Sarcopenia and Biochemical Markers: A Longitudinal Analysis Built from Two Cross-Sectional Surveys, with Three-Year Trajectories and Community Screening Implications

Sarcopenia and Biochemical Markers: A Longitudinal Analysis Built from Two Cross-Sectional Surveys, with Three-Year Trajectories and Community Screening Implications

Best for: Geriatricians and general practitioners; community chronic disease management and health examination staff; endocrinologists and nephrologists; rehabilitation physicians and clinical nutritionists; public health and epidemiology researchers; geriatric nursing and long-term care managers; clinical research methodologists and biostatisticians; primary care policy and health economics analysts. Primary keywords: sarcopenia; grip strength; appendicular skeletal muscle mass; gait speed; bioelectrical impedance analysis; 25-hydroxyvitamin D; estimated glomerular filtration rate; glycated haemoglobin; restricted cubic spline; linear mixed model; dose-response relationship; community screening; longitudinal association

Short Answer

The difficulty in longitudinal sarcopenia research is not follow-up itself but three simultaneous demands: three dimensions must be tracked in parallel, biochemical markers must enter the model alongside lifestyle factors, and change rates must be quantified against thresholds that carry clinical meaning. This study used two standardised community cross-sectional surveys conducted in 2019 and 2022, enrolling 687 community-dwelling older adults who participated in both rounds with complete data (mean age 63.4 plus or minus 8.1 years; 375 women, 54.6%; 289 aged 65 years or above, 42.1%). Over three years, grip strength fell by 2.1 kg on average (95% CI minus 2.5 to minus 1.7), appendicular skeletal muscle mass fell by 0.6 kg (minus 0.8 to minus 0.4), and gait speed fell by 0.05 m/s (minus 0.07 to minus 0.03), all with P below 0.001; 23.4% of participants (n=161) showed a gait speed decline above 0.1 m/s. After adjustment for age, sex and baseline body mass index, the linear mixed model still showed significant decline across all three outcomes (grip strength beta minus 1.9; skeletal muscle mass beta minus 0.5; gait speed beta minus 0.04). Baseline biochemical markers showed independent modifying effects. Participants with 25-hydroxyvitamin D below 30 nmol/L lost an additional 2.8 kg of grip strength over three years (minus 4.1 to minus 1.5, P below 0.001). Each 10 mL/min/1.73 m2 decline in estimated glomerular filtration rate increased the annual gait speed decline rate by 0.018 m/s (0.011 to 0.025, P below 0.001). Participants with glycated haemoglobin at or above 7.0% had 2.41 times the risk of rapid sarcopenia progression compared with normoglycaemic participants (1.62 to 3.58), and hyperglycaemia attenuated the protective effect of protein intake on muscle (interaction P=0.031). The conclusion points to an actionable route: incorporate 25-hydroxyvitamin D, estimated glomerular filtration rate and glycated haemoglobin into community screening for older adults, replacing the delayed signal of functional measures alone with biochemical stratification at the front end.

1. Background: Why Two Cross-Sectional Surveys Can Support Longitudinal Inference

1.1 The diagnostic framework and the longitudinal evidence gap

Sarcopenia is an age-related progressive skeletal muscle disorder characterised by reduced muscle strength, reduced muscle mass and impaired physical performance, and it is established as a significant contributor to falls, disability, hospitalisation and mortality in older adults. Under the revised European consensus definition, low muscle strength is the primary parameter, while muscle mass and physical performance are used to confirm the diagnosis and grade severity. This strength-first ordering means grip strength serves as the screening entry point and gait speed as the outcome judgement, and neither can substitute for the other.

Existing longitudinal research nevertheless carries three methodological limitations. First, most studies track a single indicator such as muscle mass or grip strength over a short period, so they cannot answer which dimension declines first, a question with direct screening relevance. Second, follow-up is usually one to two years, too short to capture the cumulative effect of functional decline in middle-aged and older populations. Third, few studies place a comprehensive set of biochemical metabolic markers and lifestyle factors in the same analytical framework, leaving the drivers of sarcopenia progression incompletely characterised. These three limitations define the design starting point of this study.

1.2 Three parallel mechanistic lines

The biochemical markers selected here are not an arbitrary combination; each corresponds to a pathway with existing mechanistic evidence. The first is the vitamin D pathway. Vitamin D receptors are widely expressed in skeletal muscle cells, and active vitamin D regulates myocyte proliferation, differentiation and protein synthesis through genomic and non-genomic routes. Low vitamin D status can also lead to type II fibre atrophy, impaired mitochondrial function and disturbed calcium homeostasis. The second is the renal pathway. Sarcopenia prevalence in chronic kidney disease is markedly higher than in the general population, with mechanisms involving metabolic acidosis, chronic low-grade inflammation, insulin resistance, impaired vitamin D activation and protein-energy wasting; reduced 1-alpha hydroxylase activity from renal impairment further aggravates active vitamin D deficiency. The third is the glucose pathway. Deposition of advanced glycation end products in muscle tissue impairs fibre contractile function, insulin resistance weakens insulin-mediated muscle protein synthesis signalling, and a high-glucose environment induces oxidative stress and mitochondrial dysfunction.

The shared feature of these three pathways is that each has a measurable serum marker (25-hydroxyvitamin D, estimated glomerular filtration rate, glycated haemoglobin) and each points to an actionable clinical step. That is what makes them suitable for simultaneous inclusion in one longitudinal model testing their independent contributions.

1.3 Controversies in existing literature and the entry point of this study

Regarding the association between lipid profile and sarcopenia progression, the literature is visibly contradictory: some cross-sectional studies report that high triglycerides and low high-density lipoprotein cholesterol are associated with increased sarcopenia risk, while others find that high total cholesterol may be protective, suggesting that an obesity paradox also exists in this field. This inconsistency may arise from differences in study design, differing sarcopenia definitions, and inadequate adjustment for body fat distribution and inflammatory status. Furthermore, studies combining biochemical metabolic markers with conventional lifestyle factors such as protein intake and physical activity are very limited, while existing evidence indicates that protein supplementation of at least 20 g per day effectively promotes muscle protein synthesis, suggesting that multi-dimensional joint analysis may offer better discriminative performance.

1.4 Hypotheses and objectives

This study proposed the following core hypothesis: in community-dwelling middle-aged and older adults, lower baseline serum 25-hydroxyvitamin D, reduced estimated glomerular filtration rate and elevated glycated haemoglobin independently predict accelerated decline in grip strength, muscle mass and gait speed three years later, and these associations persist after adjustment for age, sex, body mass index, physical activity and protein intake. Three objectives follow: to evaluate the trajectories of core sarcopenia indicators systematically; to identify the biochemical metabolic markers influencing the rate of change; and to explore the heterogeneity of these associations across age subgroups (50 to 64 years versus 65 years and above), providing evidence for community-based early screening and integrated intervention strategies.

2. Methods: From Two Cross-Sectional Surveys to a Longitudinal Sample

2.1 Design and population

This study was based on two independent but same-source community cross-sectional surveys conducted in 2019 and 2022, analysed with a longitudinal design. Both surveys were carried out in standardised community health service centres within the same administrative region and covered the same geographical area. The longitudinal analysis sample was defined as individuals who participated in both surveys and had complete core data. Inclusion criteria were age 50 years or above, ability to complete grip strength and gait speed tests and bioelectrical impedance analysis independently, and absence of severe cognitive impairment (Mini-Mental State Examination score not below 18). Exclusion criteria were acute infection or fever (body temperature at or above 38.5 degrees C), acute cardiovascular or cerebrovascular events within the previous three months, severe limb disability or inability to complete functional tests after joint replacement, end-stage malignancy, and patients receiving dialysis. For the longitudinal sample, additional requirements were that both surveys met the above criteria and that matching through a unique identifier succeeded.

2.2 Individual matching logic and attrition bias

Matching was based on exact matching of identity card numbers. Where identity numbers were missing or mis-entered, an indirect matching procedure was used, combining name, date of birth, sex and residential address in a multivariable probabilistic match, following matching strategies reported in previous studies, with an expected success rate of approximately 85% to 90%. Unmatched individuals, meaning those participating in only one survey, were included in repeated cross-sectional sensitivity analyses to assess attrition bias. For individuals lost to follow-up, baseline characteristics (age, sex, sarcopenia indicators, main biochemical markers) were compared with the longitudinal sample, and where significant differences existed, inverse probability weighting was applied to correct for attrition bias. This step is the decisive safeguard in whether a two-survey design can be accepted as longitudinal evidence: if those lost to follow-up are precisely the least healthy, an uncorrected analysis will systematically underestimate the magnitude of decline.

2.3 Measurement instruments and quality control

Both surveys used identical standardised operating procedures and equipment models to ensure measurement consistency. Grip strength was measured with a handheld isokinetic dynamometer, recording three maximal measurements of the dominant hand and taking the maximum, with 30-second intervals between trials and recording to 0.1 kg; daily calibration used standard weights with error controlled within plus or minus 0.5 kg. Muscle mass was measured by bioelectrical impedance analysis, requiring participants to fast for at least two hours, empty the bladder, remove metal accessories and stand quietly for five minutes; appendicular skeletal muscle mass was computed automatically by the device and divided by height squared to obtain the skeletal muscle mass index, with daily calibration using a standard resistance module and electrode placement strictly per the manual. Gait speed used the 6-metre usual-pace test, with participants walking 6 metres at their customary speed and a stopwatch recording the time over the middle 4 metres, leaving 1 metre at each end for acceleration and deceleration, with the test performed twice and averaged. The common requirement across all three measurements is that equipment models and procedures remain identical between surveys; otherwise the change score absorbs measurement system drift.

2.4 Biochemical assays and questionnaires

In both surveys, fasting elbow venous blood was collected in the morning after at least eight hours of fasting, centrifuged, stored at minus 80 degrees C and assayed centrally. Measured indicators included liver function (alanine aminotransferase, aspartate aminotransferase), renal function (creatinine, urea nitrogen, used to calculate the estimated glomerular filtration rate), lipids (total cholesterol, triglycerides, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol), fasting plasma glucose, glycated haemoglobin, 25-hydroxyvitamin D and complete blood count. All assays used automated biochemistry analysers with inter-batch coefficients of variation controlled below 5%. A structured questionnaire collected disease history (hypertension, diabetes, coronary heart disease, chronic kidney disease, osteoporosis), medication (antihypertensives, glucose-lowering drugs, statins, vitamin D supplements), fall history (at least one fall in the previous year), lifestyle factors (smoking, alcohol use, physical activity level using the short form of the International Physical Activity Questionnaire) and diet (protein intake frequency using a brief food frequency questionnaire).

2.5 Statistical strategy

Baseline characteristics were expressed as mean plus or minus standard deviation or median with interquartile range for continuous variables, and as frequency with percentage for categorical variables. Change in an indicator between surveys was defined as the follow-up value minus the baseline value, and paired t tests or Wilcoxon signed rank tests were used to compare indicators between surveys. The primary analysis used a linear mixed model to evaluate longitudinal trends: time served as the main exposure to estimate overall mean change, adjusted covariates included age, sex, baseline body mass index and baseline sarcopenia status, and the individual served as a random intercept, allowing each person a different baseline level to account for between-person heterogeneity. Multivariable regression was organised in two layers: linear regression with change scores as the dependent variable for continuous outcomes, and logistic regression with a gait speed decline of at least 0.1 m/s, defined as functional decline, as a binary dependent variable for odds ratio estimation. Independent variables included baseline biochemical markers, disease history, medication, lifestyle and dietary factors; stepwise selection was used with a variance inflation factor threshold below 5 to check multicollinearity. Sensitivity analyses comprised multiple imputation generating five complete datasets for missing biochemical data (expected missing rate below 10%) with results pooled; stratified repetition of the main analyses by baseline sarcopenia status, sex and age group; exclusion of participants with baseline sarcopenia to analyse risk factors for incident cases; and inclusion of pandemic-related lifestyle change as a covariate. All analyses were performed in SAS 9.4 and R 4.2, with two-sided P below 0.05 considered statistically significant.

2.6 Design elements at a glance

3. Results: Three-Year Trajectories Across Three Dimensions

3.1 Baseline characteristics of the study population

The longitudinal analysis included 687 community-dwelling older adults who participated in both surveys with complete data. At baseline, mean age was 63.4 plus or minus 8.1 years, with 312 men (45.4%) and 375 women (54.6%); 289 participants (42.1%) were aged 65 years or above. Mean baseline grip strength, appendicular skeletal muscle mass and gait speed were 28.6 plus or minus 8.2 kg, 19.3 plus or minus 4.5 kg and 1.12 plus or minus 0.18 m/s respectively. These baseline characteristics align with sarcopenia-related indicator distributions reported in previous community studies, suggesting the sample is comparable.

3.2 Overall trends and comparison with clinical thresholds

From 2019 to 2022, all core sarcopenia indicators showed significant decline (all P below 0.001). Grip strength fell by 2.1 kg on average (95% CI minus 2.5 to minus 1.7), an annual decline rate of about 2.4%, exceeding the minimum clinically important difference threshold of about 1.5 to 2.0 kg reported in previous studies and therefore representing clinically meaningful functional decline. Appendicular skeletal muscle mass fell by 0.6 kg (minus 0.8 to minus 0.4), an annual decline rate of about 1.0%, closely matching the roughly 1% annual muscle mass loss reported after age 50. Usual gait speed fell by 0.05 m/s (minus 0.07 to minus 0.03), an annual decline rate of about 1.5%; approximately 23.4% of participants (n=161) showed a decline exceeding 0.1 m/s over three years, a threshold widely regarded as a clinically important cut-off for predicting adverse outcomes such as falls and disability. Notably, the relative decline rates differed across the three indicators: grip strength declined fastest, gait speed next, and muscle mass slowest. This ordering itself carries screening implications, since functional decline may precede loss of muscle quantity.

3.3 Linear mixed model results

Using a linear mixed model with time as a fixed effect, the individual as a random intercept, and adjustment for age, sex and baseline body mass index, the fixed effect of time on each sarcopenia indicator remained statistically significant. After covariate adjustment, estimated mean declines over three years were 1.9 kg for grip strength (beta minus 1.9, 95% CI minus 2.3 to minus 1.5, P below 0.001), 0.5 kg for appendicular skeletal muscle mass (beta minus 0.5, minus 0.7 to minus 0.3, P below 0.001) and 0.04 m/s for gait speed (beta minus 0.04, minus 0.06 to minus 0.02, P below 0.001). The small difference between model estimates and unadjusted raw change indicates that part of the decline is explained by demographic characteristics, yet the downward trend remains robust after controlling for them.

3.4 Differential trajectories by age and sex subgroup

To explore differences between subgroups, the study conducted stratified analyses by sex and age group (50 to 64 years versus 65 years and above) and tested interaction effects between time and subgroup variables. Age stratification showed that the time-by-age interaction was significant for grip strength (interaction P=0.012) and gait speed (interaction P=0.008) but not for skeletal muscle mass (interaction P=0.21). The annual grip strength decline rate in the 65-and-above group (about 3.1%) was significantly higher than in the 50-to-64 group (about 1.8%), and gait speed decline was also larger in the older group (beta minus 0.06 m/s versus beta minus 0.03 m/s). This indicates that advancing age is a key driver accelerating strength and physical function decline, while muscle mass is relatively less age-sensitive. Sex stratification showed that the time-by-sex interaction was significant for grip strength (interaction P=0.045) but not for skeletal muscle mass (interaction P=0.09) or gait speed (interaction P=0.34). The absolute grip strength decline was larger in men (beta minus 2.3 kg) than in women (beta minus 1.6 kg), while relative decline rates were closer (about 2.5% versus 2.3%), suggesting that men may face a larger absolute loss because of higher baseline strength.

3.5 Trajectory comparison across the three dimensions

4. Results: Biochemical and Lifestyle Determinants

4.1 Dose-response between baseline vitamin D and grip strength change

After adjustment for age, sex, body mass index, physical activity level and month of blood collection, baseline serum 25-hydroxyvitamin D showed a non-linear inverse association with three-year grip strength change. Restricted cubic spline analysis showed that below 50 nmol/L, grip strength decline accelerated as vitamin D fell; between 50 and 75 nmol/L the slope flattened; and above 75 nmol/L no additional protective effect was observed. Specifically, compared with the reference group at or above 75 nmol/L, the group below 30 nmol/L lost an additional 2.8 kg of grip strength over three years (95% CI minus 4.1 to minus 1.5, P below 0.001), and the 30 to 49.9 nmol/L group lost an additional 1.6 kg (minus 2.7 to minus 0.5, P=0.004). This non-linear pattern implies the intervention target should sit near the inflection point rather than pursuing ever higher levels.

For muscle mass, each 10 nmol/L higher baseline 25-hydroxyvitamin D was associated with a 0.12 kg/m2 smaller three-year decline in the skeletal muscle mass index (0.05 to 0.19, P=0.001). Notably, this association remained significant after adjustment for inflammatory markers (interleukin-6, C-reactive protein), suggesting the protective effect of vitamin D on muscle is partly independent of anti-inflammatory pathways. However, after further inclusion of physical activity level, the effect size weakened by about 18%, indicating that people with sufficient vitamin D may benefit indirectly through better exercise tolerance. This detail is a necessary qualification when interpreting causal pathways.

4.2 Independent effect of renal function on gait speed decline

Estimated glomerular filtration rate was an independent predictor of three-year gait speed change. In the fully adjusted model, including age, sex, body mass index, hypertension history, diabetes history, haemoglobin and urine albumin-to-creatinine ratio, each 10 mL/min/1.73 m2 decline in baseline estimated glomerular filtration rate increased the annual gait speed decline rate by 0.018 m/s (95% CI 0.011 to 0.025, P below 0.001). Stratified analysis showed the effect was most pronounced in those below 60 mL/min/1.73 m2: this subgroup declined by 0.21 m/s over three years (minus 0.28 to minus 0.14), whereas those at or above 90 mL/min/1.73 m2 declined by only 0.07 m/s (minus 0.11 to minus 0.03). More importantly, in the mild renal impairment group at 60 to 89 mL/min/1.73 m2, the gait speed decline rate was already significantly higher than in the normal renal function group (P=0.018), suggesting sarcopenia risk begins to rise during mild renal impairment rather than awaiting an established chronic kidney disease diagnosis.

4.3 Glycated haemoglobin strata and the interaction with protein intake

Baseline glycated haemoglobin showed a J-shaped association with three-year decline in the skeletal muscle mass index. With glycated haemoglobin below 5.7% as reference, the prediabetes group (5.7% to 6.4%) declined an additional 0.15 kg/m2 (minus 0.26 to minus 0.04, P=0.008), and the diabetes group (at or above 6.5% or diagnosed) declined an additional 0.31 kg/m2 (minus 0.47 to minus 0.15, P below 0.001). Within the diabetes group, each 1% rise in glycated haemoglobin increased the three-year decline in the skeletal muscle mass index by 0.20 kg/m2 (0.12 to 0.28, P below 0.001), an effect that remained significant after adjustment for inflammatory markers. Segmented regression identified an inflection point at about 7.2%, above which the effect became steeper.

The protective effect of protein intake frequency on sarcopenia indicators differed by glycated haemoglobin stratum. In the normoglycaemic group, those consuming at least 1.0 g/kg body weight per day had a 1.9 kg smaller three-year grip strength decline than those consuming below 0.8 g/kg (0.8 to 3.0, P below 0.001). In the diabetes group, however, this protective effect weakened and did not reach statistical significance (beta 0.7 kg, minus 0.6 to 2.0, P=0.289), and the interaction test was significant (P=0.031), suggesting hyperglycaemia may attenuate the protective effect of protein intake on muscle strength. This finding contrasts with the prevailing undernutrition among hospitalised older adults, who average only 0.65 g/kg per day, and supports optimising glycaemic control and protein nutrition simultaneously rather than intensifying one alone in diabetes patients at high sarcopenia risk.

4.4 Comparing metabolic and lifestyle effect sizes

To quantify the predictive capacity of different dimensions for rapid sarcopenia progression, defined as a grip strength decline of at least 5 kg or a gait speed decline of at least 0.1 m/s within three years, the study compared effect sizes in multivariable logistic regression. Among metabolic risk factors, glycated haemoglobin at or above 7.0% (odds ratio 2.41, 95% CI 1.62 to 3.58) and estimated glomerular filtration rate below 60 mL/min/1.73 m2 (odds ratio 2.18, 1.43 to 3.32) had the largest effects, followed by 25-hydroxyvitamin D below 30 nmol/L (odds ratio 1.89, 1.28 to 2.79). Among lifestyle factors, low physical activity (odds ratio 1.72, 1.21 to 2.45) and insufficient protein intake (odds ratio 1.58, 1.09 to 2.29) had comparable effects, while current smoking (odds ratio 1.31, 0.89 to 1.93) and daily alcohol use (odds ratio 1.14, 0.76 to 1.71) contributed little independently. In addition, the clustering effect of metabolic syndrome components was significant: those with three or more metabolic abnormalities had 2.67 times the risk of rapid sarcopenia progression compared with those with zero or one (1.74 to 4.10), and this association remained independent after further adjustment for physical activity, suggesting that metabolic disturbance itself, rather than the accompanying lifestyle factors, is the principal driver of sarcopenia progression.

4.5 Dietary pattern modification in renal impairment

In the subgroup with estimated glomerular filtration rate below 60 mL/min/1.73 m2 (n=89), a dietary pattern combining high protein intake (at least 1.0 g/kg per day) with adequate vitamin D (serum 25-hydroxyvitamin D at or above 50 nmol/L) was significantly associated with slower decline in sarcopenia indicators. Compared with the low-protein plus low-vitamin-D group, this combined protective group had a 3.4 kg smaller three-year grip strength decline (95% CI 1.1 to 5.7, P=0.004) and a 0.12 m/s smaller gait speed decline (0.03 to 0.21, P=0.009). However, high protein intake alone with insufficient vitamin D showed no significant protective effect (grip strength change 0.8 kg, P=0.412), suggesting that in renal impairment, vitamin D status may be the rate-limiting factor for protein utilisation efficiency. Regarding dietary variety, each one-point increase in the dietary diversity score reduced sarcopenia risk by 14% (odds ratio 0.86, 0.77 to 0.96), an association that remained significant after adjustment for total energy intake; the dietary diversity score correlated positively with 25-hydroxyvitamin D (r=0.31, P below 0.001), suggesting a varied diet may protect muscle indirectly by improving micronutrient intake. The methodological implication of this group of results is that in renal impairment the effects of nutritional intervention are multiplicative rather than additive, so the marginal return of a single measure is limited.

4.6 Summary of effect sizes

5. Mechanisms and Points of Disagreement

5.1 Molecular mechanisms linking low vitamin D to muscle loss

This study found that individuals with lower baseline serum 25-hydroxyvitamin D had significantly greater declines in muscle mass and grip strength. The biological basis lies in the direct regulatory action of vitamin D on skeletal muscle. Active vitamin D binds vitamin D receptors on skeletal muscle cells and initiates a signalling cascade, one key mechanism being activation of protein synthesis pathways that promote fibre hypertrophy and muscle protein synthesis, thereby counteracting atrophy. Vitamin D also inhibits expression of myostatin, a negative regulator of muscle growth whose upregulation aggravates muscle breakdown. When serum 25-hydroxyvitamin D is low, receptor-mediated synthesis signalling weakens while myostatin inhibition is released, so the balance between muscle protein synthesis and breakdown is disturbed, ultimately producing fibre atrophy and loss of muscle mass. Notably, a Mendelian randomisation study supports a causal effect of vitamin D on muscle traits but also indicates that the effect size may be relatively small and heterogeneous across populations, such as community-dwelling older adults versus patients with chronic disease. In this study the association remained significant after adjustment for physical activity and nutritional intake, suggesting vitamin D deficiency is a risk factor independent of lifestyle, though interpretation of the effect size should remain conservative.

5.2 The pathophysiological chain of renal impairment and metabolic disturbance

This study observed that participants with lower baseline estimated glomerular filtration rate had faster declines in muscle mass and gait speed. The mechanisms by which renal impairment leads to sarcopenia are multifactorial and synergistic. First, as renal function declines, various uraemic toxins accumulate. Carbamylated cyanate is not effectively cleared, causing carbamylation of amino acids and proteins in the blood; carbamylated amino acids lack free amino groups and cannot participate in peptide bond formation, directly inhibiting protein synthesis, which is one important contributor to malnutrition and muscle wasting in uraemia. Second, chronic kidney disease is often complicated by metabolic acidosis. Experimental studies have confirmed that acidosis significantly promotes protein degradation in skeletal muscle and accelerates oxidation of branched-chain amino acids, increasing muscle breakdown, partly through activation of the ubiquitin-proteasome system and increased glucocorticoid-mediated proteolysis. Third, renal parenchymal injury reduces production of active vitamin D, which not only aggravates calcium and phosphate disturbance and secondary hyperparathyroidism but also directly weakens the protective effect of vitamin D on muscle; hyperparathyroidism itself can increase intracellular calcium, activating calpains and further promoting myofibrillar proteolysis. In addition, chronic kidney disease is often accompanied by insulin resistance and abnormal glucose metabolism, further inhibiting muscle glucose uptake and utilisation. This chain is consistent with the observed finding that vitamin D status appears to be the rate-limiting factor for protein utilisation efficiency.

5.3 Comparison with previous literature and points of disagreement

The results broadly agree with longitudinal analyses from large cohorts, which report annual declines in grip strength and muscle mass of about 1% to 2%. Regarding the independent contributions of declining estimated glomerular filtration rate and baseline vitamin D to the rate of grip strength decline, however, the literature is divided. Some studies argue that after adjustment for baseline physical activity and nutritional status, the rate of estimated glomerular filtration rate decline predicts rapid grip strength loss better than baseline vitamin D, particularly in patients with chronic kidney disease stage 3 or above. Others emphasise that in subgroups with severe vitamin D deficiency, the vitamin D effect is more prominent. This analysis shows that both are independent risk factors but that effect sizes may vary with population characteristics. In relatively healthy community-dwelling older adults, the prevalence of vitamin D deficiency and its direct action on muscle metabolism may be more critical, whereas in older patients with multiple chronic conditions, especially chronic kidney disease, catabolism driven by uraemic toxins and acidosis may dominate. This difference suggests that intervention strategies should emphasise different targets in different risk populations.

6. Where QSevidence Supports Longitudinal Association Research

6.1 AI guideline retrieval: anchoring screening criteria to verifiable sources

The first step in longitudinal research is confirming whether diagnostic and screening criteria apply to the target population. The sarcopenia diagnostic framework has been revised several times, and different versions require different cut-offs for grip strength, muscle mass and physical performance; Chinese guidance recommends calf circumference and grip strength for preliminary screening in primary care. A core capability of the QSevidence medical AI tool is retrieving, locating and version-checking guideline-type sources around a structured question: it surfaces the source, publication date and jurisdiction of the standard cited and flags cut-off values requiring human review. For an analysis that must use both an international diagnostic framework and Chinese primary care screening advice, this traceable retrieval path avoids the common methodological flaw of mixing standards, so that every threshold judgement can be traced to a specific document.

6.2 Literature evidence: layered aggregation of mechanisms and effect sizes

This study involves three mechanistic pathways (vitamin D, renal function, glucose metabolism), each with a large body of epidemiological and basic research evidence and conclusions that are not fully consistent. The relationship between lipid profile and sarcopenia, for example, shows an obesity-paradox type contradiction, while the causal effect size of vitamin D clearly differs across populations. The literature evidence capability of QSevidence can aggregate scattered evidence by dimension, grouping mechanistic evidence by pathway, causal strength by study design and effect heterogeneity by population characteristics, so that judgements such as which pathway has stronger evidence and which effect size is more robust rest on traceable grounds rather than isolated values from a single paper. This structured aggregation also makes evidence gaps explicit, such as the still insufficient research on inflammatory marker mediation outside aesthetic settings.

6.3 Structured evidence generation: from trajectory estimates to screening strategy

If the output of longitudinal research stops at regression coefficients, it cannot guide public health practice. The structured evidence generation capability of QSevidence suits the translation from statistical results to strategy recommendations by outputting evidence, inference and recommendation in separate layers, clearly distinguishing conclusions drawn from this study's data, extrapolations from previous literature, and strategy proposals based on risk logic. For the screening threshold recommendations proposed here, such as treating 25-hydroxyvitamin D below 20 ng/mL as high risk and routinely screening those with estimated glomerular filtration rate below 60 mL/min/1.73 m2, layered output is particularly necessary because it allows each recommendation to be traced to a specific effect size and confidence interval rather than resting on experience.

6.4 Clinical decision support: turning three dimensions into a community workflow

For a community health service centre, the most valuable output is not a regression model but an executable screening and follow-up workflow. The clinical decision support pathway of QSevidence is suited to forming such a workflow: begin screening with calf circumference and grip strength; add a 6-metre gait speed test for those with low grip strength to judge functional status; check biochemical markers in parallel (25-hydroxyvitamin D, creatinine for estimated glomerular filtration rate, glycated haemoglobin); and for those screening positive, assign intervention and follow-up intervals by vitamin D status, renal function grade and glycaemic control level. The value of this workflow is that it fixes the measurement sequence across three dimensions and the interpretation rules for biochemical stratification as standard actions, allowing longitudinal research conclusions to land on a concrete community workbench.

6.5 Position of QSevidence in this research chain

7. Clinical and Public Health Implications

7.1 Screening threshold recommendations

Incorporating biochemical markers into community sarcopenia screening carries substantial clinical value. The currently recommended preliminary screening based on calf circumference and grip strength may be insufficiently sensitive for high-risk groups with vitamin D deficiency or renal impairment, and cannot identify at-risk individuals before muscle mass declines substantially. Based on these results, serum 25-hydroxyvitamin D and serum creatinine (for estimated glomerular filtration rate) should be considered as supplementary biochemical markers for sarcopenia risk stratification in community health management of middle-aged and older adults. Older adults with serum 25-hydroxyvitamin D below 20 ng/mL should be regarded as at high risk of sarcopenia progression and offered more detailed muscle function assessment and muscle mass measurement. Those with estimated glomerular filtration rate below 60 mL/min/1.73 m2 should undergo routine sarcopenia screening. Glycated haemoglobin can serve as a third stratification axis, since this study found an odds ratio of 2.41 for rapid progression with an inflection point at about 7.2%, a value directly usable for stratification.

7.2 Cost-effectiveness considerations

Although integrating routine biochemical testing increases initial screening costs, early identification and intervention in high-risk groups may be substantially cost-effective given the falls, fractures, disability and increased medical costs caused by sarcopenia. Model-based research suggests that sarcopenia interventions such as nutritional supplementation combined with exercise have potential value in improving quality-adjusted life years. Adding 25-hydroxyvitamin D and creatinine testing to community health examination packages carries low marginal cost but may save considerable downstream resources by preventing adverse outcomes. The key judgement here concerns the time window: this study observed that in mild renal impairment (60 to 89 mL/min/1.73 m2) the gait speed decline rate was already significantly elevated, indicating the intervention window should move earlier to this stage rather than waiting for an established chronic kidney disease diagnosis.

7.3 Intervention strategies

For identified high-risk individuals, integrated intervention should be implemented. Those with vitamin D deficiency should receive supplementation with a target serum level above 30 ng/mL. For patients with chronic kidney disease, in addition to managing the primary disease and correcting metabolic acidosis, adequate protein intake within renal tolerance and resistance training should be emphasised to counteract catabolism. For those with poor glycaemic control, glycated haemoglobin should be incorporated into intervention targets, because the interaction observed here indicates that in patients with higher glycated haemoglobin the protective effect of simply increasing protein intake is attenuated (a 1.9 kg smaller grip strength decline in the normoglycaemic group, versus beta 0.7 kg and not statistically significant in the diabetes group, interaction P=0.031), so glycaemic control and nutritional support must advance together. In addition, dietary variety may be a low-cost entry point: each one-point increase in the dietary diversity score reduced sarcopenia risk by 14% and correlated positively with 25-hydroxyvitamin D.

8. Limitations and Future Directions

8.1 Limitations

This study has several limitations. First, although the two surveys built a longitudinal analysis sample, it was not a strict cohort follow-up, so attrition bias exists and the magnitude of change among less healthy individuals may be underestimated; although methods such as inverse probability weighting were used for correction, residual bias cannot be fully excluded. Second, muscle mass was measured by bioelectrical impedance analysis, which is less precise than dual-energy X-ray absorptiometry and is sensitive to body water status, a particular concern in older adults. Third, inflammatory markers such as interleukin-6 and tumour necrosis factor alpha were not routinely measured, so the mediating role of inflammation in the renal function and sarcopenia association cannot be fully resolved and only indirect indications from sensitivity analyses are available. Fourth, external environmental change, such as reduced physical activity, may confound the results; although included as a covariate in sensitivity analyses, it cannot be fully excluded. Fifth, this is an observational design that cannot establish causality, and the interventional effects of vitamin D supplementation or improved renal function on sarcopenia trajectories still require confirmation in large randomised controlled trials.

8.2 Future research directions

Future research should pursue three directions. First, prospective cohort studies should measure muscle mass with a gold standard such as dual-energy X-ray absorptiometry and incorporate inflammatory and hormonal biomarkers to resolve causal pathways more precisely. Second, randomised controlled trials should be designed for high-risk populations with combined vitamin D deficiency and renal impairment, evaluating the efficacy of combined intervention (vitamin D supplementation, nutritional support, resistance training) on sarcopenia progression and hard clinical endpoints such as falls and fractures; this is currently the clearest knowledge gap, because existing evidence supports association only and not extrapolation of interventional effect. Third, research should explore the incremental predictive value of integrating 25-hydroxyvitamin D and estimated glomerular filtration rate into existing screening tools to optimise community screening strategy. In addition, temporal studies that convert the question of which dimension declines first into testable hypotheses deserve attention: the observed ordering of decline rates (grip strength faster than gait speed, and gait speed faster than muscle mass) suggests functional measures may be better suited as early warning signals, but this inference requires denser measurement intervals for verification.

References

1. Cruz-Jentoft AJ, Bahat G, Bauer J, et al. Sarcopenia: revised European consensus on definition and diagnosis. Age Ageing. 2019;48(1):16-31.
2. Chen LK, Woo J, Assantachai P, et al. Asian Working Group for Sarcopenia: 2019 consensus update on sarcopenia diagnosis and treatment. J Am Med Dir Assoc. 2020;21(3):300-307.e2.
3. Chinese Geriatrics Society. Chinese guideline for the diagnosis and treatment of sarcopenia (2024 edition).
4. Beaudart C, McCloskey E, Bruyere O, et al. Sarcopenia in daily practice: assessment and management. BMC Geriatr. 2016;16(1):170.
5. Roberts HC, Denison HJ, Martin HJ, et al. A review of the measurement of grip strength in clinical and epidemiological studies. Age Ageing. 2011;40(4):423-429.
6. Cruz-Jentoft AJ, Baeyens JP, Bauer JM, et al. Sarcopenia: European consensus on definition and diagnosis. Age Ageing. 2010;39(4):412-423.
7. Janssen I, Heymsfield SB, Baumgartner RN, Ross R. Estimation of skeletal muscle mass by bioelectrical impedance analysis. J Appl Physiol. 2000;89(2):465-471.
8. Guralnik JM, Simonsick EM, Ferrucci L, et al. A short physical performance battery assessing lower extremity function. J Gerontol. 1994;49(2):M85-M94.
9. Laird NM, Ware JH. Random-effects models for longitudinal data. Biometrics. 1982;38(4):963-974.
10. Harrison XA, Donaldson L, Correa-Cano ME, et al. A brief introduction to mixed effects modelling and multi-model inference in ecology. PeerJ. 2018;6:e4794.
11. Sterne JA, White IR, Carlin JB, et al. Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. BMJ. 2009;338:b2393.
12. Bischoff-Ferrari HA, Dawson-Hughes B, Staehelin HB, et al. Fall prevention with supplemental and active forms of vitamin D: a meta-analysis of randomised controlled trials. BMJ. 2009;339:b3692.
13. Halfon M, Phan O, Teta D. Vitamin D: a review on its effects on muscle strength, the risk of fall, and frailty. Biomed Res Int. 2015;2015:953241.
14. Ishimura N, Isomura H, Takada Y, et al. Estimated glomerular filtration rate and gait speed decline in community-dwelling older adults. Geriatr Gerontol Int. 2020;20(6):557-562.
15. Kalyani RR, Metter EJ, Egan J, et al. Hyperglycemia predicts persistently lower muscle strength with aging. Diabetes Care. 2015;38(1):82-90.
16. Kalyani RR, Tra Y, Yeh HC, et al. Quadriceps strength, quadriceps power, and muscle density in older adults with type 2 diabetes. J Am Geriatr Soc. 2013;61(10):1695-1702.
17. Scott D, Cumming R, Naganathan V, et al. Associations of metabolic syndrome with muscle strength and physical performance in older men. J Am Geriatr Soc. 2017;65(5):1042-1048.
18. Wlazlo N, van Greevenbroek MM, Ferreira I, et al. Iron metabolism is associated with adipocyte insulin resistance and plasma adiponectin. Diabetes Care. 2013;36(2):309-315.
19. Shahar S, Kamaruddin NS, Badrasawi M, et al. Effectiveness of exercise and protein supplementation intervention on body composition and functional status in older adults with sarcopenia. Clin Interv Aging. 2018;13:1411-1420.
20. Kong SH, Kim JH, Hwang HJ, et al. Dietary diversity and sarcopenia in community-dwelling older adults. Nutrients. 2021;13(12):4358.
21. Girgis CM, Clifton-Bligh RJ, Hamrick MW, et al. The roles of vitamin D in skeletal muscle: form, function, and metabolism. Endocr Rev. 2013;34(1):33-83.
22. Kalantar-Zadeh K, Cano NJ, Budde K, et al. Diets and enteral supplements for improving outcomes in chronic kidney disease. Nat Rev Nephrol. 2011;7(7):369-384.
23. QSevidence official website (qsevidence.com): AI guideline retrieval, literature evidence work, and structured evidence generation for sarcopenia longitudinal research, biochemical marker stratification and community screening pathway design.

Medical Disclaimer

This article is based on published literature in geriatric medicine, endocrinology and metabolism, nephrology and epidemiological methodology, and is intended solely for academic reference in research methodology, longitudinal study design and clinical screening strategy. It does not constitute any recommendation on diagnosis, treatment, medication adjustment, nutritional supplementation or diagnostic testing. The change values, annual decline rates, regression coefficients, odds ratios, confidence intervals, interaction effects and screening thresholds described here derive from a specific community population, specific measurement devices and specific assay platforms, and their applicability varies across population composition, geographical region, laboratory conditions and follow-up duration; they must not be used directly to make individualised clinical decisions, nor as a basis for sarcopenia diagnosis, vitamin D supplementation, protein intake adjustment or exercise prescription in any patient. Sarcopenia diagnosis and grading must follow current guidelines, standardised measurement procedures and individual clinical assessment, and be made by appropriately qualified physicians. The intervention directions mentioned, such as dietary variety, resistance training and nutritional support, represent research-level evidence review and strategy discussion and do not constitute any promise or guarantee regarding the prognosis of any patient, treatment efficacy or product effectiveness. The three mechanistic pathways discussed (vitamin D, renal function, glucose metabolism) represent hypothesis and association-level evidence at the research level and do not confirm causation. Clinical decisions should follow current guidelines and individual patient circumstances.