This study developed a short-form SHS questionnaire (i.e., SHSQ-SF) containing nine key items from the SHSQ-25. This simplified instrument had the highest factor loading of the original SHSQ-25, explaining 57.01% of the variance in SHS [45, 46]. Although the SHSQ-SF is a unidimensional measure, it can evaluate SHS efficiently when survey space is restricted. It is therefore suitable for use with large-scale populations. Other SHS scales such as the MSQA, SHMS V1.0, CSHES, and SSS [22, 25–27] have a large number of items. For example, the SSS comprises three symptom dimensions with 10 factors supported by 58 items; the CSHES includes three symptom dimensions with 18 factors supported by 64 items [26].
By comparison, the SHSQ-SF features nine items, based on which respondents indicate their health complaints over the past three months: 1) exhausted without a significant increase in physical activity; 2) lethargic when working; 3) stiff muscles or joints; 4) pain in shoulder/neck/waist; 5) out of breath when sitting still; 6) chest congestion; 7) poor appetite; 8) impaired short-term memory; and 9) inability to respond quickly. Items are scored as 0 = never or almost never, 1 = occasionally, 2 = often, 3 = very often, or 4 = always and then summed to obtain the total SHSQ-SF score. Participants in this study were deemed to be in either ideal health (total score < 11) or suboptimal health (total score ≥ 11). This cutoff point was determined with the ROC using the training dataset with an AUC of 0.985. For external validation, the SHSQ-SF had an AUC of 0.975, Se of 94.2%, Sp of 87.4%, and kappa of 0.720. This questionnaire was highly effective at identifying individuals with SHS overall. Its AUC exceeded that of the SSS, implying that the SHSQ-SF can more effectively identify people with SHS. Such results have not been reported for other SHS measures.
Scores for items on the SHSQ-SF differed significantly between individuals with higher SHS scores (i.e., the highest 27% of the sample) and lower SHS scores (i.e., the lowest 27% of the sample). Accordingly, this scale can potentially discriminate people at high risk of SHS from the healthy population. Each item was also significantly correlated with the total SHS score (ranging from 0.682 to 0.826), illustrating the questionnaire’s homogeneity [47]. A sensitivity analysis based on CITC further confirmed the measure’s stability.
The SHSQ-SF also showed clear reliability and validity. Regarding the former, a Cronbach’s α between 0.7 and 0.8 is fairly good and 0.8–0.9 is very good [48]. The developed questionnaire had a Cronbach’s α of 0.902 with a split-half reliability of 0.863, suggesting high internal consistency. We examined structural validity and convergent validity as well. All the modified model fit indices (i.e., χ2/dƒ, GFI, NFI, RFI, RMSEA) demonstrated good validity [49, 50]. With respect to convergent validity, although one statistic (AVE) was somewhat low, the other (CR) was high [51]. The questionnaire’s factor structure hence possessed sufficient convergent validity. The Cronbach’s α for the SHSQ-SF aligned with those of the SHSQ-25, MSQA, SHMS V1.0, CSHES, and SSS, all of which were greater than 0.8 to indicate high internal consistency [22, 25–27].
Disease onset can be prevented with early detection followed by appropriate intervention. Clinicians should be encouraged to shift from delayed intervention to early screening for individuals with SHS. To this end, PPPM/3PM presents a path forward in medicine; the concept enables early identification of populations who may have SHS and who are at risk of developing NCDs [52]. Although SHS is understood as an intermediate state between health and disease [53], its identification remains challenging. A full health evaluation is necessary to effectively implement public health interventions [54]. A nationwide investigation of health status, psychology, behavior, and social determinants has been carried out in China since 2020 [55]. SHS assessment was included under the PPPM/3PM framework in the investigation’s 2023 proposal. We developed a short-form SHS questionnaire to describe SHS in a nationwide cross-sectional study. To our knowledge, the SHSQ-SF is the first simplified SHS measure suitable for large-scale population-based research. The SHSQ-SF, as a cost-effective and powerful SHS evaluation tool, can minimize economic barriers associated with a lack of SHS-specific laboratory tests and treatments. The questionnaire can in turn facilitate PPPM/3PM to substantially influence healthcare.
Limitations
This study had several limitations. First, most participants were young; elderly individuals with somatic and psychological disorders were excluded at enrollment. Our findings may not be representative. Second, no objective or physical assessment standard is available for SHS. We took the SHSQ-25 as the gold standard because it is the most widely used scale of its kind in Africa, Asia, Europe, and Oceania. Finally, some participants may have had undiagnosed conditions. The reported SHS level is thus probably overestimated.