Figure 1 illustrates the study selection process and results via the PRISMA flow diagram. There were 2,063 articles identified in the initial literature search, which were used to screen for eligibility. After the removal of 946 duplicates, 1,077 articles were excluded according to the exclusion criteria above. The most common reasons for exclusion were case reports and animal/in vitro studies. The full text of the remaining 40 articles was reviewed and 23 of them were excluded based on the inclusion criteria. The primary reasons for exclusion were inappropriate target population, irrelevant outcomes, or unsuitable measurements. As such, a total of 17 articles (i.e., 17 studies) met the inclusion criteria and were included in this review [26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42].

PRISMA flow diagram for the identification of studies

Results of the methodological quality assessment for cross-sectional studies were shown in Table 3 and those for cohort studies were presented in Table 4. The 14 cross-sectional studies were scored between 5.5 and 8 (out of 8) and were rated as moderate or high quality. The two cohort studies were scored as 8.5 and 9 (out of 11) and were rated as moderate or high quality. That is, all studies included in this review had a low risk of bias. It is worth noting that all studies, including both cross-sectional and cohort studies, reported the outcomes in a valid and reliable way and used appropriate statistical analysis methods. One study [40], which reported the secondary data analysis of a cohort study, was not assessed for its methodology quality as that study did not report results based on primary data.

Table 2 presents the characteristics of the included studies. Of the 17 studies, 14 articles were cross-sectional [26,27,28,29,30,31,32,33,34,35,36, 38, 39, 41, 42], two articles reported a cohort study [37, 39], and one study [40] reported the secondary data analysis results. According to the study selection criteria above, no meta-analysis is required for this systematic review.

Five studies were conducted in Japan [29, 30, 33, 39, 41], three in Korea [31, 36, 42], two in China [27, 38], two in Turkey [34, 37], two in Germany [28, 32], and the others in Singapore [26], Italy [35], and USA [40] each. There were 8,392 adults with diabetes included in this review, and one study included only males [42]. The mean age of participants included in this review was reported in 16 studies [26, 27, 29,30,31,32,33,34,35,36,37,38,39,40,41,42], ranging from 52 to 75 years old. According to the body mass index [43], three studies included participants who were obese [31, 34, 40].

In addition, 14 articles specified the type of diabetes, with 12 studies including T2D only [26, 27, 30, 32,33,34,35,36,37,38, 40, 41] and two studies including both type 1 diabetes (T1D) and T2D [29, 39]. Five studies examined diabetes multimorbidity, including cognitive impairment, diabetic peripheral neuropathy, cardiovascular disease, retinopathy, and diabetic polyneuropathy [30, 33, 35, 36, 41].

The muscle function measurements varied across the included studies. The diagnosis of sarcopenia was reported in eight studies, with three studies using the European Working Group on Sarcopenia in Older People (EWGSOP) criteria [27, 34, 37] and five studies using the Asian Working Group for Sarcopenia (AWGS) criteria [26, 29, 33, 38, 39]. Two studies further specified the definition of probable sarcopenia [27, 37]. Five studies examined the gait test to assess the physical performance of participants, including the 4-m gait test [29, 37, 39] and 6-m gait speed [26,