Introduction
Sarcopenia, a chronic condition denoted by age-related loss of skeletal muscle mass, correlates with declining muscle strength and functional performance, which can result in prolonged recovery period and reduced independence [1, 2]. The prevalence of sarcopenia in older patients (aged ≥ 60 years) ranges from 9.9 to 40.4% according to different diagnostic criteria [3]. This number will certainly increase in countries with aging societies, where sarcopenia has become an increasingly serious medical and public health issue. Accumulating evidence has demonstrated a significant correlation of sarcopenia with metabolic disorders, functional decline, and premature death in the general population [4, 5]. It has been reported that sarcopenia can double individual medical costs and the direct costs attributable to low muscle mass in the US are estimated to exceed $18.5 billion per year [6].
Sarcopenic obesity (SO), a distinct condition of both reduced muscle mass and excess adiposity, also occurs and has raised a continuing concern [7]. The estimated prevalence of SO is up to 11% worldwide, although prevalence rates are numerically higher in subgroups of patients who are aged ≥ 75 years, who are hospitalized, or who are North/South Americans [8]. For middle-aged adults, obesity can accelerate the penetration of fat into muscle, lower physical function, and cause disability, thus increasing the risk of all-cause mortality [9‐11]. However, considerable debate has existed regarding the health effects of obesity on older adults, in particular on those with SO, and whether excessive weight may be of benefit [12‐14].
Anzeige
Studies have shown that obesity (defined by high body mass index [BMI]) is associated with a decreased risk of all-cause mortality among older adults compared with the control group, which is usually called ‘obesity paradox’ [15]. However, the mortality risk appears to be higher in older patients with SO than in those with sarcopenia and without obesity [16‐18]. Nearly three out of four older Americans are likely to have hypertension, which remains a leading cause of disability or death worldwide [19, 20]. Importantly, when hypertension is combined with either obesity or sarcopenia, the mortality risk can be even higher [21, 22]. Considering that ‘obesity paradox’ also disappears in patients with concomitant obesity and sarcopenia, the combination of any two of hypertension, obesity, and sarcopenia may have a synergistic effect of increasing mortality risk in older adults. Nevertheless, it remains unclear whether a copresence of obesity and sarcopenia, or obesity or sarcopenia alone, increases the mortality risk in patients with hypertension.
Obesity and insulin resistance are common in patients with hypertension and may contribute to hypertension via multiple potential mechanisms, including activation of the sympathetic nervous system and the renin-angiotensin-aldosterone system, oxidative stress, and endothelial insulin resistance [23]. The increase in free fatty acids induced by obesity can enhance insulin resistance and then inhibit autophagy by activating mammalian target of rapamycin pathway, thereby causing lysosomal degradation of proteins and organelles in muscle [24]. These findings reveal that in patients with hypertension, obesity and insulin resistance may play a vital role in the pathogenesis of sarcopenia and the prevalence of SO may be higher than the general population. However, uncertainty still exists regarding the prevalence of SO in patients with hypertension.
The aims of this study are to investigate the prevalence of SO in older adults with hypertension, and examine the association between various body composition phenotypes and all-cause mortality, with a focus on SO.
Methods
Study population
The National Health and Nutrition Examination Survey (NHANES), undertaken by the National Center for Health Statistics (NCHS) at the Centers for Disease Control and Prevention, is a nationally representative health survey of the US noninstitutionalized civilian population [25]. Based on complex, stratified, multistage, probability clusters, NHANES data are consecutively obtained since 1999 and are released every two years. Detailed information concerning the sample design, weighting, and analytic methodology have been described elsewhere [26]. The NCHS Ethics Review Board has approved the NHANES protocol [27]. Written informed consent was acquired from each participant. This study included older adults aged ≥ 60 years with hypertension from 3 cycles of the NHANES cohort (1999–2000, 2001–2002, and 2003–2004) that collected data on body composition with the full spectrum of age. All individuals were prospectively followed up for mortality through December 31, 2019, permitting more than 15 years of observation for survival outcomes. This study was performed following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for reporting cohort studies (Table S1) [28].
Anzeige
There were 4314 older adults with hypertension initially identified from the NHANES 1999–2004. Participants with unclear vital status (n = 2) and missing information about body composition or potential covariates (n = 2896) were excluded. Subsequently, we excluded individuals with cancer at baseline (n = 278), because cancer is strongly correlated with the depletion of muscle and fat storage and cachexia may generate large confounding effects. We also excluded individuals dying within the first two years after in-house interviews (n = 33), because those at high risk of all-cause mortality were extremely likely to change their dietary habits and own quite different muscle mass. The final analytical cohort included 1105 individuals (Fig. 1).
Fig. 1
Flow chart of participants included. NHANES, National Health and Nutrition Examination Survey
×
![]()
Assessment of hypertension
Following a 5-min rest in a seated position, blood pressure was measured by a certified examiner with a mercury sphygmomanometer according to the recommendations from the American Heart Association [29]. Three or four groups of systolic blood pressure (SBP) and diastolic blood pressure (DBP) were obtained and averaged. Hypertension was defined as (1) mean SBP ≥ 140 mmHg or DBP ≥ 90 mmHg, (2) self-reported physician diagnosis of hypertension, and/or (3) current use of anti-hypertensive medications [30].
Measurement of body composition
Height, weight, and waist circumference (WC) were measured by a trained examiner following standard procedures and equipment [31]. BMI was calculated as weight divided by height squared (kg/m2). WC was measured along the superior lateral border of the iliac crest using a steel tape. Appendicular lean mass (ALM) and body fat (BF) were measured by a certified radiology technologist using whole body dual-energy X-ray absorptiometry (DXA; QDR 4500 A; Hologic, Bedford, MA). ALM was calculated as appendicular lean soft tissue mass excluding body mineral content [32, 33]. Thereafter, ALM was normalized by height squared (ALM/height2), weight (ALM/weight), and BMI (ALM/BMI), as suggested by the European Working Group on Sarcopenia in Older People (EWGSOP) [34], Janssen et al. [35], and the Foundation for the National Institutes of Health (FNIH) Sarcopenia Project [36], respectively.
Sarcopenia was broadly defined based on low lean mass (LLM), and was defined using the following 3 sex-specific cutoffs, respectively: (1) ALM/height2 < 7.0 kg/m2 in males and < 5.5 kg/m2 in females; (2) ALM/weight < 0.262 in males and < 0.196 in females; and (3) ALM/BMI < 0.789 in males and < 0.512 in females. Obesity was defined as BMI ≥ 30 kg/m2 [37], BF percentage ≥ 30% in males and ≥ 42% in females [38], or WC ≥ 102 cm in males and ≥ 88 cm in females [39], as previously described [40]. Obesity was diagnosed if at least 1 out of the 3 obesity criteria was fulfilled. Subsequently, individuals were divided into 4 groups that combined muscle and fat status (Group 1 no obesity or LLM; Group 2 obesity [without] LLM; Group 3 LLM [without] obesity; and Group 4 LLM with obesity) [41].
Ascertainment of mortality
Participant outcomes were ascertained via the NHANES Public Use Linked Mortality File, which contains comprehensive information on survival based on the result of a probabilistic match between the NHANES and National Death Index death certificate records through December 31, 2019 [42]. Follow-up duration for each individual was calculated as the time of baseline to the last known date alive or censored from the mortality file.
Study covariates
To reduce potential confounding from covariates, information on age, sex, race/ethnicity, educational level, family income-to-poverty ratio, smoking status, alcohol intake, physical activity, total protein intake, and chronic health conditions (i.e., diabetes mellitus, chronic kidney disease [CKD], cognitive problem, arthritis, hyperlipidemia, cardiovascular disease, and cancer) was obtained from standardized questionnaires. Race/ethnicity was categorized as non-Hispanic White, non-Hispanic Black, Mexican American, and the other racial/ethnic group. Educational level was classified as less than high school, high school, and college or higher. Family income-to-poverty ratio was divided into < 1.30, 1.30–3.49, and ≥ 3.50, with higher ratios indicating higher levels of household income.
Participants were grouped into nonsmokers, former smokers, and current smokers by the responses about whether they smoked no less than 100 cigarettes during lifetime, and whether they still smoked currently. The mean intakes of alcohol and total protein were obtained from two 24-hour dietary recalls. Participants were divided into nondrinkers (0 g/day), moderate drinkers (0.1–27.9 g/day in males and 0.1–13.9 g/day in females), and heavy drinkers (≥ 28 g/day in males and ≥ 14 g/day in females). Physical activity was measured using metabolic equivalent for the activities relating to frequency and duration of walking/bicycling, tasks around home/yard, and muscle strengthening activities (Table S2) [43]. Physical activity was then categorized as low (< 600 MET-min/week), medium (600–1200 MET-min/week), and high (≥ 1200 MET-min/week) following the World Health Organization guideline recommendations [44].
Anzeige
Diabetes mellitus was defined as (1) self-reported diabetes, (2) hemoglobin A1c levels ≥ 6.5%, (3) fasting glucose levels ≥ 7.0 mmol/L, (4) random glucose levels ≥ 11.1 mmol/L, and/or (5) current use of anti-diabetic medications [45]. Estimated glomerular filtration rate (eGFR) was calculated according to the Chronic Kidney Disease Epidemiology Collaboration equation [46]. A diagnosis of CKD was made if eGFR was less than 60 mL/(min·1.73 m2) [47]. Diagnoses of cognitive problem, arthritis, cardiovascular disease, and cancer were determined based on self-reporting using health questionnaires. Hyperlipidemia was defined as (1) triglyceride levels ≥ 150 mg/dL, (2) total cholesterol levels ≥ 200 mg/dL, (3) low-density lipoprotein cholesterol levels ≥ 130 mg/dL, (4) high-density lipoprotein cholesterol levels < 40 mg/dL in males and < 50 mg/dL in females, and/or (5) current use of anti-hyperlipidemic medications [48]. According to the National Cholesterol Education Program’s Adult Treatment Panel III [49], individuals were diagnosed with metabolic syndrome (MetS) when meeting any 3 of the following criteria: elevated WC (≥ 102 cm in males and ≥ 88 cm in females), elevated triglyceride (≥ 150 mg/dL and/or current use of fibrate drugs), low high-density lipoprotein cholesterol (< 40 mg/dL in males and < 50 mg/dL in females and/or current use of niacin drugs), elevated blood pressure (mean SBP ≥ 130 mmHg or DBP ≥ 85 mmHg and/or current use of anti-hypertensive medications), and impaired glucose tolerance (fasting glucose levels ≥ 5.6 mmol/L and/or current use of anti-diabetic medications).
Statistical analysis
Data analysis was conducted from October 29, 2022 to February 13, 2023. According to the NHANES analytic guideline, primary sampling units, stratification, and sampling weights were used for all statistical analyses to ensure the accurate calculation of the point estimates and standard errors [50]. Based on the 3 criteria for sarcopenia, comparisons of baseline characteristics across various body composition phenotypes (i.e., normal, obesity [without LLM], LLM [without obesity], and LLM with obesity) were performed by one-way analysis of variance for continuous variables and chi-square test for categorical variables.
The prevalence of LLM with obesity, LLM, obesity, and no LLM or obesity in patients with hypertension was calculated. Rates were stratified by the 3 different ALM indexes (ALMIs). Multivariable Cox regression analysis was adopted to calculate the hazard ratios (HRs) and 95% confidence intervals (CIs) to evaluate the association between body composition phenotypes and mortality. Model 1 adjusted for age, sex, and race/ethnicity. Model 2 further adjusted for educational level, family income-to-poverty ratio, smoking status, alcohol intake, physical activity, and total protein intake. Model 3 further adjusted for chronic health conditions. Moreover, we tested the interaction between LLM and obesity with regard to all-cause mortality.
To assess the robustness of the results, two sensitivity analyses were also conducted. First, given that the recently updated clinical practice guidelines recommend the use of 130/80 mmHg as the lower threshold for hypertension, this criterion was chosen as one of the definitions of hypertension in a sensitivity analysis. Second, we investigated whether results from the main analysis were maintained, when obesity was defined using only BF percentage (≥ 30% in males and ≥ 42% in females). All analyses were carried out by R software version 4.1.1 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was defined as a two-sided P-value < 0.05.
Anzeige
Results
Baseline characteristics
Of the 1105 participants, the mean (SE) age was 69.9 (0.2) years, 44.6% of participants were male, and most participants (83.6%) were non-Hispanic White. Notably, ALM normalization methods substantially affected the association between LLM with obesity and MetS; compared with obesity alone, individuals with both LLM and obesity appeared to be at higher risk for MetS by the ALM/weight and ALM/BMI indexes, but not by the ALM/height2 index. Baseline demographic and clinical characteristics by the 3 different ALMIs are summarized in Table 1. The LLM group by the ALM/weight index was excluded from further analyses, because no participant was diagnosed with LLM and without obesity. Mean BMI was 28.3 kg/m2, mean BF percentage was 31.3% in males and 42.0% in females, and mean WC was 105.2 cm in males and 96.1 cm in females. Elevated BF percentage was the measure that defined the most individuals (57.2%) with obesity, followed by BMI (31.7%) and WC (28.1%). The prevalence of obesity defined by any of BMI, BF percentage, and WC was 76.5% in older adults with hypertension. Table 2 provides detailed information on baseline anthropometry and body composition characteristics.
Table 1
Baseline demographic and clinical characteristics of the study population according to the status of obesity and LLM defined by different ALM indexes
Characteristics | Method A (ALM/height2) | Method B (ALM/weight) | Method C (ALM/BMI) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Normal (n = 177) | Obesity (n = 731) | LLM (n = 97) | LLM with obesity (n = 100) | P-value | Normal (n = 274) | Obesity (n = 702) | LLM (n = 0) | LLM with obesity (n = 129) | P-value | Normal (n = 248) | Obesity (n = 560) | LLM (n = 26) | LLM with obesity (n = 271) | P-value | |
Age, mean, years | 69.2 (0.8) | 69.4 (0.3) | 72.4 (0.9) | 72.6 (0.8) | < 0.001 | 70.4 (0.6) | 69.6 (0.3) | N/A | 70.5 (0.6) | 0.357 | 70.3 (0.7) | 69.2 (0.3) | 71.7 (1.7) | 71.5 (0.5) | 0.004 |
Sex | < 0.001 | < 0.001 | 0.002 | ||||||||||||
Female | 39.8 (3.8) | 56.0 (1.8) | 67.7 (5.9) | 64.5 (5.6) | 50.1 (3.2) | 61.9 (1.8) | N/A | 30.2 (5.7) | 52.2 (3.3) | 60.4 (1.8) | 22.1 (11.0) | 47.4 (4.1) | |||
Male | 60.2 (3.8) | 44.0 (1.8) | 32.3 (5.9) | 35.5 (5.6) | 49.9 (3.2) | 38.1 (1.8) | N/A | 69.8 (5.7) | 47.8 (3.3) | 39.6 (1.8) | 77.9 (11.0) | 52.6 (4.1) | |||
Race/ethnicity | 0.010 | 0.156 | < 0.001 | ||||||||||||
Non-Hispanic White | 82.8 (2.7) | 83.7 (1.9) | 82.4 (5.1) | 85.4 (4.8) | 82.6 (2.7) | 82.8 (2.0) | N/A | 89.9 (2.7) | 84.2 (2.5) | 84.0 (2.1) | 62.0 (13.3) | 83.5 (3.1) | |||
Non-Hispanic Black | 10.9 (2.1) | 7.0 (1.1) | 2.9 (1.3) | 0.4 (0.4) | 7.9 (1.4) | 6.9 (1.1) | N/A | 2.0 (0.7) | 8.4 (1.5) | 7.8 (1.3) | 2.5 (2.5) | 1.5 (0.7) | |||
Mexican American | 2.1 (0.6) | 3.2 (0.7) | 2.3 (0.8) | 3.3 (1.2) | 2.2 (0.6) | 3.2 (0.8) | N/A | 3.4 (1.0) | 1.9 (0.6) | 2.1 (0.5) | 6.8 (3.2) | 6.6 (1.7) | |||
Othera | 4.2 (1.8) | 6.1 (1.5) | 12.4 (4.5) | 10.9 (4.1) | 7.2 (2.2) | 7.1 (1.6) | N/A | 4.7 (2.3) | 5.6 (1.9) | 6.1 (1.6) | 28.7 (12.8) | 8.4 (2.4) | |||
Educational level | 0.193 | 0.154 | 0.007 | ||||||||||||
Less than high school | 11.2 (2.6) | 12.3 (1.7) | 14.1 (3.8) | 11.2 (3.9) | 12.3 (2.2) | 12.1 (1.8) | N/A | 12.4 (3.1) | 12.0 (2.3) | 10.1 (1.7) | 16.7 (7.6) | 17.9 (3.3) | |||
High school | 33.8 (5.1) | 46.2 (2.2) | 40.6 (5.8) | 52.5 (7.0) | 36.3 (3.9) | 46.8 (2.4) | N/A | 48.5 (5.7) | 36.6 (3.8) | 45.3 (2.6) | 32.1 (13.5) | 52.0 (3.9) | |||
College or higher | 54.9 (5.2) | 41.5 (2.4) | 45.3 (6.5) | 36.3 (5.7) | 51.4 (3.9) | 41.2 (2.4) | N/A | 39.1 (5.5) | 51.4 (3.8) | 44.5 (2.7) | 51.3 (16.1) | 30.1 (3.5) | |||
Family income-to-poverty ratio | 0.386 | 0.030 | 0.103 | ||||||||||||
<1.30 | 16.2 (4.2) | 19.1 (1.8) | 21.0 (5.6) | 15.2 (5.1) | 18.0 (3.5) | 18.6 (1.9) | N/A | 19.1 (3.2) | 17.8 (3.7) | 16.6 (1.9) | 19.7 (9.7) | 24.7 (3.7) | |||
1.30–3.49 | 36.6 (4.8) | 44.5 (2.6) | 38.1 (5.6) | 51.4 (5.1) | 37.2 (3.1) | 43.7 (2.5) | N/A | 54.4 (4.7) | 37.1 (3.4) | 44.9 (2.7) | 38.5 (16.5) | 46.8 (3.7) | |||
≥3.50 | 47.2 (6.3) | 36.4 (2.7) | 40.9 (7.3) | 33.4 (4.8) | 44.9 (4.0) | 37.7 (2.9) | N/A | 26.4 (3.5) | 45.1 (4.7) | 38.6 (3.0) | 41.8 (16.3) | 28.5 (3.8) | |||
Smoking status | 0.242 | 0.344 | 0.690 | ||||||||||||
Never | 46.1 (4.9) | 47.8 (2.3) | 56.0 (6.0) | 46.4 (5.5) | 49.7 (4.3) | 49.0 (2.6) | N/A | 40.2 (5.5) | 50.1 (4.6) | 47.3 (2.6) | 44.2 (17.0) | 48.7 (4.5) | |||
Former | 38.2 (4.7) | 43.7 (2.4) | 36.5 (6.5) | 40.4 (6.5) | 37.6 (4.4) | 42.3 (2.9) | N/A | 48.3 (5.6) | 36.8 (4.6) | 43.9 (2.9) | 48.0 (14.6) | 41.3 (4.9) | |||
Current | 15.7 (3.9) | 8.5 (1.3) | 7.5 (2.7) | 13.2 (4.2) | 12.7 (2.8) | 8.7 (1.3) | N/A | 11.5 (3.4) | 13.0 (3.1) | 8.8 (1.5) | 7.7 (4.9) | 10.0 (2.2) | |||
Alcohol intakeb | 0.616 | 0.801 | 0.083 | ||||||||||||
Never | 79.0 (4.3) | 74.5 (1.9) | 75.3 (5.4) | 66.4 (5.9) | 77.6 (3.7) | 73.1 (1.9) | N/A | 75.6 (4.1) | 78.5 (3.9) | 70.8 (2.5) | 65.3 (11.1) | 81.2 (3.0) | |||
Moderate | 13.8 (3.8) | 14.0 (1.5) | 12.3 (3.8) | 20.2 (4.6) | 13.2 (3.0) | 15.2 (1.7) | N/A | 12.7 (3.4) | 12.2 (3.1) | 15.1 (2.0) | 27.3 (9.9) | 13.8 (2.7) | |||
Heavy | 7.2 (2.6) | 11.5 (1.8) | 12.4 (4.1) | 13.4 (4.5) | 9.1 (2.4) | 11.7 (1.8) | N/A | 11.8 (3.4) | 9.3 (2.5) | 14.0 (2.2) | 7.4 (7.2) | 5.0 (1.6) | |||
Physical activity, MET-min/week | 0.336 | 0.664 | 0.423 | ||||||||||||
Low | 51.5 (4.8) | 57.6 (2.1) | 63.2 (5.9) | 66.9 (5.8) | 55.8 (4.3) | 58.4 (2.1) | N/A | 61.3 (5.8) | 53.9 (4.6) | 57.9 (2.6) | 81.6 (10.6) | 61.6 (3.6) | |||
Medium | 21.7 (3.1) | 18.3 (1.9) | 13.2 (3.8) | 17.2 (5.0) | 18.6 (2.6) | 17.8 (1.9) | N/A | 20.4 (5.0) | 19.7 (2.6) | 18.3 (2.4) | 3.4 (2.2) | 17.8 (2.5) | |||
High | 26.8 (3.6) | 24.1 (1.9) | 23.6 (5.2) | 15.8 (4.5) | 25.6 (3.1) | 23.9 (1.8) | N/A | 18.3 (3.1) | 26.4 (3.5) | 23.9 (2.1) | 15.1 (10.3) | 20.6 (3.7) | |||
Total protein intake, mean, g/day | 76.8 (3.0) | 69.8 (1.3) | 68.5 (3.0) | 60.9 (2.8) | 0.004 | 73.7 (2.4) | 68.8 (1.4) | N/A | 68.2 (2.9) | 0.214 | 73.6 (2.7) | 70.2 (1.6) | 75.8 (5.1) | 64.3 (2.1) | 0.017 |
Diabetes mellitus | < 0.001 | < 0.001 | < 0.001 | ||||||||||||
No | 90.3 (2.7) | 75.8 (1.7) | 87.4 (3.1) | 86.5 (4.3) | 89.2 (2.0) | 78.8 (1.5) | N/A | 68.1 (5.4) | 88.7 (2.2) | 78.2 (1.8) | 96.1 (2.8) | 74.2 (3.3) | |||
Yes | 9.7 (2.7) | 24.2 (1.7) | 12.6 (3.1) | 13.5 (4.3) | 10.8 (2.0) | 21.2 (1.5) | N/A | 31.9 (5.4) | 11.3 (2.2) | 21.8 (1.8) | 3.9 (2.8) | 25.8 (3.3) | |||
Chronic kidney disease | 0.141 | 0.163 | 0.044 | ||||||||||||
No | 72.3 (3.6) | 78.8 (1.7) | 70.8 (5.6) | 69.8 (6.2) | 71.7 (3.0) | 78.2 (1.8) | N/A | 74.8 (4.3) | 72.0 (3.2) | 80.3 (1.9) | 68.7 (12.1) | 70.0 (4.5) | |||
Yes | 27.7 (3.6) | 21.2 (1.7) | 29.2 (5.6) | 30.2 (6.2) | 28.3 (3.0) | 21.8 (1.8) | N/A | 25.2 (4.3) | 28.0 (3.2) | 19.7 (1.9) | 31.3 (12.1) | 30.0 (4.5) | |||
Cognitive problem | 0.001 | 0.030 | < 0.001 | ||||||||||||
No | 94.8 (1.6) | 93.3 (1.0) | 83.5 (4.9) | 82.7 (5.1) | 90.7 (2.3) | 93.3 (1.2) | N/A | 84.3 (4.3) | 90.5 (2.4) | 95.0 (1.2) | 93.1 (5.0) | 83.1 (3.2) | |||
Yes | 5.2 (1.6) | 6.7 (1.0) | 16.5 (4.9) | 17.3 (5.1) | 9.3 (2.3) | 6.7 (1.2) | N/A | 15.7 (4.3) | 9.5 (2.4) | 5.0 (1.2) | 6.9 (5.0) | 16.9 (3.2) | |||
Arthritis | 0.404 | 0.364 | 0.533 | ||||||||||||
No | 59.4 (4.8) | 51.8 (2.4) | 54.6 (5.8) | 56.8 (5.4) | 57.6 (3.4) | 52.0 (2.7) | N/A | 54.8 (4.7) | 57.4 (3.4) | 52.2 (2.8) | 61.1 (9.9) | 53.1 (4.2) | |||
Yes | 40.6 (4.8) | 48.2 (2.4) | 45.4 (5.8) | 43.2 (5.4) | 42.4 (3.4) | 48.0 (2.7) | N/A | 45.2 (4.7) | 42.6 (3.4) | 47.8 (2.8) | 38.9 (9.9) | 46.9 (4.2) | |||
Hyperlipidemia | 0.051 | 0.049 | 0.017 | ||||||||||||
No | 16.9 (3.0) | 9.1 (1.4) | 15.8 (4.0) | 16.0 (4.4) | 16.5 (2.4) | 10.0 (1.4) | N/A | 9.7 (2.6) | 17.3 (2.4) | 10.4 (1.5) | 5.6 (3.6) | 8.8 (2.0) | |||
Yes | 83.1 (3.0) | 90.9 (1.4) | 84.2 (4.0) | 84.0 (4.4) | 83.5 (2.4) | 90.0 (1.4) | N/A | 90.3 (2.6) | 82.7 (2.4) | 89.6 (1.5) | 94.4 (3.6) | 91.2 (2.0) | |||
Metabolic syndrome | < 0.001 | < 0.001 | < 0.001 | ||||||||||||
No | 82.8 (4.1) | 39.5 (2.2) | 84.9 (3.9) | 63.5 (4.9) | 83.6 (3.1) | 43.6 (2.3) | N/A | 36.9 (5.0) | 84.7 (3.2) | 43.4 (2.5) | 67.9 (12.3) | 40.2 (4.6) | |||
Yes | 17.2 (4.1) | 60.5 (2.2) | 15.1 (3.9) | 36.5 (4.9) | 16.4 (3.1) | 56.4 (2.3) | N/A | 63.1 (5.0) | 15.3 (3.2) | 56.6 (2.5) | 32.1 (12.3) | 59.8 (4.6) | |||
Cardiovascular disease | 0.795 | 0.060 | 0.090 | ||||||||||||
No | 69.4 (3.7) | 73.5 (1.7) | 72.3 (5.7) | 73.5 (5.8) | 70.4 (3.5) | 75.1 (1.6) | N/A | 64.8 (5.1) | 70.7 (3.6) | 75.8 (1.9) | 67.6 (10.8) | 66.7 (3.9) | |||
Yes | 30.6 (3.7) | 26.5 (1.7) | 27.7 (5.7) | 26.5 (5.8) | 29.6 (3.5) | 24.9 (1.6) | N/A | 35.2 (5.1) | 29.3 (3.6) | 24.2 (1.9) | 32.4 (10.8) | 33.3 (3.9) | |||
Table 2
Baseline anthropometry and body composition characteristics of the study population according to the status of obesity and LLM defined by different ALM indexes
Characteristics | Method A (ALM/height2) | Method B (ALM/weight) | Method C (ALM/BMI) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Normal (n = 177) | Obesity (n = 731) | LLM (n = 97) | LLM with obesity (n = 100) | P-value | Normal (n = 274) | Obesity (n = 702) | LLM (n = 0) | LLM with obesity (n = 129) | P-value | Normal (n = 248) | Obesity (n = 560) | LLM (n = 26) | LLM with obesity (n = 271) | P-value | |
Height, mean, cm | 168.4 (1.0) | 166.4 (0.4) | 162.3 (1.1) | 164.0 (1.1) | < 0.001 | 166.2 (0.8) | 165.7 (0.4) | N/A | 168.2 (1.1) | 0.063 | 166.8 (0.8) | 167.9 (0.4) | 158.1 (0.9) | 160.9 (0.7) | < 0.001 |
Weight, mean, kg | 70.1 (1.0) | 84.9 (0.6) | 55.5 (1.1) | 67.1 (1.2) | < 0.001 | 64.7 (1.1) | 81.5 (0.5) | N/A | 89.0 (1.8) | < 0.001 | 65.3 (1.0) | 83.8 (0.6) | 57.5 (1.7) | 79.3 (1.4) | < 0.001 |
BMI, mean, kg/m2 | 24.6 (0.2) | 30.6 (0.2) | 21.0 (0.2) | 24.9 (0.2) | < 0.001 | 23.2 (0.2) | 29.6 (0.2) | N/A | 31.3 (0.4) | < 0.001 | 23.3 (0.2) | 29.6 (0.2) | 23.0 (0.6) | 30.5 (0.4) | < 0.001 |
Proportion with BMI ≥ 30 kg/m2 | < 0.001 | < 0.001 | < 0.001 | ||||||||||||
No | 100.0 (0.0) | 52.5 (2.2) | 100.0 (0.0) | 100.0 (0.0) | 100.0 (0.0) | 61.8 (2.1) | N/A | 40.9 (3.9) | 100.0 (0.0) | 61.0 (1.9) | 100.0 (0.0) | 51.6 (4.7) | |||
Yes | 0.0 (0.0) | 47.5 (2.2) | 0.0 (0.0) | 0.0 (0.0) | 0.0 (0.0) | 38.2 (2.1) | N/A | 59.1 (3.9) | 0.0 (0.0) | 39.0 (1.9) | 0.0 (0.0) | 48.4 (4.7) | |||
BF percentage, mean, % | 29.8 (0.5) | 39.1 (0.3) | 32.3 (0.7) | 39.9 (0.7) | < 0.001 | 30.7 (0.4) | 39.0 (0.3) | N/A | 40.3 (0.7) | < 0.001 | 30.7 (0.5) | 38.8 (0.3) | 30.0 (1.5) | 40.4 (0.5) | < 0.001 |
Proportion with BF percentage ≥ 30/42% | < 0.001 | < 0.001 | < 0.001 | ||||||||||||
No | 100.0 (0.0) | 25.8 (2.2) | 100.0 (0.0) | 20.6 (4.7) | 100.0 (0.0) | 29.5 (2.3) | N/A | 1.0 (1.0) | 100.0 (0.0) | 32.4 (2.5) | 100.0 (0.0) | 4.2 (1.4) | |||
Yes | 0.0 (0.0) | 74.2 (2.2) | 0.0 (0.0) | 79.4 (4.7) | 0.0 (0.0) | 70.5 (2.3) | N/A | 99.0 (1.0) | 0.0 (0.0) | 67.6 (2.5) | 0.0 (0.0) | 95.8 (1.4) | |||
WC, mean, cm | 89.1 (0.6) | 105.9 (0.4) | 80.7 (0.9) | 94.9 (1.1) | < 0.001 | 86.0 (0.6) | 103.3 (0.4) | N/A | 111.0 (1.3) | < 0.001 | 86.1 (0.6) | 104.3 (0.5) | 85.1 (2.0) | 105.0 (1.0) | < 0.001 |
Proportion with WC ≥ 102/88 cm | < 0.001 | < 0.001 | < 0.001 | ||||||||||||
No | 100.0 (0.0) | 60.2 (1.8) | 100.0 (0.0) | 84.4 (4.7) | 100.0 (0.0) | 68.8 (1.8) | N/A | 32.9 (5.6) | 100.0 (0.0) | 65.5 (1.8) | 100.0 (0.0) | 56.8 (4.1) | |||
Yes | 0.0 (0.0) | 39.8 (1.8) | 0.0 (0.0) | 15.6 (4.7) | 0.0 (0.0) | 31.2 (1.8) | N/A | 67.1 (5.6) | 0.0 (0.0) | 34.5 (1.8) | 0.0 (0.0) | 43.2 (4.1) | |||
ALM, mean, kg | 20.6 (0.4) | 21.2 (0.2) | 14.6 (0.4) | 15.3 (0.4) | < 0.001 | 18.4 (0.4) | 20.4 (0.2) | N/A | 20.8 (0.6) | < 0.001 | 18.6 (0.4) | 21.1 (0.2) | 15.8 (0.9) | 18.6 (0.4) | < 0.001 |
ALM/height2, mean, kg/m2 | 7.2 (0.1) | 7.5 (0.0) | 5.5 (0.1) | 5.6 (0.1) | < 0.001 | 6.5 (0.1) | 7.3 (0.1) | N/A | 7.2 (0.2) | < 0.001 | 6.6 (0.1) | 7.4 (0.0) | 6.3 (0.3) | 7.1 (0.1) | < 0.001 |
ALM/weight, mean, % | 0.3 (0.0) | 0.2 (0.0) | 0.3 (0.0) | 0.2 (0.0) | < 0.001 | 0.3 (0.0) | 0.2 (0.0) | N/A | 0.2 (0.0) | < 0.001 | 0.3 (0.0) | 0.2 (0.0) | 0.3 (0.0) | 0.2 (0.0) | < 0.001 |
ALM/BMI, mean, m2 | 0.8 (0.0) | 0.7 (0.0) | 0.7 (0.0) | 0.6 (0.0) | < 0.001 | 0.8 (0.0) | 0.7 (0.0) | N/A | 0.7 (0.0) | < 0.001 | 0.8 (0.0) | 0.7 (0.0) | 0.7 (0.0) | 0.6 (0.0) | < 0.001 |
Prevalence of LLM with obesity
We estimated the prevalence of LLM with obesity, LLM or obesity alone, and normal body phenotype based on 3 different ALMIs (Fig. 2). LLM with obesity was seen in 9.8%, 11.7%, and 19.6% of older adults with hypertension by the ALM/height2, ALM/weight, and ALM/BMI indexes, respectively. In relation to LLM alone, prevalence rates ranged from 0% by the ALM/weight index to 8.7% by the ALM/height2 index. Additionally, the respective prevalence rates of obesity were 66.7%, 64.8%, and 56.9% using these definitions. Less than one-quarter (14.9%, 23.5%, and 21.9%, respectively) of individuals had neither LLM nor obesity.
Anzeige
Fig. 2
Prevalence of LLM with obesity, LLM, obesity, and no LLM or obesity in patients with hypertension based on different ALM indexes. ALM, appendicular lean mass; BMI, body mass index; LLM, low lean mass
×
![]()
Association between LLM with obesity and all-cause mortality
After 15,148 person-years of follow-up (median duration: 15.4 years; maximum duration: 20.8 years), a total of 642 deaths (56.3%) were observed in older adults with hypertension. Figure 3 illustrates the adjusted cumulative incidence curves of all-cause mortality for various body composition phenotypes according to the ALM/height2, ALM/weight, and ALM/BMI indexes, respectively. In the fully adjusted models, LLM with obesity was significantly associated with a higher risk of all-cause mortality (HR 1.69, 95% CI 1.14–2.49, P = 0.008; HR 1.48, 95% CI 1.04–2.10, P = 0.028; HR 1.30, 95% CI 1.02–1.66, P = 0.037; respectively) compared with the normal body phenotype (Table 3). No significant differences were found in individuals with LLM or obesity alone. The adjusted HRs of each body composition phenotype for all-cause mortality were comparable among the 3 different ALMIs. No clinically relevant interaction between LLM and obesity was observed when fitting the fully adjusted model with an interaction term (P-values were 0.893 and 0.131 for the ALM/height2 and ALM/BMI indexes, respectively).
Fig. 3
Adjusted cumulative incidence curves of all-cause mortality in patients with hypertension according to the status of obesity and LLM defined by (A) ALM/height2, (B) ALM/weight, and (C) ALM/BMI, respectively. ALM, appendicular lean mass; BMI, body mass index; LLM, low lean mass
×
![]()
Table 3
HRs and 95% CIs of all-cause mortality in patients with hypertension according to the status of obesity and LLM defined by different ALM indexes
Multivariate models | Method A (ALM/height2) | Method B (ALM/weight) | Method C (ALM/BMI) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
Normal (n = 177) | Obesity (n = 731) | LLM (n = 97) | LLM with obesity (n = 100) | Normal (n = 274) | Obesity (n = 702) | LLM (n = 0) | LLM with obesity (n = 129) | Normal (n = 248) | Obesity (n = 560) | LLM (n = 26) | LLM with obesity (n = 271) | |
Deaths/person-years | 89/2516 | 408/10,308 | 72/1158 | 73/1166 | 161/3674 | 384/9947 | N/A | 97/1527 | 146/3345 | 299/8002 | 15/329 | 182/3472 |
Model 1 | 1.00 (ref.) | 1.28 0.99–1.65 P = 0.064 | 1.47 1.08-2.00 P = 0.014 | 1.65 1.09–2.49 P = 0.017 | 1.00 (ref.) | 1.05 0.86–1.27 P = 0.638 | N/A | 1.62 1.15–2.27 P = 0.005 | 1.00 (ref.) | 1.00 0.80–1.26 P = 0.971 | 0.83 0.50–1.39 P = 0.484 | 1.42 1.11–1.83 P = 0.006 |
Model 2 | 1.00 (ref.) | 1.28 1.01–1.63 P = 0.044 | 1.50 1.04–2.14 P = 0.028 | 1.77 1.22–2.58 P = 0.003 | 1.00 (ref.) | 1.05 0.87–1.27 P = 0.604 | N/A | 1.62 1.16–2.26 P = 0.004 | 1.00 (ref.) | 1.01 0.82–1.25 P = 0.928 | 0.77 0.36–1.65 P = 0.505 | 1.37 1.09–1.74 P = 0.008 |
Model 3 | 1.00 (ref.) | 1.23 0.95–1.60 P = 0.115 | 1.32 0.91–1.93 P = 0.148 | 1.69 1.14–2.49 P = 0.008 | 1.00 (ref.) | 1.08 0.88–1.33 P = 0.455 | N/A | 1.48 1.04–2.10 P = 0.028 | 1.00 (ref.) | 1.05 0.83–1.32 P = 0.682 | 0.80 0.41–1.54 P = 0.501 | 1.30 1.02–1.66 P = 0.037 |
Sensitivity analysis
When 130/80 mmHg was used as the lower threshold for hypertension, compared with individuals with normal body phenotype, the adjusted HRs (95% CI) of LLM with obesity (1.82 [1.22–2.71], 1.49 [1.07–2.07], and 1.30 [1.03–1.65], respectively) were similar to the cutoff value of 149/90 mmHg, with no significant differences seen in those with LLM or obesity alone (Table S3). Moreover, when we defined obesity exclusively using high BF percentage, LLM with obesity demonstrated a robust trend towards higher risks of mortality versus the normal body phenotype, with HRs (95% CI) of 1.54 (1.04–2.28), 1.47 (1.09–1.99), and 1.32 (1.05–1.66), respectively (Table S4). Results for LLM and obesity remained non-significant after multivariable adjustment, with the exception of the ALM/height2 index. When LLM was defined by the ALM/height2 index, the risks of all-cause mortality gradually increased from the normal body phenotype to obesity, LLM, and LLM with obesity (Ptrend = 0.014). The prevalence of LLM with obesity in the 2 sensitivity analyses was similar to the main analysis (Figure S1 and Figure S2).
Discussion
To our knowledge, this study is the first to examine the respective effects of ALM/height2, ALM/weight, and ALM/BMI in estimating the prevalence of LLM with obesity in older adults with hypertension, and to reveal the differences in mortality rates between various body composition phenotypes. The main findings are that: (1) the prevalence of LLM with obesity by the 3 different ALMIs differed significantly in older adults with hypertension, varying widely from 9.8%, 11.7%, to 19.6%; and (2) after multivariable adjustment, LLM with obesity was associated with 69%, 48%, and 30% increased risks of all-cause mortality when compared with the normal body phenotype, respectively, with no statistical differences seen in the LLM or obese groups.
There lacks a generally accepted operational definition of SO, which is now considered as a copresence of sarcopenia and obesity [51]; according to different definitions of sarcopenia or obesity, major differences may exist in anthropometric and metabolic characteristics of patients with SO. While sarcopenia has been listed as a disease by the International Classification of Diseases-10 since 2016, its diagnostic criteria are not uniform [52]. Regional expert groups, including the International Working Group on Sarcopenia [53], FNIH Sarcopenia Project [36], EWGSOP2 [34], Asian Working Group for Sarcopenia [54], and Sarcopenia Definition and Outcomes Consortium [55] have formulated a variety of definitions and diagnostic criteria for sarcopenia. Among these definitions, sarcopenia was defined by the absolute or adjusted levels of ALM (i.e., ALM/height2, ALM/weight, and ALM/BMI as measured by whole body DXA) to a greater or lesser extent, which was the primary definition used in this study.
Obesity is traditionally defined as BMI ≥ 30 kg/m2. However, considerable debate has existed concerning the efficacy of BMI in assessing the association between obesity and mortality in older adults. In a meta-analysis of 32 studies involving 197,940 older adults (aged ≥ 65 years), BMI range of 24.0-29.9 was associated with the lowest risk of mortality, and the mortality risk only began to increase when BMI exceeded ≥ 33.0 kg/m2 [56]. Ageing was associated with clinically meaningful changes in body composition, as characterized by a loss of muscle mass and an increase in visceral fat [57]. Therefore, BMI alone may not be a reliable indicator of obesity in older adults, considering that it cannot discriminate between lean mass and fat mass that have opposing effects on mortality [58]. Alternative anthropometric indicators include BF percentage (≥ 30% in males and ≥ 42% in females), WC (≥ 102 cm in males and ≥ 88 cm in females), and waist-hip ratio (≥ 0.90 in males and ≥ 0.85 in females) [59]. Considering the data availability in the NHANES, obesity was defined as any of high BMI, BF percentage, and WC in the main analysis, and was also defined by BF percentage alone in a sensitivity analysis, as previously described [40].
The prevalence of LLM with obesity in this study is hard to compare with other studies, because no related studies in patients with hypertension have been published. In a cross-sectional study in the US elderly population, the prevalence of SO varied up to 26-fold based on the different definitions used [60]. Cauley [61] demonstrated that ALM/height2, ALM/weight, and ALM/BMI yielded different distributions of SO after stratification by age, sex, and race/ethnicity; according to different populations and definitions, the prevalence of SO ranged from 0 to 41%. Meng et al. [62] compared the diagnostic efficacy of the ALM/height2 and ALM/weight indexes in Chinese older adults aged ≥ 65 years, finding that ALM/weight might be more appropriate in revealing the impact of advanced age on the prevalence of SO. A recent meta-analysis of 167,151 older adults reported the pooled prevalence of SO in the general population [63]. In their study, its prevalence was 8% and 23% based on the ALM/height2 and ALM/weight indexes, respectively, which was similar to this study. Evidence suggests that hypertension can facilitate the progression of sarcopenia via stimulating the production of multiple catabolic cytokines [64]. According to the ALM/height2 index, the prevalence of sarcopenia was approximately 20.2% in hypertensive patients aged ≥ 60 years [24]; nevertheless, the prevalence of SO is as yet unclear.
This study showed that 9.8%, 11.7%, and 19.6% of older adults with hypertension were diagnosed to have LLM with obesity by the ALM/height2, ALM/weight, and ALM/BMI indexes, respectively. Sensitivity analyses confirmed the robustness of the results. Consistent with previous studies, the prevalence of LLM with obesity by the ALM/weight and ALM/BMI indexes was numerically higher in comparison with the ALM/height2 index in our study [63]. Because the definitions of obesity were the same for the 3 diagnostic criteria and weight/BMI-based methods for ALM normalization identified rather few patients with LLM (without obesity), obesity defined by weight or BMI might be an independent risk factor for sarcopenia in older adults; these results indicated that ‘obesity paradox’ did not seem to exist in patients with SO. The number/area of intramyocellular lipid droplets and mechanical load of body weight were positively and significantly correlated with myofiber area, which led to greater absolute muscle strength and strength in obese patients [65‐67]. Therefore, when normalizing ALM with weight or BMI, the relative muscle mass and strength were weaker, thus resulting in obesity becoming a risk factor for sarcopenia [68]; in contrast, when sarcopenia was defined using the ALM/height2 index, older adults with obesity were at a relatively lower risk for sarcopenia.
From another perspective, the ALM/weight and ALM/BMI indexes clearly failed to catch LLM (without obesity), which might be explained by inherent limitations such as anthropometric indicators, statistical methodologies, and diagnostic criteria for sarcopenia and obesity. Specifically, body weight and BMI usually reflect only the sum of BF, muscle, bone, and organs, rather than BF alone [69, 70]. Considering the relatively constant mass of bone and organs, when ALM was adjusted by weight or BMI, individuals with LLM were extremely likely to have a high BF percentage. Hence, the combination of obesity and sarcopenia, as defined by the ALM/weight or ALM/BMI indexes, is of limited benefit in assessing patients with sarcopenia and without obesity. Recently, a third lumbar computed tomography (CT) scan has emerged as a promising development in identifying various body composition phenotypes [71, 72]. CT can distinguish between BF, muscle, bone, water, and air based on tissue-specific attenuation values, thus qualifying total abdominal muscle area, visceral fat area, subcutaneous fat area, and WC directly [73]. Therefore, CT scan, a routine and feasible technique in the clinical setting, may provide a more convenient and accurate diagnostic option for use in older adults with probable obesity, sarcopenia, and SO. Further research is still required to determine the optimal diagnostic criteria for sarcopenia and SO.
Accumulating evidence has demonstrated that sarcopenia was associated with 57%, 75%, and 60% increased risks of hospitalization, rehospitalization, and all-cause mortality, respectively, compared with the non-sarcopenic group [74‐76]. Unfortunately, many studies in the field did not discriminate between sarcopenia (without obesity) and SO. It could not be ruled out that a significant fraction of patients with sarcopenia in previous reports actually had SO, which should be regarded as a distinct body composition phenotype with specific clinical and metabolic characteristics. This claim was supported by results from previous studies [40]. Only a few studies have prospectively investigated the association between SO and mortality, and there were indications that older adults with SO had the highest mortality risk compared with other body phenotypes [16‐18]. This condition would be further complicated when considering the effect of SO in older adults with hypertension, as hypertension combined with either obesity or sarcopenia might have a higher mortality risk due to the synergistic effect [19, 20]. In this study involving 1105 hypertensive patients, LLM with obesity was significantly correlated with 69%, 48%, and 30% increased risks of all-cause mortality, respectively, as compared with the normal body phenotype; in contrast, no statistical differences were observed in elderly hypertensive patients with LLM or obesity alone. Consistent with our findings, Liu et al. [63] found that the risk of all-cause mortality in patients with obesity alone was comparable with those with normal BMI and without sarcopenia. The ‘obesity paradox’ correlated with mortality might be partly attributed to higher lean soft tissue mass in older adults with obesity alone [77]. For patients with LLM (without obesity), despite including 1105 older adults with hypertension, the sample size was relatively small; only 97, 0, and 26 patients were included according to the 3 ALMIs, respectively, resulting in wide CIs and less precise estimates. Results from several cohort studies showed a significantly higher risk of all-cause mortality in patients with either sarcopenia alone or with SO, with the former contradictory with our results [78, 79]. Therefore, although there was an apparent difference with respect to study population relative to other studies, further large, prospective, longitudinal comparative studies would be beneficial.
Limitations
This study has several limitations. First, sarcopenia (without obesity) was defined depending solely on muscle quantity, because data on grip strength and physical performance were inaccessible in the NHANES 1999–2004. The use of ALMI alone might not fully capture the complexity of sarcopenia and accurately reflect the true prevalence or influence of sarcopenia in hypertensive individuals. However, different definitions of sarcopenia exist in the current literature, and this definition was commonly recognized and employed in recent studies [80‐82]. Second, the findings of this study were primarily applicable to noninstitutional older adults instead of hospitalized patients that were likely to have a greater degree of sarcopenia or obesity. Third, this study was conducted in the US elderly population, presumably limiting the generalizability of the findings to patients from other countries, races/ethnicities, or demographic populations. Fourth, a number of participants (n = 1352) were precluded due to a lack of sufficient clinical data, and thus selection bias could not be completely ruled out. Finally, due to the constraint imposed by the sample size, this study did not investigate the difference in prevalence and mortality rates between men and women; however, a meta-analysis demonstrated no apparent difference in the prevalence of SO between genders in older adults [63]. Future prospective studies in a large cohort are warranted in different ethnic/racial groups.
Conclusions
The prevalence of LLM with obesity markedly differed in older adults with hypertension according to the ALM/height2, ALM/weight, and ALM/BMI indexes, amounting to 9.8%, 11.7%, and 19.6%, respectively. When compared with the normal body phenotype, LLM with obesity was significantly associated with a higher risk of all-cause mortality, with no statistical difference found in patients with LLM or obesity alone. Further large, prospective, cohort studies are required to validate these findings and uncover underlying mechanisms.
Acknowledgements
We sincerely acknowledge the personnel who contributed to the NHANES data we studied.
Declarations
Ethics approval and consent to participate
All methods were carried out in accordance with relevant guidelines and regulations (declaration of Helsinki). The National Center for Health Statistics (NCHS) Ethics Review Board has approved the NHANES protocol and written informed consent was acquired from each participant.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.