The present study is the first to report sex-specific reference values for SPPB among community-dwelling adults aged 21–80+ years old in Singapore. Among older adults aged ≥60 years, SPPB and GS subtest had varied performance in assessing sarcopenia but moderate-to-excellent performance for discriminating severe sarcopenia. We show that SPPB cut-off score of ≤11 had optimal sensitivity and specificity for discriminating sarcopenia and severe sarcopenia in this Asian cohort of community-dwelling older adults, based on various AWGS19 definitions. The study results also suggest that regardless of sex, the GS subtest could be useful in assessing sarcopenia in our population.
In our study population, more than half of participants aged 21–80 years and over a third participants aged >80 years achieved the maximum SPPB score of 12. This implies a ceiling effect for SPPB in our population, as >20% of men and women across all age groups achieved the highest possible score (23). Our findings agree with a Norwegian study which reported ceiling effects of SPPB, across age groups 40–80+ years (3). However, such a ceiling effect of SPPB was not observed in Colombian adults aged >80 years, with 19.8% of males and 7% of females with an SPPB score of 10–12 (4). Across ages 40–80+ years, mean SPPB scores in men and women were similar between our study participants and Norwegian adults (3). Compared with Colombian older adults (60–80+ years), the mean SPPB in our population was higher in both sexes (by ~2–3 points) (4). These findings suggest that SPPB scores differ by population and population-specific reference values are necessary. The presence of ceiling effects in our population support the need to report specific SPPB subtest values, rather than aggregated scores, in order to better classify physical performance in community-dwelling older adults with higher functional ability. The disparity in SPPB scores between populations could be due to socio-economic, racial or ethnicity differences. For example, poverty and lower education were associated with greater likelihood of physical functioning limitation among older adults aged >60 years (15). In older adults aged ≥65 years, non-Hispanic blacks had poorer SPPB scores and greater mobility disability than non-Hispanic whites, implying that race and genetic factors could also affect physical function (14). Therefore, it is important to report population-specific SPPB and individual subtest values in community-dwelling adults.
Sarcopenia is associated with functional decline, increased risk of frailty, falls and mortality (24), which contribute to huge personal, social and economic burdens (25). In our study, the prevalence of sarcopenia and severe sarcopenia combined ranged from 30–41% in men and 23–40% in women depending on AWGS19 definition, justifying the need for markers such as SPPB, to assess sarcopenia and poor physical function in a quicker and easier manner among the wider population. The prevalence of severe sarcopenia among Caucasian older adults was 7% and 8% according to EWGSOP2 SPPB and GS criteria respectively (26), the difference between definitions was smaller than our study of 3% and 17% for AWGS19 SPPB and GS criteria respectively. These differences could be attributed to the different cut-offs for SPPB and GS between EWGSOP2 and AWGS19 for the diagnosis of severe sarcopenia. The varied sarcopenia prevalence by definition also highlights the importance of comparison across different physical performance criteria, for the performance of SPPB and its subtests in discriminating sarcopenia.
The present study showed that SPPB cut-point of 11 gave the optimal sensitivity (42–58%) and specificity (69–71%) for assessing sarcopenia in community-dwelling adults ≥60 years. For severe sarcopenia in men and women, SPPB cut-off of ≤11 had optimal performance for AWGS19 GS and STS definitions. Although SPPB cut-off of ≤9 in men and ≤7 in women gave the best sensitivity and specificity for AWGS19 SPPB definition for severe sarcopenia, these results should be interpreted with caution due to the small sample size of men (n=8) and women (n=2) with severe sarcopenia based on the AWGS19 SPPB criterion. The cut-off of ≤11 was higher than the recommended SPPB cut-point of ≤8–9 suggested by EWGSOP and AWGS19 SPPB criteria for sarcopenia (10, 11, 26, 27). Other studies also reported SPPB cut-points of 7–9 being associated with higher mortality risk (28-30). The optimal SPPB cut-point of ≤11 for discriminating severe sarcopenia in our study was also higher than an Australian study which showed an optimal SPPB cut-point of 5–8, depending on physical performance definition such as GS and SPPB, for severe sarcopenia (26). Differences in study populations likely explain the disparity. Our study participants were community-dwelling older Asian adults with high functional ability, which differed from other studies involving Caucasians (26, 29), outpatient or hospitalised individuals who might have limited physical function (28, 30). Furthermore, SPPB scores are commonly stratified into groups (0–3, 4–6, 7–9, 10–12), with a score of 10–12 as the reference (normal) group (7, 27, 31). Within individuals with SPPB score 10–12, varying physical function, risk of sarcopenia and mortality plausibly exist. Compared with individuals with maximum SPPB score, individuals with score of 11 were 1.4 times more likely to develop mobility disability in a 3-year follow-up study (32). These results suggest that a 1-point decrease in SPPB score could impact physical function (33). Therefore, in functional community-dwelling older adults, a higher SPPB cut-off can better discriminate sarcopenia. Nonetheless, SPPB had limited performance in discriminating sarcopenia in our study, despite moderate-to-excellent performance in assessing severe sarcopenia. These results suggest that the cut-point of ≤11points might be more useful for assessing severe sarcopenia among community-dwelling older adults.
We compared the performance of individual SPPB subtests in assessing sarcopenia. Our results demonstrate that GS, but not STS subtest, generally had better performance than total SPPB score in discriminating sarcopenia, and had comparable performance with SPPB in assessing severe sarcopenia. Among our participants, GS subtest cut-off of 1.0 m/s gave optimal sensitivity (57–70%) and specificity (68–72%) in assessing sarcopenia in men and women for STS and SPPB criteria. For discriminating severe sarcopenia, optimal GS subtest cut-off was ≤1.0 m/s in women (GS and STS criteria) and ≤0.9 m/s in men (all AWGS19 criteria). Our findings agree with the recommended AWGS19 cut-off for GS criteria, despite a different walk distance of 6m in AWGS19 and 8ft in the present study (11). Other studies also reported a GS of <1.0 m/s in sarcopenic older adults (34), and found greater dementia risk and poorer health outcomes in adults >80 years with GS of <1.0 m/s (35, 36). However, GS cut-off recommendations varied according to sarcopenia-associated health outcomes, such as hospitalisation, falls, mortality, cognitive impairment (37). For example, other studies including the EWGSOP recommended a GS cut-off of 0.8 m/s (10, 12, 38), due to its association with lower life expectancy (39) and disability (37). Nonetheless, GS cut-offs are dependent on health status and demographics, supporting the need for population-specific studies investigating the diagnostic value of GS in sarcopenia. Herein, we propose that the GS subtest of SPPB might be a useful, simple and accessible tool for assessing sarcopenia in functional community-dwelling older adults.
Our study used a well-established performance-based physical function assessment and recruited randomly from the general population, suggesting a good degree of generalisability. However, the study findings cannot be generalised to people living in institutions. Future longitudinal studies should investigate the prognostic value of SPPB and its subtests in diagnosing sarcopenia and severe sarcopenia.