Abstract
Objective
To investigate the association between the Dietary Inflammation Index (DII) and sarcopenia in older adults, providing theoretical support for clinical prevention and treatment.
Methods
Elderly individuals aged 60 years and older in Urumqi were selected. Data were collected through questionnaires and physical examinations to assess sarcopenia prevalence. Logistic regression analyzed the relationship between DII and sarcopenia. A stepwise logistic regression model incorporating age, BMI, waist circumference, DII, and PSQI total score was constructed for prediction.
Results
Each 1-unit increase in DII significantly elevated sarcopenia risk (OR = 2.02, 95% CI: 1.49 to 2.73). The predictive model demonstrated good discrimination (AUC = 0.736 in training set, AUC = 0.742 in testing set) and calibration (Hosmer-Lemeshow test P > 0.05). Decision curve analysis indicated significant clinical net benefit within the 15% to 45% risk threshold range.
Conclusion
Elevated dietary inflammation index represents an independent risk factor for sarcopenia in older adults. The multifactorial prediction model demonstrates robust discriminatory capability and clinical utility. Optimizing dietary patterns to reduce DII is recommended for sarcopenia prevention.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12877-026-07108-3.
Keywords: Dietary inflammatory index, Elderly, Sarcopenia, Prediction model, Clinical decision curve
Introduction
Sarcopenia, first proposed by Rosenberg in 1989, is a degenerative disorder characterized by age-related loss of skeletal muscle mass and impairment of skeletal muscle function, constituting a complex syndrome [1]. Sarcopenia primarily manifests as diminished muscle mass, strength, and function. It carries adverse consequences including falls, fractures, disability, frailty, impaired activities of daily living, reduced quality of life, and increased mortality risk, while also elevating hospitalization rates, healthcare costs, and familial/societal burdens [2]. Research indicates that after age 50, muscle mass decreases by 1% to 2% annually, while muscle strength declines by 1.5% to 5% per year. Epidemiological data also confirm sarcopenia’s prevalence among the elderly, with its incidence rising with advancing age [3, 4].
The Dietary Inflammatory Index (DII) [5] was developed by researchers at the University of South Carolina within the Cancer Prevention and Control Program. It is a literature-derived dietary tool used to assess the overall inflammatory potential of an individual’s diet. In recent years, the DII has been extensively applied in clinical research, primarily focusing on cancer studies with a particular emphasis on cardiovascular disease [6, 7]. An Italian study on esophageal cancer found [8] that a pro-inflammatory diet increases the risk of esophageal squamous cell carcinoma. The DII was developed by an international research team that compiled dietary data from representative populations across 11 countries [9], calculating robust global daily intake means and standard deviations for 45 dietary components. Furthermore, based on 1,943 English-language publications examining the effects of dietary components on inflammation-related factors (such as IL-1β and TNF-α), the team assigned an Effect Score (e) to each of the 45 dietary components. This score ranges from − 1 to 1, with higher positive values indicating greater pro-inflammatory tendencies and higher absolute negative values indicating greater anti-inflammatory tendencies [9]. By quantifying the anti-inflammatory or pro-inflammatory properties of foods, the DII reflects how an individual’s dietary pattern influences chronic inflammatory status. Chronic inflammation is one of the key mechanisms underlying sarcopenia.
In recent years, an increasing number of studies have explored the relationship between dietary patterns and muscle health, suggesting that improving dietary quality may help slow the progression of muscle loss [10]. The Dietary Inflammatory Index (DII), as a comprehensive tool for evaluating the inflammatory potential of diets, has gradually gained attention. Multiple studies have indicated a significant association between DII and sarcopenia. A meta-analysis incorporating 11 observational studies involving 19,954 participants demonstrated that higher DII scores significantly increased the risk of sarcopenia. Each one-point increase in DII score elevated the risk of sarcopenia by 20% (OR = 1.22, 95% CI: 1.12 to 1.33, p < 0.05) [11]. Another study indicated that in Crohn’s disease patients, a pro-inflammatory diet (high DII score) is an independent risk factor for sarcopenia, particularly low muscle mass, with the association remaining highly significant after thorough adjustment for confounders (OR approaching 10-fold at its peak) [12]. A study using National Health and Nutrition Examination Survey (NHANES) data in patients with ischemic heart disease (IHD) found significantly higher DII scores in the sarcopenia group compared to the non-sarcopenia group (0.24 vs. -0.17, p = 0.020) [13]. These findings suggest that controlling dietary inflammation levels may hold particular significance for maintaining muscle health in chronic disease patients. However, the current body of research remains limited, and the underlying mechanisms are not fully elucidated. There is an urgent need for more high-quality epidemiological and experimental studies to further reveal the pathways through which DII influences muscle health across different populations and to provide a theoretical basis for preventing sarcopenia through dietary interventions.
This study will investigate the prevalence of sarcopenia among elderly individuals aged 60 and above in Urumqi, Xinjiang, using stratified cluster sampling. Multiple methodologies, including questionnaire surveys (covering basic information and dietary intake) and physical examinations, will be employed. The research aims to analyze the association between the Dietary Inflammatory Index and sarcopenia. The findings will elucidate the mechanisms underlying the development of sarcopenia in the elderly, providing a theoretical foundation for the prevention and treatment of sarcopenia among older adults in Xinjiang.
Objects and methods
Research subjects
The study selected a natural population of individuals aged 60 and above in Urumqi, Xinjiang as the survey subjects. Using a cluster sampling method, six communities were randomly chosen from the densely populated districts of Tianshan, Saybag, Xinshi, and Shuimogou in Urumqi. Elderly residents aged 60 and above who met the inclusion and exclusion criteria were surveyed within these communities.
Inclusion criteria: (1) Aged 60 years or older; (2) Permanent residents who have lived in Urumqi for ≥ 10 years; (3) Individuals who provided informed consent for the survey.
Exclusion criteria: (1) Diagnosed with Alzheimer’s disease, Parkinson’s disease, or mental disorders that impair normal communication; (2) Individuals with hearing or visual impairments; (3) Those with physical disabilities, amputations, or hip replacements that restrict physical activity; (4) Patients with hyperthyroidism, hypothyroidism, heart or renal failure, or those taking long-term steroid medications; (5) Individuals who have taken weight-loss drugs or glucocorticoids within the past three weeks, which may affect body composition analysis; (6) Those with implanted metal stents or pacemakers that interfere with body composition analysis; (7) Patients diagnosed with sleep apnea syndrome.
Formula for estimating the sample size based on the prevalence rate from the baseline survey: 
Estimate the survey sample size. Where: 
, 
, 
(based on literature reports indicating a prevalence rate of approximately 13% for sarcopenia in China) [14], yielding a survey population of 1087 individuals. In accordance with cluster sampling requirements, the sample size is increased by 10%, resulting in a minimum required survey population of 1200 individuals. The final sample size is 1370 individuals. Ethics approval number: XJYKDXR20240314047.
Research methods
Questionnaire survey
Prior to the survey, investigators explained the significance, purpose, and content of the study to the participants. After obtaining their consent, participants were asked to sign an informed consent form. Questionnaires were administered via investigator interviews.
①Basic information questionnaire
Collected basic demographic characteristics, including name, gender, date of birth, place of residence, occupation, education level, marital status, and economic status; personal disease history and family history of diabetes; and behavioral risk factors (smoking and alcohol consumption).
②Dietary survey
A Semi-quantitative Food Frequency Questionnaire (FFQ) [15] was used to assess the dietary nutritional status of participants. The FFQ was developed by referencing the “Food Frequency Questionnaire” from the 2002 China National Nutrition and Health Survey, the 2010 China National Nutrition and Health Surveillance “Food Frequency Questionnaire,” and simplified FFQs from previous large-scale epidemiological studies. It was adapted to incorporate the regional dietary characteristics of Xinjiang and the eating habits of the elderly population. Participants were asked whether they consumed specific foods over the past year, along with the frequency and average portion size. Energy and nutrient intakes were calculated using a food composition table. The analysis focused on the intake of different food types and nutrients and their association with sarcopenia in the elderly.
DII and its calculation [16]
Dietary intake data of the surveyed population were obtained via the SFFQ. The nutrient content per 100 g of each food item was converted based on the “China Food Composition Table (Standard Edition)” [17] (6th Edition), covering 20 nutrients including energy, protein, fat, carbohydrates, dietary fiber, cholesterol, vitamin A, carotene, thiamine, riboflavin, niacin, vitamin C, vitamin E, vitamin D, magnesium, iron, zinc, selenium, folate, and alcohol. The average daily intake of each food item for each subject was then combined with the respective food’s nutrient content matrix using R software to compute the average daily intake of the 20 nutrients for each study subject.
DII calculation [18]
Based on the Dietary Inflammatory Index (DII) calculation method established by Shivappa et al. [5], the DII score for each participant in this study was computed according to the following procedure: (1) Individual food intake data were collected using a food frequency questionnaire, and the daily intake of each nutrient was calculated according to the China Food Composition Table (Standard Edition), 6th edition. (2) The individual daily nutrient intake was compared with the corresponding global mean intake and standard deviation from a global dietary database to derive a standardized Z-score [Z = (individual intake – global mean intake) / global standard deviation]. (3) To correct for the right-skewed distribution commonly observed in dietary data and to achieve a symmetrical distribution of component scores ranging from − 1 to + 1, each Z-score was converted to its corresponding cumulative probability (P) using the standard normal cumulative distribution function. This probability value was subsequently linearly transformed to obtain a centered proportion (centered proportion = P × 2–1), where 0 represents the global mean intake level. (4) This centered proportion was multiplied by the food parameter-specific inflammatory effect score (see Table 1), which is derived from the literature evidence (positive weights for pro-inflammatory components and negative weights for anti-inflammatory components), thereby yielding the component-specific DII score. Finally, the component-specific DII scores for all examined nutrients were summed to obtain an overall DII score that reflects the inflammatory potential of the individual’s overall diet. A higher DII score indicates a more pro-inflammatory dietary pattern, whereas a lower score suggests a more anti-inflammatory tendency.
Table 1.
Scores of food-specific inflammatory effects
| Food parameter | Weighted number of articles |
Raw inflammatory effect score |
Overall inflammatory effect scoret | Global daily mean Intake (units/d) |
SD |
|---|---|---|---|---|---|
| Alcohol (g) | 417 | -0.278 | -0.278 | 13.98 | 3.72 |
| Vitamin B12 (µg) | 122 | 0.205 | 0.106 | 5.15 | 2.70 |
| Vitamin B6 (mg) | 227 | -0.379 | -0.365 | 1.47 | 0.74 |
| β-Carotene (µg) | 401 | -0.584 | -0.584 | 3718 | 1720 |
| Caffeine (g) | 209 | -0.124 | -0.110 | 8.05 | 6.67 |
| Carbohydrate (g) | 211 | 0.109 | 0.097 | 272.2 | 40.0 |
| Cholesterol (mg) | 75 | 0.347 | 0.110 | 279.4 | 51.2 |
| Energy (kcal) | 245 | 0.180 | 0.180 | 2056 | 338 |
| Eugenol (mg) | 38 | -0.868 | -0.140 | 0.01 | 0.08 |
| Total fat (g) | 443 | 0.298 | 0.298 | 71.4 | 19.4 |
| Fibre (g) | 261 | -0.663 | -0.663 | 18.8 | 4.9 |
| Folic acid (µg) | 217 | -0.207 | -0.190 | 273.0 | 70.7 |
| Garlic (g) | 277 | -0.412 | -0.412 | 4.35 | 2.90 |
| Ginger (g) | 182 | -0.588 | -0.453 | 59.0 | 63.2 |
| Fe (mg) | 619 | 0.032 | 0.032 | 13.35 | 3.71 |
| Mg (mg) | 351 | -0.484 | -0.484 | 310.1 | 139.4 |
| MUFA (g) | 106 | -0.019 | -0.009 | 27.0 | 6.1 |
| Niacin (mg) | 58 | -1.000 | -0.246 | 25.90 | 11.77 |
| n-3 Fatty acids (g) | 2588 | -0.436 | -0.436 | 1.06 | 1.06 |
| n-6 Fatty acids (g) | 924 | -0.159 | -0.159 | 10.80 | 7.50 |
| Onion (g) | 145 | -0.490 | -0.301 | 35.9 | 18.4 |
| Protein (g) | 102 | 0.049 | 0.021 | 79.4 | 13.9 |
| PUFA (g) | 4002 | -0.337 | -0.337 | 13.88 | 3.76 |
| Riboflavin (mg) | 22 | -0.727 | -0.068 | 1.70 | 0.79 |
| Saffron (g) | 33 | -1.000 | -0.140 | 0.37 | 1.78 |
| Saturated fat (g) | 205 | 0.429 | 0.373 | 28.6 | 8.0 |
| Se (µg) | 372 | -0.191 | -0.191 | 67.0 | 25.1 |
| Thiamin (mg) | 65 | -0.354 | -0.098 | 1.70 | 0.66 |
| Trans fat (g) | 125 | 0.432 | 0.229 | 3.15 | 3.75 |
| Turmeric (mg) | 814 | -0.785 | -0.785 | 533.6 | 754.3 |
| Vitamin A (RE) | 663 | -0.401 | -0.401 | 983.9 | 518.6 |
| Vitamin C (mg) | 733 | -0.424 | -0.424 | 118.2 | 43.46 |
| Vitamin D (µg) | 996 | -0.446 | -0.446 | 6.26 | 2.21 |
| Vitamin E (mg) | 1495 | -0.419 | -0.419 | 8.73 | 1.49 |
| Zn (mg) | 1036 | -0.313 | -0.313 | 9.84 | 2.19 |
| Green/black tea (g) | 735 | -0.536 | -0.536 | 1.69 | 1.53 |
| Flavan-3-ol (mg) | 521 | -0.415 | -0.415 | 95.8 | 85.9 |
| Flavones (mg) | 318 | -0.616 | -0.616 | 1.55 | 0.07 |
| Flavonols (mg) | 887 | -0.467 | -0.467 | 17.70 | 6.79 |
| Flavonones (mg) | 65 | -0.908 | -0.250 | 11.70 | 3.82 |
| Anthocyanidins (mg) | 69 | -0.449 | -0.131 | 18.05 | 21.14 |
| Isoflavones (mg) | 484 | -0.593 | -0.593 | 1.20 | 0.20 |
| Pepper (g) | 78 | -0.397 | -0.131 | 10.00 | 7.07 |
| Thyme/oregano (mg) | 24 | -1.000 | -0.102 | 0.33 | 0.99 |
| Rosemary (mg) | 9 | -0.333 | -0.013 | 1.00 | 15.00 |
③Pittsburgh Sleep Quality Index (PSQI)
Developed by Buysse et al. [19], the PSQI was used to assess sleep quality over the past month. The scale consists of 24 items across 7 dimensions, with a total score ranging from 0 to 21. Higher scores indicate poorer sleep quality. A total score < 7 suggests no sleep disturbance, while a score ≥ 7 indicates the presence of sleep disturbance. In this study, the Cronbach’s alpha coefficient for the scale was 0.807.
Diagnostic criteria for sarcopenia
According to the 2019 consensus of the Asian Working Group for Sarcopenia (AWGS), the diagnostic criteria for sarcopenia include reduced appendicular skeletal muscle mass, accompanied by decreased muscle strength or impaired physical performance [20].
①Appendicular Skeletal Muscle Mass (ASM)
ASM was measured using the Korean dual bioelectrical impedance analysis device (InBody 770). The obtained ASM represents the sum of skeletal muscle mass in both upper and lower limbs, normalized by height (Ht). The normalization formula is: ASM/Ht². Low skeletal muscle mass was defined as ASM/Ht² < 7.0 kg/m² for men and < 5.7 kg/m² for women [20].
②Handgrip strength
Handgrip strength was measured using a German JOTTM electronic handgrip dynamometer on the dominant hand. The grip span was adjusted to an appropriate range before testing. During the test, participants sat upright with their upper arms and forearms forming a 90-degree angle. The testing arm was slightly abducted but not exceeding 30 degrees. Participants squeezed the dynamometer with maximum force, and the test was performed twice for each hand. The highest value was recorded. If the difference between the two hands was small, the right-hand data were used for statistical analysis. Handgrip strength was considered reduced if < 28 kg for men and < 18 kg for women [20].
③Gait speed
The 6-meter gait speed test was used to assess physical performance. A 12-meter straight distance was marked with colored tape, indicating the starting point, 3-meter point, 9-meter point, and endpoint. Timing began when the participant reached the 3-meter mark and ended at the 9-meter mark. The test was conducted three times, and the fastest result was used for statistical analysis. Gait speed < 1.0 m/s was defined as reduced physical performance [20].
Statistical methods
The questionnaires were returned and entered in a uniform Epidata double pass for initial sorting and screening of invalid questionnaires, and analyzed using SPSS 25.0 statistical software, and the quantitative information was described using (
), the data that did not satisfy normal distribution were described using M (P25, P75), and comparisons between groups were made using the rank-sum test. Count data were described using percentages, and qualitative data were compared between groups using chi-square test; quantitative data that did not obey normal distribution were used or rank sum test, and multifactorial analyses were analyzed using logistic regression analysis at the test level of ɑ, and P<0.05 was considered statistically significant difference. The 1370 patients were randomly divided into training set and validation set in the ratio of 7:3. The potential risk factors for sarcopenia were first identified by univariate analysis, and then these factors were included in the training set, and the predictive model was constructed by multifactor logistic regression. After the model was constructed, column-line plots were drawn using the rms package for easy visualization of the predictions. To test the reliability of the model, it was evaluated from three perspectives: first, to draw the subject work characteristics (ROC) curve and calculate the area under the curve (AUC) to determine whether the model could accurately distinguish between diseased and non-diseased conditions; second, to draw the calibration curve to see how well the model’s predicted probability matched the actual onset of the disease; and third, to carry out a decision curve analysis (DCA) to calculate how much actual value the model could bring much practical value to clinical decision-making. All statistical tests were performed using two-sided tests, and a significant difference between the two groups of data was recognized when the P value was less than 0.05.
Main research results and conclusions
Basic information
A total of 1,370 elderly people over 60 years of age were investigated, and the prevalence of sarcopenia was 14.2% (194 cases), with its risk increasing significantly with age (P < 0.001). With regard to social factors, the highest prevalence was found in the widowed population (29.2%, χ²=46.292, P < 0.001), the risk was higher in pensioners than in wage earners, and the risk was significantly higher in nursing home residents and in those living with their children than in couples living together. Metabolic and physiological indicators showed that BMI, waist circumference, diastolic blood pressure and uric acid were significantly lower in the sarcopenia group, accompanied by increased inflammatory indices, decreased hemoglobin and increased creatinine. Lifestyle, alcohol drinkers were at lower risk, while poor sleep quality and chronic disease burden (P = 0.043) were also significantly associated with the disease. Specific results are shown in Table 2.
Table 2.
Comparison of clinical characteristics between the sarcopenia group and non-sarcopenia group 
| Variables | Non sarcopenia group (1176) | Sarcopenia group (194) | statistical value | P |
|---|---|---|---|---|
| Age | 65.00(61.00, 70.00) | 70.00(64.00, 74.00) | -7.961 | <0.001 |
| Gender | ||||
| Male [cases (%)] | 493(84.6%) | 90(15.4%) | 1.361 | 0.243 |
| Female [cases (%)] | 683(86.8%) | 104(13.2%) | ||
| Sources of Income | ||||
| Wages | 91(97.8%) | 2(2.2%) | 12.062 | 0.007 |
| Pensions | 1071(84.9%) | 190(15.1%) | ||
| Low-income | 1(100.0%) | 0(0.0%) | ||
| Other | 13(86.7%) | 2(13.3%) | ||
| Marriage | ||||
| Married | 1002(88.4%) | 131(11.6%) | 46.292 | <0.001 |
| Divorced | 25(92.6%) | 2(7.4%) | ||
| Widowed | 148(70.8%) | 61(29.2%) | ||
| Unmarried | 1(100.0%) | 0(0.0%) | ||
| Education | ||||
| Middle school and below | 137(86.7%) | 21(13.3%) | 5.650 | 0.059 |
| High school and secondary | 592(83.7%) | 115(16.3%) | ||
| College and above | 447(88.5%) | 58(11.5%) | ||
| Smoking | ||||
| No | 881(85.8%) | 146(14.2%) | 0.010 | 0.919 |
| Yes | 295(86.0%) | 48(14.0%) | ||
| Alcohol consumption | ||||
| No | 1093(85.2%) | 190(14.8%) | 6.989 | 0.008 |
| Yes | 83(95.4%) | 4(4.6%) | ||
| Employment Status | ||||
| Full-Time Employment | 68(97.1%) | 2(2.9%) | 7.903 | 0.019 |
| Part-Time Employment | 24(82.8%) | 5(17.2%) | ||
| Separated or Retired | 1084(85.3%) | 187(14.7%) | ||
| Residence Status | ||||
| Individuals living alone | 94(81.0%) | 22(19.0%) | 26.102 | <0.001 |
| Couples living together | 893(88.6%) | 115(11.4%) | ||
| Living with children | 186(77.2%) | 55(22.8%) | ||
| Nursing homes (welfare homes) | 3(60.0%) | 2(40.0%) | ||
| DII | -0.11(-1.39, 1.01) | 0.90(-0.48, 2.13) | -7.313 | <0.001 |
| PSQI | 5.00(3.00, 8.00) | 6.00(4.00, 10.00) | -4.192 | <0.001 |
| Number of chronic diseases | 2.00(1.00, 3.00) | 2.00(1.00, 3.00) | -2.027 | 0.043 |
| Cr(µmol/L) | 72.50(64.00, 83.00) | 75.50(65.00, 91.25) | -2.091 | 0.037 |
| UA(µmol/L) | 321.00(271.00, 377.00) | 300.00(252.00, 358.00) | -3.184 | 0.001 |
| glucose content | 4.98(4.59, 5.46) | 5.02(4.62, 5.70) | -0.661 | 0.509 |
| Hb(g/L) | 140.00(131.00, 150.00) | 138.00(128.00, 149.00) | -2.246 | 0.025 |
| TG(mmol/L) | 1.43(1.05, 2.00) | 1.44(1.06, 1.87) | -0.052 | 0.958 |
| TC(mmol/L) | 5.17(4.42, 5.88) | 5.16(4.39, 5.94) | -0.047 | 0.962 |
| HDL-C(mmol/L) | 1.39(1.21, 1.62) | 1.45(1.25, 1.61) | -1.314 | 0.189 |
| LDL-C(mmol/L) | 3.05(2.59, 3.61) | 3.05(2.62, 3.43) | -0.580 | 0.562 |
| Systolic | 141(130, 154) | 139(129, 153) | -0.658 | 0.510 |
| Diastolic | 83.00(76.00, 90.00) | 81(73, 88) | -2.637 | 0.008 |
| BMI | 25.12(23.21, 27.35) | 22.95(20.53, 25.19) | -7.998 | <0.001 |
| Waist circumference | 88.00(81.00, 95.00) | 85.00(77.75, 93.00) | -3.252 | 0.001 |
Multifactor logistic regression analysis
The outcome variable for the analysis was the presence of sarcopenia in the population (0 = No, 1 = Yes), and then all significant variables with P-values less than 0.05 in the univariate analyses were included in the multivariate logistic regression model. A diagnosis of covariance was made before the analysis, and it was found that the VIF values of all the independent variables did not exceed 10, which indicated that there was no serious problem of “mutual interference” between these variables. For the categorical variables, gender was assigned as 1 for males and 2 for females in the reference group, and continuous variables such as BMI, blood glucose, and duration of illness were directly analyzed in the model.Specific results are shown in Table 3.
Table 3.
Multifactor logistic regression analysis
| Variables | B | S.E. | Waldχ2 | OR | 95%CI. | P-value |
|---|---|---|---|---|---|---|
| Constants | -4.740 | 1.208 | 15.392 | 0.009 | <0.001 | |
| Age | 0.070 | 0.014 | 24.270 | 1.073 | (1.043, 1.103) | <0.001 |
| BMI | -0.253 | 0.036 | 48.955 | 0.777 | (0.724, 0.834) | <0.001 |
| Waist circumference | 0.042 | 0.012 | 12.090 | 1.043 | (1.019, 1.068) | 0.001 |
| Total PSQI score | 0.048 | 0.022 | 4.835 | 1.049 | (1.005, 1.095) | 0.028 |
| DII | 0.307 | 0.056 | 30.392 | 1.359 | (1.219, 1.516) | <0.001 |
Dose-response relationship between PSQI, dietary inflammation index and sarcopenia
In this study, we used a restricted cubic spline function (RCS) set at 4 nodes to analyze the dietary inflammation index (DII) and the risk of sarcopenia in a consistently increasing nonlinear relationship (P < 0.001) Dose-response characteristics: OR = 0.62 (95% CI:0.51 to 0.75) for DII=-2, OR = 1.0 (reference point) for DII = 0, DII = 3. OR = 2.94 (95% CI:2.17 to 3.98). Risk acceleration interval: 48% higher risk per unit increase after DII > 1 (see Fig. 1). The nonlinear association of PSQI total score and dietary inflammation index (DII) with the risk of sarcopenia was analyzed. After adjusting for confounders such as age, sex, and BMI, the results showed that PSQI total score (sleep quality) was significantly nonlinearly associated with the risk of sarcopenia (P < 0.001) Dose-response characteristics: the OR stabilized at the baseline level (OR ≈ 1) when the PSQI total score was < 5, and the risk increased exponentially after the PSQI total score was > 5 (for PSQI = 10, OR = 1.82; at PSQI = 15, OR = 3.15). Key inflection point: PSQI = 5 is the threshold point for significant rise in risk (Fig. 2).
Fig. 1.
Dose-response relationship between dietary inflammation index and sarcopenia based on restrictive cubic spline analysis
Fig. 2.
Dose-response relationship between PSQI and sarcopenia based on restricted cubic spline analysis
Column chart prediction model
The column-line plots constructed from the multifactorial logistic regression model showed that the weights of the predictor variables contributing to the total score were as follows: for every 1-year increase in age, the weight increased by 4 points; for every 1-unit increase in dietary inflammation index, the weight increased by 13 points; for every 1-point increase in the total PSQI score, the weight increased by 4 points; for every 1-kilogram-m² decrease in BMI, the weight increased by 12 points; for every 10-cm increase in waist circumference, the weighting increases by 20 points, see Fig. 3 for details.
Fig. 3.
Nomogram prediction model
Validation of the column map prediction model
ROC curves were used to test the discriminative ability of the column-line graph prediction model. The results showed that the area under the curve (AUC) of the training set was 0.784 (95%CI 0.74–0.828), and the AUC of the test set was 0.752 (95%CI 0.68 to 0.824), with AUC > 0.7 indicating that the model has more than moderate discriminative ability, and the difference between AUC of the training set and the test set was small (Δ = 0.032), which indicated that the model has good predictive differentiation ability (Fig. 4). Then the Bootstrap method was used to draw the calibration curve with 1000 repeated samples, the calibration error of the training set: MAE = 0.009 (n = 273), the calibration curve of the B = 1000 times self-help method showed that the predicted probability was highly consistent with the actual probability, and the calibration error of the test set: MAE = 0.019 (n = 273), the calibration curve was still close to the ideal diagonal but slightly lower than that of the training set (as expected), MAE < 0.05 indicates that the predicted probability error is within the clinically acceptable range, and the test set MAE (0.019) is significantly lower than the critical value of 0.05, and the model is well calibrated (Fig. 5). The Hosmer-Lemeshow test was then used to see how well the model fit, the χ² value for the training set was 9.098 with a p-value of 0.334, and the χ² value for the validation set was 12.319 with a p-value of 0.138, which showed that the model predicted probabilities were well-fitted to the actual probability of occurrence.
Fig. 4.
ROC curves of the nomogram prediction model
Fig. 5.
Hosmer–Lemeshow calibration curves of the nomogram prediction. Note: (A) is the training set and (B) is the validation set
DCA curve analysis of the model
Decision curve analysis (DCA) was used to figure out how much clinical net benefit this model got under different risk thresholds. As can be seen in Fig. 6, the prediction model constructed in this study has stable clinical decision value in the range of 15% to 65% risk thresholds (training set net benefit: 0.12 to 0.45; test set: 0.10 to 0.38), which is higher than the reference lines of “All” and “None”, and its standardized net benefit is significantly better than that of empirical decision strategies. The standardized net gain of “All” and “None” is significantly better than that of empirical decision-making strategies, which can provide an evidence-based basis for individualized interventions.
Fig. 6.
Decision curve analysis (DCA) of the nomogram prediction model
Discussion
In this study, the prevalence of sarcopenia and the factors affecting it were analyzed in 1370 elderly people aged 60 years and above. The results showed that the prevalence of sarcopenia in the elderly was 14.2%, and the risk increased significantly with age. In terms of social factors, the highest prevalence was found in the widowed group, and the risk was relatively high among pensioners, nursing home residents and those living with children; in terms of metabolic and physiological indicators, the sarcopenia group showed abnormalities in BMI, waist circumference and other indicators; in terms of lifestyle, the risk was lower in those who consumed alcohol, and the poor quality of sleep and the burden of chronic diseases were significantly associated with the prevalence of sarcopenia. The dose-response relationship between dietary inflammation index (DII) and Pittsburgh sleep quality index (PSQI) and the risk of sarcopenia was analyzed by restricted cubic spline function (RCS), and the results showed that there was a significant nonlinear positive correlation between DII and the risk of sarcopenia. The column-line plots constructed by multifactorial Logistic regression analysis clarified the contribution weights of each predictor variable, and the model was validated to have good predictive discriminatory ability and calibration, with stable clinical decision-making value within a certain risk threshold.
In terms of diet, Xie et al. have reported that dietary patterns are associated with muscle mass and found that a high dietary inflammation index was strongly associated with reduced muscle mass and strength [21].A study by Chen et al. showed that for each 1-unit increase in DII, there was an OR of 1.12 for the combination of low muscle mass and low muscle strength, with a 95% CI (1.01 to 1.25), suggesting that dietary inflammation has a negative impact of dietary inflammation [22]. The elevated inflammation index in patients with sarcopenia is consistent with the idea that chronic inflammation is involved in the pathogenesis of sarcopenia as suggested by related studies, and that inflammatory responses may lead to increased muscle proteolysis and metabolism, which can trigger sarcopenia [23]. These results support the conclusion of our study that improving dietary composition and lowering DII could contribute to the prevention of sarcopenia.A study by Pu et al. [21] in people over 65 years of age showed that higher DII was associated with an increased risk of muscle strength loss, with an OR for the occurrence of muscle strength loss in the group with the highest DII score of 1.65 (95% CI: 1.04 to 2.61, p < 0.05). Studies conducted found that DII was negatively correlated with grip strength (HGS) and extremity skeletal muscle mass (ASM), i.e., higher DII scores were associated with less grip strength and lower extremity skeletal muscle mass, while DII was positively correlated with timed get-up-and-go test (TUG) time, suggesting that pro-inflammatory diets may be associated with poorer physical performance [24]. This again reflects the importance of maintaining a healthy diet in the elderly population. The OR in this study (2.02) was higher than most of the literature, which may stem from geographic differences in dietary patterns (e.g., high-fat and high-salt diets are prevalent in Xinjiang) or higher baseline levels of inflammation in the population.
In contrast, the finding that poor sleep quality is associated with sarcopenia is consistent with the results of a study of community-dwelling older adults, which found that the prevalence of sarcopenia was approximately 27.3% higher in older adults with poor sleep quality than in those with good sleep quality [25]. One study found [26] that older women ≥ 60 years of age with a PSQI score > 5 (representing poor sleep quality) were at greater risk of developing sarcopenia. This is further supported by the results of the present study, especially the significant elevation of risk after a PSQI threshold of 5 points is in line with previous studies, which provides a good clinical rationale for improving the health of older adults.
Some studies have shown that the prevalence of sarcopenia in the elderly population ranges from about 10% to 27%, which is consistent with 14.2% in this study [27]. Regarding the age factor, numerous studies have shown that increasing age is an important risk factor for sarcopenia, and it was found that the average age of the sarcopenia population was greater than that of the non-sarcopenia population, and the difference was statistically significant (P < 0.001), indicating that the risk of sarcopenia increases with increasing age [28]; with regard to the social factors, the finding of a high prevalence in widowed populations is in line with some of the studies pointing out that the Marital status has a significant impact on the health of older adults, the companionship of the spouse plays an important role in the care and psychological support of older adults, and widowhood may lead to changes in lifestyle and poor psychological status, which may increase the risk of the disease [29]. A study showed that divorce or widowhood or other marital status was a risk factor for 10-year all-cause mortality (HR = 1.24, 95% CI = 1.05 to 1.45) [29]. In terms of metabolic and physiological indicators, the present study found that BMI was significantly lower in the sarcopenia group, which is consistent with the results of most studies that concluded that low BMI is a risk factor for sarcopenia.A cross-sectional and retrospective cohort analysis found that BMI was significantly negatively correlated with the risk of sarcopenia, i.e., the lower the BMI, the higher the risk of sarcopenia, and after adjusting for potential confounders, the sarcopenia risk was significantly higher in the group with a BMI < 18.5 than in the group with a BMI18.5. After adjusting for potential confounders, the risk of sarcopenia in the BMI < 18.5 group was significantly higher than that in the BMI 18.5 to 23 group [30]. This idea is also reflected in the present study where the weight of each 1 kg/m² decrease in BMI was increased by 12 points. The association of chronic disease burden with the prevalence of sarcopenia is also in line with several studies showing that multiple chronic diseases increase the risk of sarcopenia, which may lead to muscle loss by affecting nutrient absorption, mobility, and other pathways [31].
The innovation of this study is that a column-line graph prediction model was constructed through multifactor Logistic regression analysis, and the model was comprehensively validated to clarify the contribution weights of each predictor variable, which provides a more intuitive and effective tool for individualized prediction and intervention of sarcopenia. Meanwhile, social factors, metabolic and physiological indicators, lifestyle and other factors were comprehensively considered to reveal the influencing factors of sarcopenia more comprehensively. However, there are some shortcomings in this study. First, the study was a cross-sectional survey, which could not clarify the causal relationship between the factors and sarcopenia; second, the sample originated from a specific region, which may have geographical limitations and the extrapolation of the results may be affected to a certain extent; furthermore, in the survey of lifestyle factors, the indicators such as alcohol consumption and sleep quality relied mainly on self-reporting of the subjects, which may have a recall bias; and lastly, the study did not investigate the influence of some potential factors such as exercise intensity and specific components of nutritional intake were not analyzed in depth, which needs to be improved in further studies.
Supplementary Information
Acknowledgements
We thank all study participants and the staff of the community health service centers for their assistance.
Clinical trial number
Not applicable.
Authors’ information
Ping Tang (2001-) is a postgraduate student majoring in public health at the School of Public Health, Xinjiang Medical University, with research interests in epidemiology and statistics.
Authors’ contributions
Ping Tang: study design, data collection, data analysis, manuscript writing. SONGDA LI: data collection. XUEMEI YAO: technical equipment support, manuscript writing.
Funding
This work was supported by the Natural Science Foundation of the Xinjiang Uygur Autonomous Region (Grant No. 2024D01C131).
The 14th Five-Year Plan Distinctive Program of Public Health and Preventive Medicine in Higher Education Institutions of Xinjiang Uygur Autonomous Region.
Data availability
The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.
Declarations
Ethics approval and consent to participate
This study was approved by the Ethics Review Committee of Xinjiang Medical University (Approval No.: XJYKDXR20240314047). The procedures used in this study adhere to the tenets of the Declaration of Helsinki. Informed consent was obtained from all individual participants included in the study.
Consent for publication
All participants consented to the publication of the anonymized data and results.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Yuan S, Larsson SC. Epidemiology of sarcopenia: Prevalence, risk factors, and consequences. Metabolism. 2023;144:155533. [DOI] [PubMed] [Google Scholar]
- 2.Zhang H, Hu JW, Zhao XH, Wang PY, Ji Y, Shi YH. Prevalence and influencing factors of possible sarcopenia among middle-aged and elderly adults in China. Med Soc. 2023;36(08):44–48. [Google Scholar]
- 3.Zhou YX, Sun C, Zhao R, Dong C, Gu ZF, Gao JL. Research progress on the health impacts of sarcopenic obesity in the elderly. Chin J Osteoporos. 2025;31(06):913–917. [Google Scholar]
- 4.Pang T, Liu Y, Jiang X, Chen JY, Li JY, Zhang B, Li M. Correlation analysis between nutrition, dietary status and sarcopenia risk among elderly in Chengdu communities. In: The 14th Asian Congress of Nutrition. 2023:14–17. Chengdu, Sichuan, China.
- 5.Shivappa N, Steck SE, Hurley TG, Hussey JR, Hébert JR. Designing and developing a literature-derived, population-based dietary inflammatory index. Public Health Nutr. 2014;17(8):1689–1696. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Wang Y, Xie MX, Lu J, Wang Y. Observation on the application of dietary nutrition intervention based on dietary inflammatory index in postoperative chemotherapy patients with colorectal cancer. Mod Med J. 2025;53(06):927–935. [Google Scholar]
- 7.Yang M. Patients with coronary heart disease should pay attention to the dietary inflammatory index. Fam Med Happy Health. 2024;(10):16–17. [Google Scholar]
- 8.Jeanes YM, Reeves S. Metabolic consequences of obesity and insulin resistance in polycystic ovary syndrome: diagnostic and methodological challenges. Nutr Res Rev. 2017;30(1):97–105. [DOI] [PubMed] [Google Scholar]
- 9.Khan S, Wirth MD, Ortaglia A, Alvarado CR, Shivappa N, Hurley TG, Hebert JR. Design, development and construct validation of the children’s dietary inflammatory index. Nutrients. 2018;10(8):993. [DOI] [PMC free article] [PubMed]
- 10.Pang T, Jiang X, Zhang QW, He ZJ, Liu Y, Li M. Study on the possible correlation between nutritional dietary status and sarcopenia in community-dwelling elderly. Chin J Prev Control Chronic Dis. 2024;32(11):848– 852, 856.
- 11.Diao H, Yan F, He Q, Li M, Zheng Q, Zhu Q, Fang F, Cui W. Association between dietary inflammatory index and sarcopenia: a meta-analysis. Nutrients. 2023;15(1):219. [DOI] [PMC free article] [PubMed]
- 12.Bian D, Liu X, Wang C, Jiang Y, Gu Y, Zhong J, Shi Y. Association between dietary inflammatory index and sarcopenia in crohn’s disease patients. Nutrients. 2022;14(4):901. [DOI] [PMC free article] [PubMed]
- 13.Zheng XY, Chen Y, Gao Y, Zhou M, Guan WZ. Association between dietary inflammatory index (DII) and sarcopenia in ischemic heart disease: an 1999–2004 NHANES study of 1088 US individuals. Minerva Cardiol Angiol. 2025;73(1):65–72. [DOI] [PubMed] [Google Scholar]
- 14.Ma Y, Bian AL, Wang XF, Wang JX, Zhou XZ. Differences in the prevalence of sarcopenia by gender and age among the elderly in Tianjin. Chin Gen Pract. 2017;20(26):3267–3271. [Google Scholar]
- 15.Mu KL. Reliability and validity analysis of a semi-quantitative food frequency questionnaire for Xinjiang characteristic diets. Master’s Thesis. Xinjiang Medical University; 2023.
- 16.Kenđel Jovanović G, Pavičić Žeželj S, Klobučar Majanović S, Mrakovcic-Sutic I, Šutić I. Metabolic syndrome and its association with the dietary inflammatory index (DII)(®) in a Croatian working population. J Hum Nutr Diet. 2020;33(1):128–137. [DOI] [PubMed] [Google Scholar]
- 17.Editorial Department: Publication of the China Food Composition Table Standard Edition (6th Edition), Books 1 & 2. Acta Nutr Sin. 2019;41(05):426. [Google Scholar]
- 18.Pawlow X, Ott R, Winkler C, Ziegler AG, Hummel S. A new mathematical approach to improve the original dietary inflammatory index (DII) calculation. PLoS ONE. 2021;16(11):e0259629. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Qiu JQ, Zhou M, Qiu PY, Zhang N, Wan Y. Reliability and validity analysis of the Pittsburgh sleep quality index in a rural elderly population. Mod Prev Med. 2016;43(10):1835–1838. [Google Scholar]
- 20.Chen LK, Liu LK, Woo J, Assantachai P, Auyeung TW, Bahyah KS, Chou MY, Chen LY, Hsu PS, Krairit O, et al. Sarcopenia in asia: consensus report of the Asian working group for sarcopenia. J Am Med Dir Assoc. 2014;15(2):95–101. [DOI] [PubMed] [Google Scholar]
- 21.Xie H, Wang H, Wu Z, Li W, Liu Y, Wang N. The association of dietary inflammatory potential with skeletal muscle strength, mass, and sarcopenia: a meta-analysis. Front Nutr. 2023;10:1100918. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chen L, Ming J, Chen T, Hébert JR, Sun P, Zhang L, Wang H, Wu Q, Zhang C, Shivappa N, et al. Association between dietary inflammatory index score and muscle mass and strength in older adults: a study from National health and nutrition examination survey (NHANES) 1999–2002. Eur J Nutr. 2022;61(8):4077–4089. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Pu R, Man Q, Song S, Jia S, Liu Z, Zhang X, Zhang J, Song P. The dietary inflammatory index and sarcopenia in older adults in four Chinese provinces: a cross-sectional study. Nutrients. 2025;17(3):478. [DOI] [PMC free article] [PubMed]
- 24.Linton C, Wright HH, Wadsworth DP, Schaumberg MA. Dietary inflammatory index and associations with sarcopenia symptomology in community-dwelling older adults. Nutrients. 2022;14(24):5319. [DOI] [PMC free article] [PubMed]
- 25.Cacciatore S, Calvani R, Mancini J, Ciciarello F, Galluzzo V, Tosato M, Marzetti E, Landi F. Poor sleep quality is associated with probable sarcopenia in community-dwelling older adults: results from the longevity check-up (lookup) 8. Exp Gerontol. 2025;200:112666. [DOI] [PubMed] [Google Scholar]
- 26.Yao XM, Li MY, Wu D, Xiao H. Relationship between sleep quality and sarcopenia in the elderly. Chin Gen Pract. 2020;23(22):2804–2808, 2814.
- 27.Petermann-Rocha F, Balntzi V, Gray SR, Lara J, Ho FK, Pell JP, Celis-Morales C. Global prevalence of sarcopenia and severe sarcopenia: a systematic review and meta-analysis. J Cachexia Sarcopenia Muscle. 2022;13(1):86–99. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Wu XJ, Wang MP, Zeng JB. Risk factor analysis and prediction model construction for sarcopenia in Chinese elderly. Shandong Med J. 2025;65(06):33–36, 41.
- 29.Kong J, Li SQ, Huang PD, Liu MH, Ren DJ. An empirical analysis of the association between sarcopenia and all-cause mortality among elderly residents in China. Shanghai J Prev Med. [Online ahead of print].
- 30.Lu B, Li J, Liang X, Wen M, Luo D, Jia H, Zhang J, Li G. Association between atherogenic index of plasma, body mass index, and sarcopenia: a cross-sectional and longitudinal analysis study based on older adults in China. Aging Clin Exp Res. 2025;37(1):122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Zhao WZ, Zhang Y, Dai C, Xie LF, Chen R, Lou X, Shen GD. Association between Multimorbidity and sarcopenia among older adults in rural Anhui. J Shandong First Med Univ Shandong Acad Med Sci. 2024;45(12):741–746.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The datasets generated and/or analysed during the current study are available from the corresponding author on reasonable request.