Study population
From July 1 to July 9, 2019, we conducted a face-to-face survey including men and women aged ≥40 years residing in Settsu City. These participants were recruited either when they participated in a specific health examination conducted at the Settsu Health Center or via Settsu’s public relations magazine (Fig. 1). We assessed the validity of the finger-circle test for assessing sarcopenia by comparing the test data obtained from participants aged 40−91 years with their body composition-based assessment data (n = 525).
The present study is a multisite population-based cross-sectional anonymous mail survey of men and women aged ≥40 years residing in the cities of Settsu and Hannan in Osaka Prefecture, Japan. Settsu is a municipality in Mishima District (population: 85007; area: 14.87 km2) in northern Osaka Prefecture [26], while Hannan is a municipality in Sennan District in southern Osaka Prefecture (population: 54276; area: 36.17 km2) [26]. Settsu and Hannan have 10 and 8 elementary school zones, respectively. To obtain a representative sample of the general population, we randomly selected 1000 subjects aged ≥40 years stratified by sex and age group from each school zone (Fig. 1). Survey forms and envelopes were color-coded by school zone to identify each school zone respondents when their survey forms were returned. The Settsu mail survey was conducted from February 22, 2019 to March 5, 2019. Although 4502 survey forms were returned by March 7, 2019, additional 1,307 forms that were returned after that date were also included in the analysis (response rate: 58.1%). The Hannan mail survey was conducted from January 20, 2020 to February 7, 2020. Survey forms were included only if they arrived by February 13, 2020 (response rate: 60.0%). For both surveys, we sent a reminder postcard one week prior to the end of the survey period.
Of these participants, we excluded those with missing data on finger-circle test, age, or sex, and those who had a self-reported “needed long-term care” or “needed support”. Ultimately, we included 9337 participants to this study.
Survey content
In the Settsu and Hannan mail surveys, we collected the following basic characteristics: age (years); sex (male or female); certification of needed support or long-term care (none, needed support level 1−2, needed long-term care level 1−5); height (centimeters); body weight (kilograms); drinking status (“Do you drink alcohol?”: almost daily, sometimes, almost never, never); smoking status (“Do you smoke?”: almost daily; sometimes; used to, but quit; never); living status (“Do you live alone?”: yes, no); socioeconomic status (“Economically, how does your life feel currently?”: hard, somewhat hard, somewhat easy, easy); self-reported health (“How healthy do you normally feel?”: very healthy, somewhat healthy, not very healthy, unhealthy); self-reported physical fitness (“How confident are you in your current physical fitness?”: extremely confident, somewhat confident, slightly anxious, very anxious); exercise habits (“Do you go walking or engage in other exercise at least once per week?”: yes, no); mastication function (“Is it more difficult for you to eat hard foods now than it was six months ago?”: yes, no); number of meals (“How many meals do you eat per day (excluding snacks)?”: [number of meals]); sleep status (“In the past month, have you been getting enough rest from sleep?”: yes, somewhat, not very much, not at all); short-term cognitive ability (“Can you remember what happened 5 minutes ago?”: yes, no); and gait speed (“Do you feel like your gait speed is slower than it used to be?”: yes, no). BMI was calculated as body weight (kg) divided by the square of height (m).
Definition of sarcopenia
In the face-to-face survey, height was measured in 0.1 cm increments using a stadiometer with the participants’ shoes removed. Body composition was assessed via bioimpedance (MC-780A, TANITA Corp., Tokyo, Japan) with participants wearing as light clothing as possible [27]. Sarcopenia was defined by a skeletal muscle mass index (SMI) (evaluated by the bioimpedance method) of <7.0 kg/m2 in men and <5.7 kg/m2 in women, in accordance with the AWGS 2019 Consensus [2]. SMI was calculated as the ALM (kg) divided by the square of the height (m2).
In the mail survey, sarcopenia was assessed by the self-reported finger-circle test, that was validated in adults aged ≥65 years [14, 15]. The finger-circle (yubi-wakka) test determines whether the circumference of a participant’s finger, formed by their index fingers and thumbs, is smaller than the participant’s maximum non-dominant calf circumference. This test is explained in detail elsewhere [14, 15]. By referring to an illustration printed on the survey form, we asked the participants to conduct the finger-circle test according to the following instructions (Fig. 2): “Put your two index fingers together and your two thumbs together to make a circle around your calf. Compare the size of this finger-circle to the thickness (circumference) of the thickest part of your calf (calf is smaller [smaller], calf and finger-circle are about the same size [same size], or calf is bigger [bigger]).” A prospective cohort study revealed an association between the finger-circle test outcomes and sarcopenia and mortality risks in adults aged ≥65 years [14].
Statistical analysis
Settsu and Hannan mail survey data were analyzed separately. Continuous and ordinal participant characteristics were classified for: age (40−49, 50−59, 60−69, 70−79, and 80 years); BMI (<18.5, 18.0−24.9, 25.0−29.9, and ≥30.0 kg/m2); alcohol consumption (“almost daily,” “sometimes,” or “almost never” = alcohol drinkers and “never” = non-drinkers); smoking (“almost daily” or “sometimes” = current smokers, “used to, but quit” = past smokers, and “never” = never-smokers); socioeconomic status (“easy” or “somewhat easy” = high socioeconomic status, and “somewhat hard” or “hard” = low socioeconomic status); self-reported health (“very healthy” or “somewhat healthy” = good self-reported health, and “not very healthy” or “unhealthy” = poor self-reported health); self-reported physical fitness (“extremely confident” or “somewhat confident” = good self-reported physical fitness, and “slightly anxious” or “very anxious” = poor self-reported physical fitness); number of meals (≥3, 2, and 1 meal); sleep status (“yes” and “somewhat” = good sleep status, while “not very much” and “not at all” = poor sleep status). Participant characteristics were described for the variables mentioned above. Continuous variables are shown as means and standard deviations, while categorical variables are shown as numbers and percentages. Missing values for these variables were created from five datasets by multiple imputation by chained equation [28]. All missing values were assumed to be missing at random. In addition, complete cases (cases with no missing values) and cases with missing values were compared by age, sex, and prevalence of sarcopenia.
In the face-to-face survey, height, and body composition data for each group of finger-circle test responses (bigger, same size, smaller) are shown as means and standard deviations. The p-values for linear trends were calculated by reporting the finger-circle test (the exposure variable) as a continuous variable. To examine the validity of the finger-circle test in assessing sarcopenia, we calculated the area under the receiver operating characteristic curves (AUROC) for the finger-circle test in relation to sarcopenia defined using SMI (the direct measurement). Previous studies verified the validity of the finger-circle test only in adults aged ≥65 years [14, 15]. Therefore, in our analyses, we stratified the participants by age (≥65 or <65 years) and sex (male or female).
Prevalence of sarcopenia was reported as number of cases and percentages for the participants in the mail survey. To adjust for the distribution ratio of the sample in the mail survey participants to the distribution ratio of Settsu and Hannan populations, we calculated the prevalence of sarcopenia weighted by the distribution ratios according to age (40−49, 50−59, 60−69, 70−79, and ≥80 years) and sex (male or female) in each school zone [29]. We used the chi-square test to compare characteristics between the sarcopenia and non-sarcopenia groups.
We used multiple logistic regression analysis to examine factors associated with the prevalence of sarcopenia in the mail survey participants. In our multivariate analysis, the dependent variable was the prevalence of sarcopenia as assessed by the finger-circle test; while the explanatory variables were age, sex, BMI, alcohol consumption, smoking status, living status, socioeconomic status, self-reported health, self-reported physical fitness, exercise habits, mastication function, number of meals, sleep status, short-term cognitive ability, and gait speed. We chose this model because a previous study reported that most of the above variables are associated with the prevalence of sarcopenia [30]. The results of these analysis were reported as odds ratios (OR) and 95% confidence intervals (CI). For each variable, the following reference groups were used to calculate the OR: age 40−49 years, male sex, BMI 18.5−24.9 kg/m2, non-drinker, never-smoker, living together, high socioeconomic status, good self-reported health, good self-reported physical fitness, “yes” to exercise habits, good mastication function, ≥3 meals, good sleep status, good short-term cognitive ability, and normal gait speed. We also conducted a similar analysis including all the participants. To conduct the sensitivity analysis for assessing the robustness of the above results, we conducted a similar analysis using the complete cases dataset [31].
In the statistical analyses, p-values <0.05 were considered significant. All analyses were performed using the JMP Pro, version 13.2 (SAS Institute, Inc., Cary, NC, USA) and/or R software 3.4.3 (R Development Core Team, Vienna, Austria).
Ethical considerations
Regarding the Settsu and Hannan mail surveys, we filed an application for an ethical review with the National Institutes of Biomedical Innovation, Health and Nutrition Institutional Review Board (Number: kenei 89, December 3, 2018). An ethical review was deemed unnecessary for the following reasons: 1) no one other than the supervisors at Settsu and Hannan had access to the residents’ information; and 2) the mail surveys were anonymous and contained no information that could be used to identify the individuals. When we sent out the survey forms, we included a document that explained the following: the objective of the study, our methods for selecting the participants, that the returning of the survey form constituted consent to participate in the study, and that the survey forms were color-coded by school zone. The Settsu City face-to-face survey was approved by the Research Ethics Committee of the National Institutes of Biomedical Innovation, Health and Nutrition (Ikikenhatsu-178-1) and has been registered with the University Hospital Medical Information Network in Japan clinical trial registration system (UMIN000036880). Study procedures as well as the risks associated with participating were explained and written informed consent was obtained from all the participants. Moreover, all the study procedures were performed in accordance with relevant guidelines/regulations.