocker-sole shoes more comfortable than semirigid rocker-sole shoes [68]. However, no substantial differences in comfort were found across three different heel curvature designs (short-parallel, long-parallel and oblique) in running shoes [66], adding foam to the plantar midfoot region of the outersole [52], or following the addition of a small (2 degree) lateral heel wedge designed to alter knee joint moments when walking [44].

Three studies evaluated the effect of outsole geometry on comfort [38, 59, 87]. In soccer boots, de Clerq et al. [38] assessed the effect of various stud configurations on comfort when performing cutting manoeuvres, and found that the sole design with the least number of studs was the most comfortable. Kryger et al. [87] compared two soccer boot designs that varied according to stud shape, upper material and boot mass, and found that the most comfortable boot generated lower pressures under the first and fifth metatarsal heads. Similarly, in rugby, Kinchington et al. found that hybrid turf shoes were more comfortable than studded boots [59].

Two studies evaluated the effect of lacing on comfort. Dobson et al. [40] reported lace-up boots to be more comfortable than slip-on boots in coal miners, while Hagen et al. [58] compared different styles of lacing in seven-eyelet running shoes, and found that comfort varied according to the number of eyelets laced and how tightly the laces were tied. The most comfortable technique involved the use of three eyelets laced while keeping the upper two eyelets unlaced, although this was perceived to be less stable and was associated with higher pronation velocity while running compared to tightly lacing all eyelets.

Four studies demonstrated that shoes with more compliant upper materials may be preferable. Jordan et al. [58] compared three styles of casual footwear and found that shoes generating lower pressures on the dorsum of the foot were perceived as more comfortable, while Melvin et al. [75] found that shoes constructed from soft suede were more comfortable than shoes constructed from a stiffer leather upper. In soccer boots, Sterzing et al. [97] found that despite two models having identical stud configurations, the model with the softer heel counter was perceived to be more comfortable. Finally, Saeedi et al. [93] evaluated people with hallux valgus wearing their own shoes, shoes with a round toe-box and shoes with a stretchable fabric upper, and found that the shoes with the stretchable upper generated lower toe pressures and were perceived to be the most comfortable.

Four studies evaluated the influence of shoe microclimate (i.e., temperature, moisture and ventilation characteristics) on comfort, generally in the context of footwear worn in cold environments (such as trekking boots [15], safety boots [45] and ski boots [51]), but also in sandals used in combination with footwarmers indoors [103]. The findings of these studies were inconsistent, in that higher temperatures were found to be associated with improved comfort perceptions in ski boots [51], when adding insulation and toecaps to safety boots [45] and in indoor sandals [103], but lower temperatures were found to be more comfortable in trekking boots [15]. Likely explanations for this inconsistency are the range of ambient temperatures each study was performed in (which varied from − 6.8 to 23.3 °C) and different methods for measuring in-shoe temperature. The role of moisture and ventilation has been less studied and is inherently difficult to delineate from temperature effects. However, the study of trekking boots by Arezes et al. [15] found moisture retention to be of secondary importance to temperature in determining comfort levels.

Twenty-six studies have been conducted to assess the effects of insoles on footwear comfort. However, it is difficult to draw clear conclusions from the available literature due to the variation in populations studied and wide range of insole designs used. Broadly, the evidence suggests that the addition of insoles improves footwear comfort in casual footwear [95], factory footwear [12], running shoes [25, 26, 49, 84], high heels [53, 54] and police boots [94]. However, no significant improvements in comfort have been reported when adding flat cushioning insoles to walking shoes [74] or running shoes [86], or contoured insoles to cycling shoes [20]. Furthermore, the effect of insoles on footwear comfort is influenced by the fit of the shoe, as it has been observed that insoles may decrease comfort if they make the shoe too tight [34].

Studies comparing different insole designs have generally found that softer, more flexible insoles are perceived as more comfortable (in casual footwear [99], running shoes [31, 67, 80] and military footwear [42, 69]). However, comfort perceptions related to insole hardness may vary according to an individual’s occupation, as Anderson et al. [14] have reported that people standing for long periods at work prefer soft materials under the heel and forefoot but firmer materials under the arch. The influence of insole shape is uncertain, with studies reporting flat insoles to be more comfortable than contoured [48], contoured more comfortable than flat [69, 100], or no difference between the two [73]. Furthermore, the effect of insole customisation is unclear. Fully customised orthoses have been reported to be more comfortable than semi-customised insoles in runners [36], while reductions in comfort have been observed with the addition of anterior wedges when performing a load lifting task [46] and lateral forefoot posting when cycling [21].

One study assessed the effect of repeated wear on comfort while wearing badminton shoes [55]. Badminton players performed direction change manoeuvres while wearing new shoes and then the same shoes after 96 h of wear (6, 2-h training sessions per week for 8 weeks). The worn shoes were perceived to be significantly less comfortable in relation to in-shoe climate, medio-lateral stability, and overall fit, although performance in direction change manoeuvres was not adversely affected.