Abstract
Background:
The relationship of insulin resistance with bone mineral density (BMD) remains unclear, offering an opportunity for novel indices to shed light on the matter. The aim of this review was to evaluate the association between surrogate indices of insulin resistance and BMD.
Methods:
A systematic review was conducted to evaluate observational studies that examined the relationship between insulin resistance surrogate indices and BMD in adults. Databases including PubMed, Web of Science, Scopus, and Embase were searched. Quality assessment was performed using Joanna Briggs Institute (JBI) critical appraisal tools.
Results:
This systematic review included 27 cohorts and cross-sectional studies with 71,525 participants to assess the potential link between insulin resistance surrogate indices like HOMA-IR, HOMA-β, TyG, TyG-BMI, TyG-WtHR, and TyG-WC, along with METS-IR, and VAI, and BMD at various sites. There seems to be no link between BMD and the HOMA index, despite being extensively studied in various studies (adjusted β ranging from -0.49 to 0.103). Most literature suggests that a higher TyG index is associated with decreased BMD levels (adjusted β ranging from -0.085 to 0.0124). Despite limited evidence, other insulin resistance indices such as VAI (adjusted β ranging from 0.007 to 0.016), TyG-BMI (adjusted β ranging from 0.002 to 0.415), METS-IR (adjusted β ranging from 0.005 to 0.060), TyG-WtHR (β = 0.012) and TyG-WC (β = 0.0001) have shown a positive association with BMD in a few studies.
Conclusion:
This systematic review emphasizes the intricate connection between insulin resistance and BMD. The lack of ability to perform a meta-analysis and the dependence on cross-sectional studies hinder the robustness of the findings, hence necessitating well-designed longitudinal studies.
Systematic review registration:
https://www.crd.york.ac.uk/prospero/, identifier CRD42024512770.
Introduction
Osteoporotic fractures are associated with low BMD, making them a risk factor. The age-standard rate of osteoporosis incidence in 2019 was 49.2 million (). The amount of global deaths and DALYs linked to low bone mineral density rose by 111.16% and 93.82% from 207,367 and 8,588,936 in 1990 to 437,884 and 16,647,466 in 2019 (). Several variables can influence bone health and contribute to the development of osteoporosis (). While certain characteristics like age, sex, or ethnicity have an inescapable impact on bone health, various modifiable factors, such as weight, alcohol consumption, or a sedentary lifestyle, can be improved by appropriate diet and physical activity to enhance bone condition (). Multiple investigations have assessed the substantial impact of insulin resistance on various health-related conditions, including bone health.
The literature was clued into the inverse association of insulin resistance and bone health when the hypothesis of the negative impact of insulin resistance on bone remodeling was first presented (, ). Visceral fat mass (), an altered lipid profile (), and metabolic syndrome () are additional factors that might negatively impact bone health, leading to osteoporosis and an increased risk of fractures in the population. These parameters are sometimes referred to as surrogate indices for insulin resistance (, ). Multiple investigations have assessed the substantial impact of insulin resistance surrogate indices on various health-related conditions, such as vascular damage (), hypertension (), and metabolic syndrome (). Furthermore, the correlation between these factors and bone health is receiving significant attention and has been thoroughly examined in several research studies (, ).
Given that the findings of studies on the association of insulin resistance and its surrogates with BMD are not consistent (–) and lack systematic reviews on this association to summarize the available evidence, the objective of this study was to assess the impact of surrogate indices of insulin resistance, including Homeostatic Model Assessment for Insulin Resistance (HOMA-IR), Homeostatic Model Assessment for Beta-Cell Function (HOMA-β), Triglyceride-Glucose Index (TyG), Triglyceride-Glucose Body Mass Index (TyG-BMI), Triglyceride-Glucose Waist-to-Height Ratio (TyG-WtHR), Triglyceride-Glucose Waist Circumference (TyG-WC), Metabolic Score for Insulin Resistance (METS-IR), and Visceral Adiposity Index (VAI), on BMD to clarify the possible role of insulin resistance in the determination of bone health.
Materials and methods
The PRISMA statement guidelines were followed during the conduct of this meta-analysis (). This study adhered to a predetermined process outlined in the prospective register of systematic reviews (PROSPERO) (CRD42024512770).
Search strategy and screening
We conducted a search on electronic databases, including PubMed, Embase, Web of Science, and Scopus, until July 2024. The search utilized the following terms or relevant Medical Subject Headings (MeSH): (“visceral adiposity index” [Title/Abstract] OR “VAI” [Title/Abstract] OR “lipid accumulation product” [Title/Abstract] OR “LAP” [Title/Abstract] OR “triglyceride glucose index” [Title/Abstract] OR “TyG” [Title/Abstract] OR “triglyceride-glucose index” [Title/Abstract] OR “TyG-body mass index” [Title/Abstract] OR “TyG-BMI” [Title/Abstract] OR “TyG-waist circumference” [Title/Abstract] OR “TyG-WC” [Title/Abstract] OR “Homeostatic Model Assessment for Insulin Resistance” [Title/Abstract] OR “HOMA-IR” [Title/Abstract] OR “Metabolic Syndrome Insulin Resistance” [Title/Abstract] OR “MetS-IR” [Title/Abstract] OR “Lipoprotein Insulin Resistance Index” [Title/Abstract] OR “LP-IR” [Title/Abstract] OR “TyG-NC” [Title/Abstract] OR “TyG-NHtR” [Title/Abstract] OR “triglycerides to HDL cholesterol ratio” [Title/Abstract] OR “TG/HDL-C” [Title/Abstract] OR “Adipose insulin resistance index” [Title/Abstract] OR “Adipo-IR” [Title/Abstract] OR “lipid indices” [Title/Abstract] OR “Insulin Resistance index” [Title/Abstract] OR “Insulin Resistance indices” [Title/Abstract]) AND (“bone density”[MeSH] OR “fractures, bone”[MeSH] OR “osteoporosis”[MeSH] OR “osteoporosis”[Title/Abstract] OR “osteoporotic”[Title/Abstract] OR “osteoporoses”[Title/Abstract] OR “bone loss”[Title/Abstract] OR “fracture”[Title/Abstract] OR “bone demineralisation”[Title/Abstract] OR “bone demineralization”[Title/Abstract] OR “metabolic bone disease*”[Title/Abstract] OR “osteopenia”[Title/Abstract] OR “osteopenic”[Title/Abstract] OR “osteopaenia”[Title/Abstract] OR “osteopaenic”[Title/Abstract] OR “bone density”[Title/Abstract] OR “bone deterioration”[Title/Abstract] OR “bone mass density”[Title/Abstract] OR “bone mineral density”[Title/Abstract] OR “BMD”[Title/Abstract]). Further articles were screened for eligibility by referencing the included studies. Rayyan, a free online web tool for systematic reviewing, was used to screen the studies. It is accessible at https://www.rayyan.ai. Two reviewers (A.G.R. and A.A.) independently assessed each study and thoroughly examined the entire text to remove any duplicate materials. Studies meeting the inclusion-exclusion criteria were chosen. The third author (A.H.H.) conducted consensus meetings to address any potential disagreements among reviewers.
Inclusion and exclusion criteria
Population (P): Adult participants; Exposure (E): Studies assessing insulin resistance surrogate indices (e.g., HOMA-IR, TyG, VAI).; Comparison (C): Individuals with different levels of insulin resistance surrogate indices; Outcomes (O): BMD of different areas when beta coefficient was reported by multivariate analysis; Type of Design (T): Observational studies. Here are the formulas illustrating the definitions for the insulin surrogate indices.
For men:
For women:
Exclusion criteria were studies that involved patients with diabetes, studies without adjustment for covariates such as multivariable logistic regression analysis, non-English studies, studies that did not report BMD values (T score or Z score), FRAX or osteoporosis (low T score +/- fracture), as well as case reports, reviews, editorials, commentaries, and conference abstracts that lacked original research data or detailed methodologies.
Data extraction and quality assessment
After conducting a full-text screening, two researchers (S.M.A., Y.T.) separately entered the data into an existing Excel spreadsheet. This document included two categories of data (): demographic details like journal, authors, publication year, study location, design, sample size, gender, patient age, IR index, and outcomes; and (2) specific results such as covariates, beta coefficient, insulin resistance indices (HOMA-IR, HOMA-β, VAI, TyG, TyG-BMI, TyG-WTHR, TyG-WC, and METS-IR) levels, and BMD of various anatomical sites. The third reviewer (A.H.H.) assessed the conflicts.
Two authors (N.Z., A.A.) individually assessed the studies’ quality using the critical appraisal checklists developed by the Joanna Briggs Institute (JBI) for cohort studies (). The JBI critical evaluation checklist consists of eleven components for cohort studies and eight for cross-sectional studies. The checklist assesses specific research topics to detect possible bias risks and offers direct binary answers. If the answer was yes, the question received a score of 1. Responses that were negative, ambiguous, or irrelevant were assigned a score of 0 (). Scheduled meetings were arranged in order to come to a common agreement and settle any disagreements.
Results
Study selection
The initial systematic search of databases, including PubMed, Web of Science, Scopus, and Embase, identified 7,518 studies. After removing duplicates (n = 4,021), 3,497 records proceeded to title and abstract screening, of which 3,380 studies were excluded as irrelevant. A full-text assessment was conducted for the remaining 117 studies. Twenty-seven of these studies were deemed eligible for systematic review (Figure 1).
Figure 1
Baseline characteristics and quality assessment
This systematic review included studies conducted across various geographic regions, comprising the USA (, , –), China (–), Korea (, –), Brazil (, ), Japan (), Singapore (), Serbia () and Spain (). The studies utilized cohort (, , , ) and cross-sectional (, , –, –, , ) methodologies. Data for 71,525 participants (52.1% female) were analyzed. The mean age of our population ranged from 30.3 () to 73.6 (). Different indices, such as HOMA-IR, HOMA-β, VAI, TyG, TyG-BMI, TyG-WTHR, TyG-WC, and METS-IR, were used to indicate insulin resistance. Various outcome measures and additional information are summarized in Table 1.
Table 1
| Author | Year | Country | Study Design | Analytic sample | Age (mean±SD) | Female % | IR index | Outcomes |
|---|---|---|---|---|---|---|---|---|
| Kim | 2013 | South Korea | Cross-sectional | 14485 | 45.4±0.3 | 54.03 | HOMA-IR | BMD (Lumbar spine, Total hip, and Femoral neck) |
| Shin | 2014 | South Korea | Cross-sectional | 3113 | 49.7±0.6 | 0 | HOMA-IR | BMD (Whole body, Lumbar spine, and Femoral neck) |
| Srikanthan | 2014 | USA | Cross-sectional | 717 | 56.8±11.3 | 51.6 | HOMA-IR | BMD (Lumbar spine and Femoral neck) |
| Shanbhogue | 2016 | USA | Cross-sectional | 146 | 60.3±2.7 | 100 | HOMA-IR | BMD (Lumbar spine and Total hip) |
| Choo | 2017 | South Korea | Cross-sectional | 2750 | 30.3±0.2 | 56.07 | HOMA-IR | BMD (Total hip, Femoral neck, Lumbar spine, Femoral trochanter, and Femoral intertrochanter) |
| Iki | 2012 | Japan | Prospective Cohort | 1683 | 72.9±5.2 | 0 | HOMA-IR, HOMA-β | BMD (Lumbar spine) |
| Seoung | 2018 | South Korea | Cross-sectional | 137 | 55.6±5.8 | 100 | HOMA-IR | BMD (Lumbar spine and Femoral neck) |
| Kalimeri | 2018 | China-Singapore | Cohort | 96 | 60.7±4.2 | 100 | HOMA-IR | BMD (Lumbar spine, Total hip and Femoral neck) |
| Napoli | 2019 | USA | Prospective Cohort | 2398 | 73.6±2.9 | 53 | HOMA-IR | BMD (Total hip) |
| Yang | 2019 | China | Cross-sectional | 892 | 55±1.1 | 100 | HOMA-IR | BMD (Lumbar spine and Femoral neck) |
| de Araújo | 2020 | Brazil | Cross-sectional | 56 | 47±14 | 64.28 | HOMA-IR | BMD (Lumbar spine, Total hip, and Femoral neck, Radius) |
| Wang | 2020 | China | Cross-sectional | 2122 | 45.1±15 | 59.94 | HOMA-IR | BMD (Forearm) |
| Campillo-Sánchez | 2020 | Spain | Cross-sectional | 381 | 62±8.6 | 100 | HOMA-IR | BMD (Total hip and Femoral neck) |
| Yoon | 2021 | Korea | Cross-sectional | 4810 | 62.7±8.7 | 46.94 | HOMA-IR, TyG | BMD (Whole body, Lumbar spine, and Femoral neck, Total hip) |
| Giudici | 2021 | Brazil | Cross-sectional | 298 | 57.5±8.8 | 55.7 | HOMA-IR, HOMA-β | BMD (Whole body, Lumbar spine, and Femoral neck) |
| Ye | 2023 | China | Cross-sectional | 437 | 53.5±1.2 | 100 | HOMA-IR | BMD (Femoral neck) |
| Sun | 2023 | USA | Cross-sectional | 3341 | 69.8±6.9 | 47.4 | VAI | BMD (Total femur, Femur neck, Trochanter, Intertrochanter) |
| Pu | 2023 | USA | Cross-sectional | 1114 | 58.6±12.2 | 50.9 | METS-IR | BMD (Lumbar spine and Total femur) |
| Chen | 2023 | USA | Cross-sectional | 6257 | 50.2±15.5 | 50.1 | VAI | BMD (Total femur) |
| Zhan | 2023 | USA | Cross-sectional | 3646 | 37.4±11.2 | 46.4 | TyG | BMD (Lumbar, Whole body) |
| Xuan | 2024 | USA | Cross-sectional | 1182 | 60.3±8.2 | 0 | TyG-BMI | BMD (Femoral neck) |
| Tian N | 2024 | USA | Cross-sectional | 5456 | 30.3±13.5 | 44.35 | TyG, TyG-BMI, TyG-WHTR, TyG-WC | BMD (Whole body) |
| Chen | 2024 | USA | Cross-sectional | 1844 | 60.7±0.3 | 37 | TyG | BMD (Femoral neck) |
| Tian C | 2024 | USA | Cross-sectional | 6501 | 51.2±16.9 | 49.3 | TyG-BMI | BMD (Total femur, Femur neck, Trochanter, Intertrochanter) |
| Wen | 2022 | China | Prospective cohort | 832 | 59.3±7.8 | 43.0 | TyG-BMI | BMD (Total hip, Lumbar spine, Femoral neck) |
| Shao | 2024 | China | Cross-sectional | 6769 | 47.4±17.5 | 49.0 | METS-IR | BMD (Total femur, Femoral neck, Total spine) |
| Sretenović | 2021 | Serbia | Cross-sectional | 62 | 71.2±4.8 | 100 | HOMA-IR | BMD, T score and Z score (Hip, Lumbar spine) |
Baseline characteristics of the included studies.
25(OH)D, 25-Hydroxyvitamin D; 25(OH)D2, 25-Hydroxyvitamin D2; 25(OH)D3, 25-Hydroxyvitamin D3; ALP, Alkaline Phosphatase; ALT, Alanine Transaminase; AST, Aspartate Transaminase; BFM, Body Fat Mass; BMD, Bone Mineral Density; BMI, Body Mass Index; BUN, Blood Urea Nitrogen; CPK, Creatine Phosphokinase; CRP, C-Reactive Protein; CVD, Cardiovascular Disease; DBP, Diastolic Blood Pressure; eGFR, Estimated Glomerular Filtration Rate; FMI, Fat Mass Index; FPG, Fasting Plasma Glucose; FSH, Follicle-Stimulating Hormone; HDL-C, High-Density Lipoprotein Cholesterol; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance; HOMA-β, Homeostatic Model Assessment for Beta-cell Function; Hs-CRP, High-sensitivity C-Reactive Protein; IFG, Impaired Fasting Glucose; LMI, Lean Mass Index; LDL-C, Low-Density Lipoprotein Cholesterol; MetS-IR, Metabolic Syndrome Insulin Resistance; PIR, Poverty Income Ratio; SBP, Systolic Blood Pressure; SCr, Serum Creatinine; SMM, Skeletal Muscle Mass; SUA, Serum Uric Acid; TC, Total Cholesterol; TG, Triglycerides; TBS, Trabecular Bone Score; tOC, Total Osteocalcin; TyG, Triglyceride Glucose Index; TyG-BMI, TyG-Body Mass Index; TyG-WC, TyG-Waist Circumference; TyG-WHtR, TyG-Waist-to-Height Ratio; UA, Uric Acid; VAI, Visceral Adiposity Index; VFA, Visceral Fat Area; WC, Waist Circumference; WHR, Waist-to-Hip Ratio; ucOC, Undercarboxylated Osteocalcin; anti-TPO, Anti-Thyroid Peroxidase Antibody; TSH, Thyroid-Stimulating Hormone; Ft4, Free Thyroxine; anti-Tg, Anti-Thyroglobulin Antibody; IGF, Insulin-like Growth Factor; PTH, Parathyroid Hormone.
The quality assessment of cross-sectional studies in this review indicated a uniformly high standard, with all studies meeting each of the eight JBI quality criteria. The cohort studies also demonstrated high quality, with most criteria consistently met across studies. However, there were some limitations in addressing incomplete follow-up in two of the cohort studies (, ), which did not employ strategies to address this issue. Despite these minor limitations, the overall methodological quality of the included studies was robust (Supplementary Tables S1, S2).
HOMA-IR and bone mineral density
A number of 17 studies reported the association between HOMA-IR and BMD (, , , , –, –). Several results of studies reported that elevated levels of HOMA-IR were inversely associated with BMD (, , , , , , ), with the adjusted regression coefficient ranging from -1.11 () to -0.021 (). As for specific sites of BMD, higher levels of HOMA-IR had a negative effect on whole-body BMD, with the adjusted regression coefficient ranging from -0.041 () to -0.025 (), femoral neck BMD [from -1.09 () to -0.021 ()], and lumbar BMD [from β = -0.49 () to β = -0.084 ()]. Numerous results showed that the level of HOMA-IR does not statistically correlate with BMD (, , , , –). Kim et al., with a sample size of 14,485, revealed that in premenopausal women, no association was found between the HOMA-IR index and lumbar spine BMD (β = -0.16, P = 0.352). Lumbar spine, total hip, and femoral neck BMD were not correlated with the HOMA-IR level in postmenopausal women (β = 0.2, P = 0.482; β = 0.03, P = 0.940; and β = -0.45, P = 0.409) (). Yoon et al., with a sample size of 4,810, reported that HOMA-IR in men’s lumbar spine (β = -0.014, P = 0.499) and total hip (β = -0.021, P = 0.344), as well as in women’s lumbar spine (β = 0.006, P = 0.813), femoral neck (β = -0.043, P = 0.115), total hip (β = 0.013, P = 0.620), whole body (β = -0.026, P = 0.260) BMD, were not associated with HOMA-IR levels (). In contrast, Yang et al. identified a direct correlation between HOMA-IR and BMD of the lumbar spine and femoral neck (β = 0.103, P = 0.002; β = 0.091, P = 0.009) (). Likewise, Ye et al. found that greater insulin resistance was associated with increased femoral neck BMD in nondiabetic postmenopausal women (β [95%CI] = 0.025 [0.003, 0.047], P = 0.026) ().
TyG and bone mineral density
Four studies reported on the relationship between TyG index and BMD (, , , ). Yoon et al. found that an inverse association existed between TyG index and femoral neck, total hip, and whole-body BMD in non-diabetic men (β = -0.085, P < 0.001; β = -0.046, P = 0.037; β = -0.098, P < 0.001) and femoral neck and whole-body BMD in women (β = -0.071, P = 0.008; β = -0.065, P = 0.005), but no association was observed in the total hip area BMD of women (β = -0.003, P = 0.911). Lumbar spine BMD was found to have no relationship with TyG index in men (β = -0.028, P = 0.168) or women (β = 0.016, P = 0.500) (). According to Zhan et al., an inverse correlation existed between TyG index and BMD of the lumbar spine and whole body (β [95%CI] = -0.008 [-0.017, 0], β [95%CI] = -0.007 [-0.012, -0.001]) (). Conversely, Tian N et al. found a positive correlation between whole-body BMD and TyG (β [95%CI] = 0.0124 [0.001, 0.024]) (). Furthermore, Chen et al. found no association between femoral neck BMD and TyG index (). Therefore, most of the literature suggests that a higher TyG index, reflecting higher insulin resistance, is generally associated with lower BMD.
VAI and bone mineral density
Two studies reported on VAI and BMD (, ). There was a positive association between VAI and BMD of total femur, femoral neck, trochanter, and intertrochanter, Sun et al. reported (β [95%CI] = 0.006 [0.004, 0.009], P < 0.001; β [95%CI] = 0.004 [0.002, 0.006], P = 0.001; β [95%CI] = 0.005 [0.003, 0.007], P < 0.001; β [95%CI] = 0.007 [0.004, 0.010], P < 0.001) (). Chen et al. displayed a substantially positive association between femoral BMD and VAI (β [95%CI] = 0.016 [0.014, 0.019], P < 0.001) (). Hence, a greater VAI index, indicating increased insulin resistance, was generally linked to higher BMD.
Other insulin surrogate indices and bone mineral density
HOMA-β (, ), TyG-BMI (–, ), METS-IR (), TyG-WTHR (), and TyG-WC () were reported in very few studies. Giudici et al. reported on HOMA-β and BMD of the total body, lumbar spine, and femur (β = -0.006, P = 0.631; β = 0.071, P = 0.206; β = -0.021, P = 0.089), and found no statistically significant association (). Similarly, Iki et al. found no correlation between HOMA-β and lumbar spine BMD (β = 0.043, P = 0.0948) (). Therefore, it can be deduced that no association exists between HOMA-β and BMD.
As for TyG-BMI, Xuan et al. assessed insulin resistance by TyG-BMI, which demonstrated a positive association between TyG-BMI and femoral neck BMD as well (β [95%CI] = 0.058 [0.045, 0.072], P < 0.001) (), similar to Tian N et al. assessing whole-body BMD and TyG-BMI (β [95%CI] = 0.0004 [0.0003, 0.0004], P <0.0001) () and Tian C et al. assessing total femur, femur neck, trochanter, intertrochanter BMD and TyG-BMI (β [95%CI] = 0.002 [0.002, 0.002], P<0.00001; β [95%CI] = 0.001 [0.001, 0.002], P<0.00001; β [95%CI] = 0.001 [0.001, 0.001], P<0.00001; β [95%CI] = 0.002 [0.002, 0.002], P<0.00001) (). Likewise, Wen et al. indicated positive correlation between TyG-BMI and femoral neck, lumbar spine and total hip BMD in both male (respectively β = 0.224, 0.185, 0.271; p < 0.001) and female (respectively β = 0.279, 0.192, 0.415; p < 0.001) ().
Regarding METS-IR, Pu et al. illustrated a direct correlation between elevated levels of METS-IR, indicating higher insulin resistance, and increased total femoral and lumbar spine BMD (β [95%CI] = 0.005 [0.004, 0.006]; β = 0.005 [0.004, 0.006]) (). Similarly, Shao et al. found positive association between METS-IR and BMD of total femur, femoral neck and total spine (β[95%CI] = 0.060 [0.057, 0.064]; β[95%CI] = 0.049 [0.045, 0.052]; β[95%CI] = 0.040 [0.036, 0.044]; P<0.001) ().
Lastly, a positive association was seen between whole-body BMD and TyG-WTHR (β[95%CI] = 0.012 [0.008, 0.016], P<0.0001) and TyG-WC (β[95%CI] = 0.0001 [0.0001, 0.0001], P<0.0001) (). Summarized information regarding the qualitative synthesis of the included data is visibale in Table 2.
Table 2
| Study | IR index | Adjustment Covariates | Adjusted β | Conclusion |
|---|---|---|---|---|
| Giudici 2021 | HOMA-IR, HOMA-β | Age, Sex, LMI, FMI, and 25(OH)D | Whole body BMD and HOMA-IR = -0.011 Lumbar spine BMD and HOMA-IR= -0.018 Femoral BMD and HOMA-IR= -0.002 Whole body BMD and HOMA-β = -0.006 Lumbar spine BMD and HOMA-β = 0.071 Femoral BMD and HOMA-β = -0.021 | No association between HOMA-IR/HOMA-β and BMD |
| Campillo-Sánchez 2020 | HOMA-IR | Past And Present Medication, Lifestyle Factors, Age, BMI, the Presence of Osteoporosis and Osteoporosis Risk Factors. | Femoral neck BMD and HOMA-IR= 0.224 Total hip BMD and HOMA-IR= 0.225 | No association between HOMA-IR and BMD |
| Kim 2013 | HOMA-IR | Age, BMI, Smoking, Alcohol, and Regular Exercise | Men Lumbar spine BMD and HOMA-IR= -0.49 Femoral neck BMD and HOMA-IR= -1.09 Total hip BMD and HOMA-IR= -1.11 Premenopausal women Lumbar spine BMD and HOMA-IR= -0.16 Femoral neck BMD and HOMA-IR= -0.33 Total hip BMD and HOMA-IR= -0.37 Postmenopausal Lumbar spine BMD and HOMA-IR= 0.20 Femoral neck BMD and HOMA-IR= -0.45 Total hip BMD and HOMA-IR= 0.03 | In men, negative association between HOMA-IR and BMD. In premenopausal women negative association between HOMA-IR and femoral neck and total hip, no association with lumbar spine BMD. In postmenopausal women, no association between HOMA-IR and BMD. |
| Shin 2014 | HOMA-IR | Age, Weight, Height, Smoking, Alcohol, Income, Physical Activity, Calcium Intake, 25(OH)D, Diabetes, Percent Fat Mass, TC, HDL-C, and TG | Whole-Body BMD and HOMA-IR= -0.03 Femoral Neck BMD and HOMA-IR= -0.02 Lumbar Spine BMD and HOMA-IR= -0.01 | Negative association between HOMA-IR and whole body and femoral neck BMD, but no association with lumbar spine BMD |
| Srikanthan 2014 | HOMA-IR | Age, Sex, Race/Ethnicity, Menopause Transition Stage in Women, And Study Site | Lumbar spine BMD and HOMA-IR=-0.09 (-0.07, -0.02) Femoral neck BMD and HOMA-IR= -0.05 (-0.13, 0.027) | Negative association between lumbar spine BMD and HOMA-IR, no association between femoral neck BMD and HOMA-IR |
| Shanbhogue 2016 | HOMA-IR | Weight, Time Since Menopause, Tobacco, Alcohol, Physical Activity, Prior Use Of Osteoporosis Medications, Systemic HRT, or Glucocorticoids. | Lumbar spine BMD and HOMA-IR= 0.01 (-0.17, 0.19) Total hip BMD and HOMA-IR = 0.166 (-0.01, 0.34) | No association between BMD and HOMA-IR |
| Choo (men) 2017 | HOMA-IR | Gender, Age, Height, Weight, BFM, SBP, DBP, TC, TG, HDL-C, LDL-C, 25(OH)D, Smoking, Alcohol, Physical Activity, Education Level, Household Income, Use of Oral Contraceptives, and Age at Menarche in Females | Total hip BMD and HOMA-IR= -0.07 Femoral neck BMD and HOMA-IR= -0.1 Lumbar spine BMD and HOMA-IR= -0.08 Femoral trochanter BMD and HOMA-IR= -0.066 Femoral intertrochanter BMD and HOMA-IR= -0.054 | Negative association between HOMA-IR and BMD |
| Choo 2017 (women) | HOMA-IR | Gender, Age, Height, Weight, BFM, SBP, DBP, TC, TG, HDL-C, LDL-C, 25(OH)D, Smoking, Alcohol, Physical Activity, Education Level, Household Income, Use of Oral Contraceptives, and Age at Menarche in Females | Total hip BMD and HOMA-IR= -0.044 Femoral neck BMD and HOMA-IR= -0.051 Lumbar spine BMD and HOMA-IR= -0.035 Femoral trochanter BMD and HOMA-IR= -0.04 Femoral intertrochanter BMD and HOMA-IR= -0.03 | Negative association between femoral neck BMD and HOMA-IR, no association between total hip, femoral trochanter and intertrochanter BMD and HOMA-IR |
| Iki 2012 | HOMA-IR, HOMA-β | Age, BMI and TBS | Lumbar spine BMD and HOMA-IR= 0.013 Lumbar spine BMD and HOMA-β= 0.043 | No association between HOMA-IR, HOMA-β and BMD |
| Seoung 2018 | HOMA-IR | Age, Years Since Menopause, BMI, Smoking, Alcohol, WHR, VFA, BFM, SMM, TC, TG, HDL C, LDL-C, FPG, Hs-CRP, Adiponectin, Leptin, tOC, ucOC, Hypertension, And Lipid-Lowering Therapy | Lumbar spine BMD and HOMA-IR= 0.012 (-0.021, 0.024) Femoral neck BMD and HOMA-IR= 0.059 (-0.021, 0.038) | No association between BMD and HOMA-IR |
| Kalimeri 2018 | HOMA-IR | Lean Body Mass, Age | Femoral neck BMD and HOMA-IR= -0.064 Total hip BMD and HOMA-IR= -0.096 Lumbar spine BMD and HOMA-IR= -0.199 | Negative association between lumbar spine BMD and HOMA-IR, no association for femoral neck and total hip BMD and HOMA-IR |
| Napoli 2019 | HOMA-IR | Age, Sex, Race, eGFR, Clinic Site, BMI | Total hip BMD and HOMA-IR= Reference Total hip BMD and HOMA-IR= -0.006 Total hip BMD and HOMA-IR= 0.004 Total hip BMD and HOMA-IR= 0.007 | No association between HOMA-IR and BMD |
| Yang 2019 | HOMA-IR | First-degree FHD, Age, SUA, BMI, Menopausal Period, eGFR | Lumbar spine BMD and HOMA-IR= 0.103 Femoral neck BMD and HOMA-IR= 0.091 | Positive correlation between HOMA-IR and BMD |
| de Araújo 2020 | HOMA-IR | Age and BMI | L1-L4 BMD and HOMA-IR= 0.013 Total hip BMD and HOMA-IR= 0.018 Femoral neck BMD and HOMA-IR= 0.01 1/3 radius BMD and HOMA-IR= 0.01 L3 BMD and HOMA-IR= 0.038 | No association between HOMA-IR and BMD |
| Zhan 2023 | TyG | Age, Gender, Race, Education, Moderate Recreational Activities, Diabetes, UA, Calcium, Phosphorus, WC, LDL-C, Smoking, 25(OH)D, Antihyperlipidemic Agents | Lumbar spine BMD and TyG= -0.01(-0.02, 0) Whole-body BMD and TyG= -0.01(-0.01, -0.001) Subtotal BMD and TyG= -0.005(-0.011, 0) | Negative association between TyG index and BMD |
| Xuan 2024 | TyG-BMI | Age, Race/Ethnicity, Education, Marital Status, Drinking, Smoking, SBP, DBP, TC, HDL-C, LDL-C, Family of Osteoporosis, Physical Activity, PIR, Calcium, Phosphorus, 25(OH)D3 | Femoral neck BMD and TyG-BMI = 0.058 (0.045, 0.072) | Positive association between TyG-BMI and BMD |
| Chen 2023 | VAI | Race, Gender, Age, Education Level, Smoked At Least 100 Cigarettes, Moderate Activities, Diabetes, Family PIR, BUN, AST, ALP, ALT, Cr, Phosphorus, TC, Calcium, and Total Protein | Femoral BMD and VAI= 0.016 | Positive association between VAI and BMD |
| Ye 2023 | HOMA-IR | Age, FSH, CRP, And IFG, Physical Activity, Drinking and Smoking, BMI | Femoral neck BMD and HOMA-IR= 0.025 (0.003, 0.047) | Positive association between HOMA-IR and BMD |
| Yoon 2021 (men) | TyG, HOMA-IR | Age, BMI, 25(OH)D, Physical Activity, Smoking, and Drinking | Lumbar spine BMD and TyG= -0.028 Femoral neck BMD and TyG= -0.085 Total hip BMD and TyG= -0.046 Whole body BMD and TyG= -0.098 | In men, negative association between TyG and femoral neck, total hip, and whole-body BMD and HOMA-IR and femoral neck and whole body. No association for others. |
| Yoon 2021 (women) | TyG, HOMA-IR | Age, BMI, 25(OH)D, Physical Activity, Smoking, and Drinking | Men Lumbar spine BMD and TyG= -0.028 Femoral neck BMD and TyG= -0.085 Total hip BMD and TyG= -0.046 Whole body BMD and TyG= -0.098 Women Lumbar spine BMD and TyG= 0.016 Femoral neck BMD and TyG= -0.071 Total hip BMD and TyG= -0.003 Whole body BMD and TyG= -0.065 Men Lumbar spine BMD and HOMA-IR= -0.014 Femoral neck BMD and HOMA-IR= -0.051 Total hip BMD and HOMA-IR= -0.021 Whole body BMD and HOMA-IR= -0.041 Women Lumbar spine BMD and HOMA-IR= 0.006 Femoral neck BMD and HOMA-IR= -0.043 Total hip BMD and HOMA-IR= 0.013 Whole body BMD and TyG= -0.026 | In women, negative association between TyG and femoral neck and whole-body BMD, No association for others. |
| Sun 2023 | VAI | Gender, Age, Race, Education Level, Marital Status, PIR, Smoking, Work Activity, BUN, Calcium, Phosphorus, and SUA | Total femur BMD and VAI= 0.006 (0.004-0.009) Femur neck BMD and VAI= 0.004 (0.002-0.006) Trochanter BMD and VAI= 0.005 (0.003-0.007) Intertrochanter BMD and VAI= 0.007 (0.004-0.010) | Positive association between VAI and BMD |
| Pu 2023 | METS-IR | Age, Race, Education, Marital Status, PIR, Smoking, Alcohol, Hypertension, Calcium, 25(OH)D, TC, LDL-C, SCr, SUA, and BUN | Total femoral BMD and METS-IR= 0.005(0.004, 0.006) Lumbar spine BMD and METS-IR= 0.005(0.004, 0.006) | Positive association between METS-IR and BMD |
| Wang 2020 | HOMA-IR | Age, BMI, Height, and Weight | HOMA-IR and forearm BMD= -0.15 | Negative association of HOMA-IR and BMD |
| Tian N 2024 | TyG, TyG-BMI, TyG-WHtR, TyG-WC | Age, Gender, Race, Education Level, Family PIR, ALP, BUN, CPK, Creatinine, Phosphorus, Calcium, UA, Total Bilirubin, Glycohemoglobin, TC, HDL-C, LDL-C and 25OHD2, 25OHD3 | Whole-body BMD and TyG = 0.0124 (0.001, 0.024) Whole-body BMD and TyG-BMI = 0.0004 (0.0003, 0.0004) Whole-body BMD and TyG-WHTR = 0.012 (0.008, 0.016) Whole-body BMD and TyG-WC = 0.0001 (0.0001, 0.0001) | Positive association of TyG, TyG-BMI, TyG-WHTR, TyG-WC and BMD |
| Chen 2024 | TyG | Age, BMI, Race, CVD history, FPG, fasting insulin, and 25(OH)D3 | Men Femur neck BMD and TyG = −0.0003 (−0.02-0.01) Women Femur neck BMD and TyG = −0.002 (−0.03-0.02) | No association between TyG and BMD |
| Tian C 2024 | TyG-BMI | Age, Sex, Race, PIR, Education Attainment, Central Obesity, Calcium, Phosphorus, ALP, ALT, AST, 25(OH)D3, Physical Activity Level, Use of Glucocorticoid, Fracture History, Smoked At Least 100 Cigarettes in Life, Had At Least 12 Alcohol Drinks Past 1 Year, Hypertension | Total femur BMD and TyG-BMI= 0.002 (0.002, 0.002) Femur neck BMD and TyG-BMI= 0.001 (0.001, 0.002) Trochanter BMD and TyG-BMI= 0.001 (0.001, 0.001) Intertrochanter BMD and TyG-BMI= 0.002 (0.002, 0.002) | Positive association between TyG-BMI and BMD |
| Wen 2022 | TyG-BMI | Age, Smoking, Drinking, Previous Fracture, Parental Hip Fracture. | Men Femur neck BMD and TyG-BMI= 0.224 Lumbar spine BMD and TyG-BMI= 0.185 Total hip BMD and TyG-BMI= 0.271 Women Femur neck BMD and TyG-BMI= 0.279 Lumbar spine BMD and TyG-BMI= 0.192 Total hip BMD and TyG-BMI= 0.415 | Positive association between TyG-BMI and BMD |
| Shao 2024 | METS-IR | Gender, Age, Race, Education, Marital Status, Smoking, Activity, Hypertension, Diabetes, Calcium, UA, BUN, TC, Insulin or Glucose-Lowering Drugs, Prednisone or Cortisone, History of Osteoporosis | Total femur BMD Q1 = Reference Q2 = 0.052 (0.041, 0.062) Q3 = 0.094 (0.084, 0.105) Q4 = 0.150 (0.140, 0.161) Femur neck BMD Q1 = Reference Q2 = 0.027 (0.016, 0.037) Q3 = 0.061 (0.050, 0.071) Q4 = 0.115 (0.105, 0.126) Total spine BMD Q1 = Reference Q2 = 0.051 (0.041, 0.062) Q3 = 0.064 (0.053, 0.074) Q4 = 0.108 (0.098, 0.119) | Positive association between METS-IR and BMD |
| Sretenović 2021 | HOMA-IR | FPG, Vit D, Somatropin, anti-TPO, TSH, Ft4, anti-Tg, IGF, PTH | Hip BMD and HOMA-IR = 0.036 (-6.036, 7.239) Lumbar spine BMD and HOMA-IR = 0.032 (-5.719, 6.807) Femoral neck BMD and HOMA-IR = 0.311 (-.093, 8.762) Hip T-score and HOMA-IR = 0.066 (-2.732, 3.005) Spine T-score and HOMA-IR = 0.387 (-2.156, 3.332) Hip Z-score and HOMA-IR = 0.274 (-2.218, 3.373) Spine Z-score and HOMA-IR = -0.154 (-3.011, 2.544) | No association between HOMA-IR and hip and spine BMD |
Summary of main findings.
25(OH)D, 25-Hydroxyvitamin D; 25(OH)D2, 25-Hydroxyvitamin D2; 25(OH)D3, 25-Hydroxyvitamin D3; ALP, Alkaline Phosphatase; ALT, Alanine Transaminase; AST, Aspartate Transaminase; BFM, Body Fat Mass; BMD, Bone Mineral Density; BMI, Body Mass Index; BUN, Blood Urea Nitrogen; CPK, Creatine Phosphokinase; CRP, C-Reactive Protein; DBP, Diastolic Blood Pressure; eGFR, Estimated Glomerular Filtration Rate; FMI, Fat Mass Index; FPG, Fasting Plasma Glucose; FSH, Follicle-Stimulating Hormone; HDL-C, High-Density Lipoprotein Cholesterol; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance; HOMA-β, Homeostatic Model Assessment for Beta-cell Function; Hs-CRP, High-sensitivity C-Reactive Protein; IFG, Impaired Fasting Glucose; LMI, Lean Mass Index; LDL-C, Low-Density Lipoprotein Cholesterol; MetS-IR, Metabolic Syndrome Insulin Resistance; PIR, Poverty Income Ratio; SBP, Systolic Blood Pressure; SCr, Serum Creatinine; SMM, Skeletal Muscle Mass; SUA, Serum Uric Acid; TC, Total Cholesterol; TG, Triglycerides; TBS, Trabecular Bone Score; tOC, Total Osteocalcin; TyG, Triglyceride Glucose Index; TyG-BMI, TyG-Body Mass Index; TyG-WC, TyG-Waist Circumference; TyG-WHtR, TyG-Waist-to-Height Ratio; UA, Uric Acid; VAI, Visceral Adiposity Index; VFA, Visceral Fat Area; WC, Waist Circumference; WHR, Waist-to-Hip Ratio; ucOC, Undercarboxylated Osteocalcin.
Discussion
A total of 27 cohorts and cross-sectional studies consisting of 71,525 participants were included in this systematic review to evaluate the possible association of insulin resistance surrogate indices such as HOMA (IR and β), TyG and its derivates (including TyG-BMI, TyG-WtHR, and TyG-WC), as well as METS-IR and VAI, with BMD of different sites. The HOMA index, which has been examined excessively in several studies, appears to have no association with BMD. The majority of the literature indicates that a higher TyG index is linked to a lower BMD. Although other insulin resistance indices (TyG derivates, METS-IR, and VAI) have been investigated in few studies and there is limited evidence of their association, they have a positive association with BMD (Figure 2).
Figure 2
Insulin resistance, defined as peripheral tissue’s failure to respond to insulin, is a key feature of metabolic syndrome and an increasing risk factor for BMD loss (, ). Conversely, insulin promotes osteoblast proliferation and survival, resulting in increased bone mass. Because insulin resistance and hyperinsulinism are major causes of diabetes, bone mass could be increased in type 2 diabetic patients (, ). Therefore, insulin resistance could conceivably have mixed effects on bone mass. In IR, insulin signaling in osteoblasts is impaired, reducing their activity and leading to decreased bone formation (). Additionally, IR promotes chronic inflammation, elevating pro-inflammatory cytokines like TNF-α and IL-6, which stimulate osteoclasts (bone-resorbing cells), accelerating bone resorption (). Elevated levels of advanced glycation end products (AGEs) in IR further degrade bone quality by impairing collagen structure, exacerbating bone fragility (). Insulin resistance promotes the differentiation of mesenchymal stem cells into adipocytes rather than osteoblasts within the bone marrow. This shift increases marrow adipose tissue (MAT), which negatively correlates with BMD and contributes to overall bone loss. The presence of excess adipose tissue in the marrow can disrupt the delicate balance between bone formation and resorption (). Elevated glucose levels can further impair osteoblastogenesis by activating pathways that promote adipogenesis while inhibiting the expression of key osteogenic transcription factors like Runx2. This dual effect exacerbates the decline in bone mass associated with IR (). Due to the lack of a standardized definition of insulin resistance, various indices with different components have been established that can probably predict it and are widely used in studies ().
The HOMA-IR and HOMA-β homeostasis models rely on basal plasma insulin and glucose measurements to assess insulin resistance and pancreatic β cell activity, respectively (58). They have been widely validated and utilized in clinical and epidemiological investigations (59–61). The triglyceride and glucose index, also known as the TyG index, is another insulin resistance surrogate index that has moderate predictive accuracy (62). The TyG index’s accuracy can be improved by combining it with adiposity indicators such as BMI, WC, and waist-to-height ratio, resulting in TyG-BMI, TyG-WT, and TyG-WtHR (63, 64). As for TyG-BMI, high triglycerides mechanistically lead to lipotoxicity by accumulating in non-adipose tissues, which impairs insulin signaling and disrupts pancreatic β-cell function, increasing IR. High glucose levels also heighten oxidative stress by raising reactive oxygen species (ROS) levels, which harm β-cells. Additionally, excess visceral fat—often reflected in elevated TyG-BMI—contributes to chronic low-grade inflammation, further inhibiting insulin pathways. The TyG-WtHR combines TyG with waist-to-height ratio, highlighting central obesity’s role in IR. Central obesity elevates circulating free fatty acids, impairing insulin signaling in muscles and liver. This also leads to inflammation, oxidative stress, and reduced metabolic flexibility—the body’s ability to alternate between fat and carbohydrate oxidation—which disrupts glucose homeostasis. Lastly, the TyG-WC index incorporates waist circumference to assess IR risks tied to abdominal obesity. High waist circumference, linked to greater visceral fat, increases pro-inflammatory cytokine secretion, contributing to systemic IR. It also correlates with altered lipid metabolism, with elevated triglycerides and reduced HDL cholesterol worsening insulin sensitivity.
The visceral adiposity index, or VAI, is another insulin resistance surrogate composed of BMI, WC, triglyceride, and HDL levels, VAI is another metric that probably predicts insulin resistance (65). The METS-IR index incorporates components of metabolic syndrome, including waist circumference, blood pressure, fasting glucose, and lipids, to gauge IR. This syndrome affects IR through adipokine dysregulation, where altered adipokine secretion (e.g., elevated leptin and reduced adiponectin) fosters systemic inflammation and leptin resistance, leading to hyperglycemia. It also induces endothelial dysfunction, reducing nitric oxide and impairing glucose uptake (66). Although the variety of insulin resistance surrogates and their components enables the assessment of many aspects of insulin resistance in individuals, it also brings significant heterogeneity in their implications and associated expected outcomes, such as BMD. The next section attempts to construct a framework for the interpretation and implementation of each insulin resistance surrogate in predicting BMD.
Among studies that evaluated the association between HOMA-IR and BMD, 8 found no association, 7 found a negative association, and 2 found a positive association. A total of two studies that evaluated the association between BMD and HOMA-β also indicated no association. Apparently, the formulas of HOMA-IR and HOMA-β, which contain fasting plasma glucose and insulin, can be valid and reliable in the prediction of insulin resistance (67). However, they failed to estimate BMD accurately, as the majority of studies indicated no association between them.
The review of the studies that assessed the TyG index and BMD association identified three negative associations, two no associations, and one positive association. As a result, the serum triglyceride and glucose index, which has moderate to low accuracy in diagnosing insulin resistance (62), is unlikely to be associated with BMD. However, when the details of the results were reviewed, it turned out that all three studies that found a negative relationship between TyG and BMD investigated whole-body BMD. None of the studies that found no association between TyG and BMD did so with whole-body BMD. This topic could be the focus of future research, as three studies may not be enough to provide evidence of a negative association between TyG and whole-body BMD.
TyG-BMI was found to be positively associated with BMD in three studies, while TyG-WtHR and TyG-WC were positively associated with BMD in one study. The potential positive correlation between TyG-BMI and higher BMD could be explained by incorporating BMI into the TyG formula, as increased body weight increases mechanical load and activates osteoblasts, and each unit increase in BMI is associated with a 0.0082 g/cm2 increase in BMD (68–70). Two studies found a positive connection between VAI and BMD, while one found a favorable association between METS-IR and BMD. Although there is limited evidence of an association between BMD, METS-IR, and VAI, this can be explained by the fact that both indices include an integrated BMI component (the VAI formula contains HDL, WC, triglyceride, and BMI, and the METS-IR formula contains fasting glucose, fasting triglyceride, HDL-c, and BMI).
This study carried several limitations. First, the variety in populations and insulin resistance indices across the included studies precluded the possibility of performing a meta-analysis. Additionally, the cross-sectional nature of most included studies restricts the ability to infer causality between insulin resistance and BMD. Variability in the covariates adjusted for in the different studies may also introduce bias, as some relevant confounders might not have been consistently accounted for. Furthermore, the diverse characteristics of the sample size of studies in sex, physiologic parameters (e.g., menstruation status), and underlying disease contribute to the heterogeneity of studies on the association between insulin resistance surrogates and BMD. Another limitation is that the relationship between insulin resistance indices and specific bone quality parameters, such as bone microarchitecture and strength, has not been explored in the included studies, calling for further research focusing on how insulin resistance affects bone health beyond bone mineral density. To establish a firm conclusion, the link between insulin resistance indices and restricted evidence, particularly the TyG index, should be investigated in cohort studies.
Conclusion
Despite the heterogenicity of studies on the association between the HOMA index and BMD, there is probably no association. Although there are few studies on other insulin resistance surrogate indices, TyG may have a negative association with whole BMD. Other insulin resistance indices, such as TyG derivates, VAI, and METS-IR, are observed to have positive associations, which may be due to the addition of BMI into their formula. Future research should prioritize conducting longitudinal studies in order to explain causation and gain a deeper understanding of the mechanisms involved in these correlations. Research should investigate how lifestyle interventions, like diet and exercise, can affect insulin resistance levels and their influence on bone health.
Statements
Data availability statement
The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.
Author contributions
AS: Writing – review & editing, Writing – original draft, Validation, Supervision, Methodology, Investigation, Conceptualization. AA: Writing – review & editing, Writing – original draft, Project administration, Methodology, Formal analysis, Data curation. AG-R: Writing – review & editing, Writing – original draft, Visualization, Resources, Conceptualization. FK: Writing – review & editing, Writing – original draft, Investigation. NZ: Writing – review & editing, Writing – original draft, Validation, Software, Investigation, Data curation. SA: Writing – review & editing, Writing – original draft, Data curation. YT: Writing – review & editing, Writing – original draft, Data curation. AH: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Investigation.
Funding
The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article.
Conflict of interest
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
Supplementary material
The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fendo.2024.1499479/full#supplementary-material
Abbreviations
BMD, Bone Mineral Density; HOMA-IR, Homeostatic Model Assessment for Insulin Resistance; TyG, Triglyceride Glucose Index; VAI, Visceral Adiposity Index; BMI, Body Mass Index; WC, Waist Circumference; METS-IR, Metabolic Syndrome Insulin Resistance; LP-IR, Lipoprotein Insulin Resistance Index; HDL-C, High-Density Lipoprotein Cholesterol; Adipo-IR, Adipose Insulin Resistance Index; DM, Diabetes Mellitus; IR, Insulin Resistance; TyG-BMI, Triglyceride Glucose Index-Body Mass Index; TyG-WTHR, Triglyceride Glucose Index-Waist-to-Height Ratio; TyG-WC, Triglyceride Glucose Index-Waist Circumference; TyG-NC, Triglyceride Glucose Index-Neck Circumference; TyG-NHtR, Triglyceride Glucose Index-Neck to Height Ratio; HOMA-β, Homeostatic Model Assessment for Beta-Cell Function.
References
1
ZhuZYuPWuYWuYTanZLingJet al. Sex specific global burden of osteoporosis in 204 countries and territories, from 1990 to 2030: an age-period-cohort modeling study. J nutrition Health aging. (2023) 27:767–74. doi: 10.1007/s12603-023-1971-4
2
ShenYHuangXWuJLinXZhouXZhuZet al. The global burden of osteoporosis, low bone mass, and its related fracture in 204 countries and territories, 1990-2019. Front Endocrinol (Lausanne). (2022) 13:882241. doi: 10.3389/fendo.2022.882241
3
Wilson-BarnesSLLanham-NewSALambertH. Modifiable risk factors for bone health & fragility fractures. Best Pract Res Clin Rheumatol. (2022) 36:101758. doi: 10.1016/j.berh.2022.101758
4
PouresmaeiliFKamalidehghanBKamareheiMGohYM. A comprehensive overview on osteoporosis and its risk factors. Ther Clin Risk Manage. (2018) 14:2029–49. doi: 10.2147/TCRM.S138000
5
SrikanthanPCrandallCJMiller-MartinezDSeemanTEGreendaleGABinkleyNet al. Insulin resistance and bone strength: findings from the study of midlife in the United States. J Bone Miner Res. (2014) 29:796–803. doi: 10.1002/jbmr.2083
6
YangJHongNShimJ-SRheeYKimHC. Association of insulin resistance with lower bone volume and strength index of the proximal femur in nondiabetic postmenopausal women. J Bone Metab. (2018) 25:123. doi: 10.11005/jbm.2018.25.2.123
7
JainRKVokesT. Visceral adipose tissue is negatively associated with bone mineral density in NHANES 2011-2018. J Endocr Soc. (2023) 7:bvad008. doi: 10.1210/jendso/bvad008
8
KimJHaJJeongCLeeJLimYJoKet al. Bone mineral density and lipid profiles in older adults: a nationwide cross-sectional study. Osteoporos Int. (2023) 34:119–28. doi: 10.1007/s00198-022-06571-z
9
ChinKYWongSKEkeukuSOPangKL. Relationship between metabolic syndrome and bone health - an evaluation of epidemiological studies and mechanisms involved. Diabetes Metab Syndr Obes. (2020) 13:3667–90. doi: 10.2147/DMSO.S275560
10
SinghBSaxenaA. Surrogate markers of insulin resistance: A review. World J Diabetes. (2010) 1:36–47. doi: 10.4239/wjd.v1.i2.36
11
HuangRChengZJinXYuXYuJGuoYet al. Usefulness of four surrogate indexes of insulin resistance in middle-aged population in Hefei, China. Ann Med. (2022) 54:622–32. doi: 10.1080/07853890.2022.2039956
12
FiorentinoTVMariniMASuccurroEAndreozziFSestiG. Relationships of surrogate indexes of insulin resistance with insulin sensitivity assessed by euglycemic hyperinsulinemic clamp and subclinical vascular damage. BMJ Open Diabetes Res Care. (2019) 7:e000911. doi: 10.1136/bmjdrc-2019-000911
13
RattanathamRTangpongJChatatikunMSunDKawakamiFImaiMet al. Assessment of eight insulin resistance surrogate indexes for predicting metabolic syndrome and hypertension in Thai law enforcement officers. PeerJ. (2023) 11:e15463. doi: 10.7717/peerj.15463
14
AntoniolliLPNedelBLPazinatoTCde Andrade MesquitaLGerchmanF. Accuracy of insulin resistance indices for metabolic syndrome: a cross-sectional study in adults. Diabetol Metab Syndrome. (2018) 10:65. doi: 10.1186/s13098-018-0365-y
15
ZhuoMChenZZhongM-LLeiFQinJ-JLiuSet al. Association of insulin resistance with bone mineral density in a nationwide health check-up population in China. Bone. (2023) 170:116703. doi: 10.1016/j.bone.2023.116703
16
GuPPuBXinQYueDLuoLTaoJet al. The metabolic score of insulin resistance is positively correlated with bone mineral density in postmenopausal patients with type 2 diabetes mellitus. Sci Rep. (2023) 13:8796. doi: 10.1038/s41598-023-32931-8
17
FuYHLiuWJLeeCLWangJS. Associations of insulin resistance and insulin secretion with bone mineral density and osteoporosis in a general population. Front Endocrinol (Lausanne). (2022) 13:971960. doi: 10.3389/fendo.2022.971960
18
Barrett-ConnorEKritz-SilversteinD. Does hyperinsulinemia preserve bone? Diabetes Care. (1996) 19:1388–92. doi: 10.2337/diacare.19.12.1388
19
ChooMSChoiSRHanJHLeeSHShimYS. Association of insulin resistance with near peak bone mass in the femur and lumbar spine of Korean adults aged 25-35: The Korean National Health and Nutrition Examination Survey 2008-2010. PLoS One. (2017) 12:e0177311. doi: 10.1371/journal.pone.0177311
20
NapoliNConteCPedoneCStrotmeyerESBarbourKEBlackDMet al. Effect of insulin resistance on BMD and fracture risk in older adults. J Clin Endocrinol Metab. (2019) 104:3303–10. doi: 10.1210/jc.2018-02539
21
MoherDLiberatiATetzlaffJAltmanDG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. (2009) 6:e1000097. doi: 10.1371/journal.pmed.1000097
22
MaL-LWangY-YYangZ-HHuangDWengHZengX-T. Methodological quality (risk of bias) assessment tools for primary and secondary medical studies: what are they and which is better? Military Med Res. (2020) 7:7. doi: 10.1186/s40779-020-00238-8
23
PorrittKGomersallJLockwoodC. JBI's Systematic Reviews: Study selection and critical appraisal. Am J Nurs. (2014) 114:47–52. doi: 10.1097/01.NAJ.0000450430.97383.64
24
ShanbhogueVVFinkelsteinJSBouxseinMLYuEW. Association between insulin resistance and bone structure in nondiabetic postmenopausal women. J Clin Endocrinol Metab. (2016) 101:3114–22. doi: 10.1210/jc.2016-1726
25
ChenZHZhouTFBuYTYangL. Bone mineral density saturation as influenced by the visceral adiposity index in adults older than 20 years: a population-based study. Lipids Health Dis. (2023) 22:170. doi: 10.1186/s12944-023-01931-y
26
PuBGuPYueDXinQLuWTaoJet al. The METS-IR is independently related to bone mineral density, FRAX score, and bone fracture among U.S. non-diabetic adults: a cross-sectional study based on NHANES. BMC Musculoskelet Disord. (2023) 24:730. doi: 10.1186/s12891-023-06817-9
27
SunAHuJWangSYinFLiuZ. Association of the visceral adiposity index with femur bone mineral density and osteoporosis among the U.S. older adults from NHANES 2005-2020: a cross-sectional study. Front Endocrinol (Lausanne). (2023) 14:1231527. doi: 10.3389/fendo.2023.1231527
28
ZhanHLiuXPiaoSRongXGuoJ. Association between triglyceride-glucose index and bone mineral density in US adults: a cross sectional study. J Orthop Surg Res. (2023) 18:810. doi: 10.1186/s13018-023-04275-6
29
XuanXSunRPengCLiuLHuangTHuangC. The nonlinear association between triglyceride glucose-body mass index and femoral neck BMD in nondiabetic elderly men: NHANES 2005-March 2020. PLoS One. (2024) 19:e0296935. doi: 10.1371/journal.pone.0296935
30
TianNChenSHanHJinJLiZ. Association between triglyceride glucose index and total bone mineral density: a cross-sectional study from NHANES 2011-2018. Sci Rep. (2024) 14:4208. doi: 10.1038/s41598-024-54192-9
31
TianCLiuJMaMWangSZhangYFengZet al. Association between surrogate marker of insulin resistance and bone mineral density in US adults without diabetes. Arch Osteoporosis. (2024) 19:42. doi: 10.1007/s11657-024-01395-2
32
ChenHHuJLiJLiQLanL. Association between triglyceride-glucose index and femoral bone mineral density in community-dwelling, nondiabetic men and women: a NHANES analysis of 1,928 US individuals. Menopause. (2024) 31:626–33. doi: 10.1097/GME.0000000000002374
33
YangLHuXZhangHPanWYuWGuX. Association of bone mineral density with a first-degree family history of diabetes in normoglycemic postmenopausal women. Menopause. (2019) 26:1284–8. doi: 10.1097/GME.0000000000001396
34
WangYChenFWangHYuCShaoSZhaoMet al. Association between forearm bone mineral density and metabolic obesity in a northern chinese population. Metab Syndr Relat Disord. (2020) 18:251–9. doi: 10.1089/met.2019.0128
35
YeSShiLZhangZ. Effect of insulin resistance on gonadotropin and bone mineral density in nondiabetic postmenopausal women. Front Endocrinol (Lausanne). (2023) 14:1235102. doi: 10.3389/fendo.2023.1235102
36
WenZLiYXuLYueCWangQChenRet al. Triglyceride glucose-body mass index is a reliable indicator of bone mineral density and risk of osteoporotic fracture in middle-aged and elderly nondiabetic chinese individuals. J Clin Med. (2022) 11:5694. doi: 10.3390/jcm11195694
37
ShaoYZhongHLiLDuXZhouXHuaY. Association between the metabolic score of insulin resistance index and bone mineral density: A cross-sectional study. Int J Gerontology. (2024) 18:118–23. doi: 10.6890/IJGE.202404_18(2).0011
38
KimYHChoKHChoiYSKimSMNamGELeeSHet al. Low bone mineral density is associated with metabolic syndrome in South Korean men but not in women: The 2008-2010 Korean National Health and Nutrition Examination Survey. Arch Osteoporos. (2013) 8:142. doi: 10.1007/s11657-013-0142-3
39
ShinDKimSKimKHLeeKParkSM. Association between insulin resistance and bone mass in men. J Clin Endocrinol Metab. (2014) 99:988–95. doi: 10.1210/jc.2013-3338
40
SeoungJYLeeSWKangYMKimMJParkJMMoonHMet al. Association between metabolic risks and bone mineral density in postmenopausal women. CEOG. (2018) 45:671–6. doi: 10.12891/ceog4341.2018
41
YoonJHHongARChoiWParkJYKimHKKangHC. Association of triglyceride-glucose index with bone mineral density in non-diabetic koreans: KNHANES 2008-2011. Calcif Tissue Int. (2021) 108:176–87. doi: 10.1007/s00223-020-00761-9
42
de AraújoIMParreirasESLTCarvalhoALEliasJJr.SalmonCEGde PaulaFJA. Insulin resistance negatively affects bone quality not quantity: the relationship between bone and adipose tissue. Osteoporos Int. (2020) 31:1125–33. doi: 10.1007/s00198-020-05365-5
43
GiudiciKVde FrançaNAGPetersBSEFisbergRMMartiniLA. Associations between markers of glucose metabolism and bone measures among diabetic and non-diabetic adults. J Diabetes Metab Disord. (2021) 20:1247–55. doi: 10.1007/s40200-021-00849-5
44
IkiMTamakiJFujitaYKoudaKYuraAKadowakiEet al. Serum undercarboxylated osteocalcin levels are inversely associated with glycemic status and insulin resistance in an elderly Japanese male population: Fujiwara-kyo Osteoporosis Risk in Men (FORMEN) Study. Osteoporos Int. (2012) 23:761–70. doi: 10.1007/s00198-011-1600-7
45
KalimeriMLeekFWangNXKohHRRoyNCCameron-SmithDet al. Association of insulin resistance with bone strength and bone turnover in menopausal chinese-Singaporean women without diabetes. Int J Environ Res Public Health. (2018) 15:889. doi: 10.3390/ijerph15050889
46
Bazic-SretenovicDVeselinovicMCekerevacINikolic TurnicTAzanjacAKoricanacAet al. The relationship between insulin resistance, bone mineral density and fracture risk in postmenopausal women. Vojnosanitetski pregled. (2021) 79:41–. doi: 10.2298/VSP210216041B
47
Campillo-SánchezFUsategui-MartínRRuiz-de TemiñoÁGilJRuiz-MambrillaMFernández-GómezJMet al. Relationship between insulin resistance (HOMA-IR), trabecular bone score (TBS), and three-dimensional dual-energy X-ray absorptiometry (3D-DXA) in non-diabetic postmenopausal women. J Clin Med. (2020) 9:1732. doi: 10.3390/jcm9061732
48
TahaniNChoudharySBoivinCDawsonCGittoesNGeberhiwotT. Very high bone mineral density in a monogenic form of obesity-associated insulin resistance. Bone. (2021) 143:115756. doi: 10.1016/j.bone.2020.115756
49
CherifRMahjoubFSahliHCheourEVicoLSaklyMet al. Positive association of obesity and insulin resistance with bone mineral density in Tunisian postmenopausal women. J Clin Densitometry. (2018) 21:163–71. doi: 10.1016/j.jocd.2017.05.015
50
NapoliNChandranMPierrozDDAbrahamsenBSchwartzAVFerrariSL. Mechanisms of diabetes mellitus-induced bone fragility. Nat Rev Endocrinol. (2017) 13:208–19. doi: 10.1038/nrendo.2016.153
51
FulzeleKRiddleRCDiGirolamoDJCaoXWanCChenDet al. Insulin receptor signaling in osteoblasts regulates postnatal bone acquisition and body composition. Cell. (2010) 142:309–19. doi: 10.1016/j.cell.2010.06.002
52
ZhuSChenWMassonALiY-P. Cell signaling and transcriptional regulation of osteoblast lineage commitment, differentiation, bone formation, and homeostasis. Cell Discovery. (2024) 10:71. doi: 10.1038/s41421-024-00689-6
53
ConteCEpsteinSNapoliN. Insulin resistance and bone: a biological partnership. Acta diabetologica. (2018) 55:305–14. doi: 10.1007/s00592-018-1101-7
54
ParwaniKMandalP. Role of advanced glycation end products and insulin resistance in diabetic nephropathy. Arch Physiol Biochem. (2023) 129:95–107. doi: 10.1080/13813455.2020.1797106
55
TencerovaMOklaMKassemM. Insulin signaling in bone marrow adipocytes. Curr Osteoporos Rep. (2019) 17:446–54. doi: 10.1007/s11914-019-00552-8
56
CiprianiCColangeloLSantoriRRenellaMMastrantonioMMinisolaSet al. The interplay between bone and glucose metabolism. Front Endocrinol (Lausanne). (2020) 11:122. doi: 10.3389/fendo.2020.00122
57
MesinovicJMcMillanLBShore-LorentiCZenginADe CourtenBEbelingPRet al. Sex-specific associations between insulin resistance and bone parameters in overweight and obese older adults. Clin Endocrinol (Oxf). (2019) 90:680–9. doi: 10.1111/cen.2019.90.issue-5
58
MatthewsDRHoskerJPRudenskiASNaylorBTreacherDFTurnerR. Homeostasis model assessment: insulin resistance and β-cell function from fasting plasma glucose and insulin concentrations in man. diabetologia. (1985) 28:412–9. doi: 10.1007/BF00280883
59
YoonHJeonDJParkCEYouHSMoonAE. Relationship between homeostasis model assessment of insulin resistance and beta cell function and serum 25-hydroxyvitamin D in non-diabetic Korean adults. J Clin Biochem Nutr. (2016) 59:139–44. doi: 10.3164/jcbn.15-143
60
KhaliliDKhayamzadehMKohansalKAhanchiNSHasheminiaMHadaeghFet al. Are HOMA-IR and HOMA-B good predictors for diabetes and pre-diabetes subtypes? BMC Endocrine Disord. (2023) 23:39. doi: 10.1186/s12902-023-01291-9
61
HuHNakagawaTHondaTYamamotoSMizoueT. Should insulin resistance (HOMA-IR), insulin secretion (HOMA-β), and visceral fat area be considered for improving the performance of diabetes risk prediction models. BMJ Open Diabetes Res Care. (2024) 12:e003680. doi: 10.1136/bmjdrc-2023-003680
62
Sánchez-GarcíaARodríguez-GutiérrezRMancillas-AdameLGonzález-NavaVDíaz González-ColmeneroASolisRCet al. Diagnostic accuracy of the triglyceride and glucose index for insulin resistance: A systematic review. Int J Endocrinol. (2020) 2020:4678526. doi: 10.1155/2020/4678526
63
BalaCGheorghe-FroneaOPopDPopCCaloianBComsaHet al. The association between six surrogate insulin resistance indexes and hypertension: a population-based study. Metab Syndrome Related Disord. (2019) 17:328–33. doi: 10.1089/met.2018.0122
64
Ramírez-VélezRPérez-SousaMÁGonzález-RuízKCano-GutierrezCASchmidt-RioValleJCorrea-RodríguezMet al. Obesity-and lipid-related parameters in the identification of older adults with a high risk of prediabetes according to the American diabetes association: an analysis of the 2015 health, well-being, and aging study. Nutrients. (2019) 11:2654. doi: 10.3390/nu11112654
65
JiangKLuanHPuXWangMYinJGongR. Association between visceral adiposity index and insulin resistance: A cross-sectional study based on US adults. Front Endocrinol. (2022) 13. doi: 10.3389/fendo.2022.921067
66
BazyarHZare JavidAMasoudiMRHaidariFHeidariZHajializadehSet al. Assessing the predictive value of insulin resistance indices for metabolic syndrome risk in type 2 diabetes mellitus patients. Sci Rep. (2024) 14:8917. doi: 10.1038/s41598-024-59659-3
67
Romo-RomoAAguilar-SalinasCAGómez-DíazRABrito-CórdovaGXGómez-PérezFJLópez-CarrascoMGet al. Validity and reliability of simple surrogate indexes to evaluate beta-cell function and insulin sensitivity. Rev Med Chil. (2022) 150:1458–66. doi: 10.4067/S0034-98872022001101458
68
LloydJTAlleyDEHawkesWGHochbergMCWaldsteinSROrwigDL. Body mass index is positively associated with bone mineral density in US older adults. Arch osteoporosis. (2014) 9:1–8. doi: 10.1007/s11657-014-0175-2
69
MeyerHESøgaardAJFalchJAJørgensenLEmausN. Weight change over three decades and the risk of osteoporosis in men: the Norwegian Epidemiological Osteoporosis Studies (NOREPOS). Am J Epidemiol. (2008) 168:454–60. doi: 10.1093/aje/kwn151
70
AsomaningKBertone-JohnsonERNascaPCHoovenFPekowPS. The association between body mass index and osteoporosis in patients referred for a bone mineral density examination. J women's Health. (2006) 15:1028–34. doi: 10.1089/jwh.2006.15.1028
Summary
Keywords
insulin resistance, bone mineral density, HOMA-IR, TyG index, VAI, systematic review, osteoporosis, BMD
Citation
Shirinezhad A, Azarboo A, Ghaseminejad-Raeini A, Kanaani Nejad F, Zareshahi N, Amiri SM, Tahmasebi Y and Hoveidaei AH (2024) A systematic review of the association between insulin resistance surrogate indices and bone mineral density. Front. Endocrinol. 15:1499479. doi: 10.3389/fendo.2024.1499479
Received
20 September 2024
Accepted
26 November 2024
Published
18 December 2024
Volume
15 - 2024
Edited by
Ali Ghasem-Zadeh, University of Melbourne, Australia
Reviewed by
Kamyar Asadipooya, University of Kentucky, United States
Naibedya Chattopadhyay, Central Drug Research Institute (CSIR), India
Updates
Copyright
© 2024 Shirinezhad, Azarboo, Ghaseminejad-Raeini, Kanaani Nejad, Zareshahi, Amiri, Tahmasebi and Hoveidaei.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Amirhossein Ghaseminejad-Raeini, ahgnr1999@gmail.com
†ORCID: Alireza Azarboo, orcid.org/0000-0003-3399-2360; Amirhossein Ghaseminejad-Raeini, orcid.org/0000-0002-4322-0236; Amir Human Hoveidaei, orcid.org/0000-0003-4607-354X
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.