Vitamin D3 for Life

Live Enlightened

Research libraryVitamin D3 For Life
Study design and methods

Secondary relevance to vitamin D This paper mainly addresses another topic and also reports a vitamin D analysis, result or scientific synthesis.

This paper mainly addresses another topic and also reports a vitamin D analysis, result or scientific synthesis.

Why this paper is included

Supporting source evidence:

Women with at least 1 CMD were more likely to be Black, have a higher mean weight, have lower education and household income, have lower calcium and vitamin D intake, drink fewer alcoholic drinks per week, have lower physical activity scores, be current smokers, and greater report chronic obstructive pulmonary disease and antidepressant use.

Source: Characteristics of the sample. View original paper

Relevance classification does not establish study quality, causation or editorial approval.

Study design Cross-sectional study.

Cross-sectional study

Examines a population at a particular point in time.

Meta-analysis unclear · Blinding unclear / not reported · Human research reported

How this paper was classified

Automatic source-based classification; awaiting editorial verification. This is not a quality score.

We conducted a cross-sectional study using data from the Coronary Artery Risk Development in Young Adults (CARDIA) study.

Source: Methods.

CARDIA participants were examined approximately every 5 years, completing a battery of questionnaires, providing blood specimens, and completing imaging, including chest and abdominal CT at some exams.

Source: Methods.

View supporting evidence in the original paper

What the study types mean
Systematic review
Uses explicit methods to identify, assess and summarise research on a defined question.
Scoping review
Maps the extent and characteristics of research on a topic.
Narrative review
Discusses existing research; systematic-review methods have not been established by this classification.
Review / evidence synthesis
Reviews existing research; the exact review methods need further assessment.
Randomised controlled trial
Assigns participants or groups to comparison interventions by chance.
Non-randomised intervention study
Examines an intervention without random allocation.
Cohort study
Examines exposures and outcomes over time within a defined group.
Case-control study
Compares people with an outcome against people without it, examining exposures.
Cross-sectional study
Examines a population at a particular point in time.
Observational study
Observes exposures and outcomes without assigning the exposure; the subtype is not established here.
Case report
Describes an individual case.
Case series
Describes a group of individual cases.
Secondary analysis
Analyses existing study data; this paper is not automatically another independent trial.
Laboratory study
Investigates biological processes in a laboratory setting.
Qualitative study
Examines experiences or meanings using qualitative methods.
Genetic instrumental-variable study
Uses genetic variation as an instrument to investigate relationships; this is not a randomised treatment trial.
Study protocol
Describes planned methods and is not evidence that the planned study has reported results.
Multiple reported designs
More than one design is explicitly reported; examine the supporting methods.
Study design awaiting verification
The available source does not establish a sufficiently clear study design.

Meta-analysis combines numerical results from studies. Blinding describes who was unaware of treatment allocation. Neither is a separate quality grade. Study design, vitamin D relevance and document format are independent filters.

Full studies

The relationship between cardiometabolic disease and vertebral bone strength, mineral density, and mass in US adults

Cross-sectional study · Journal of the Endocrine Society · 2026

Original publication title

The relationship between cardiometabolic disease and vertebral bone strength, mineral density, and mass in US adults

Nicole C Wright; Rafeka Hossain; Mary L Bouxsein; Tony M Keaveny; David C Lee; Gerhard Hellemann; John Jeffrey Carr; James G Terry; Amy H Warriner; Cora E Lewis

© The Author(s) 2026. Published by Oxford University Press on behalf of the Endocrine Society.

Original publication · Publisher record · CC BY 4.0

Source wording retained. Formatting, section navigation and display-title punctuation adapted for this research record.

Authors’ conclusion

In this diverse cohort of US adults, CMD was significantly associated with lower overall and subcompartment vertebral vBMD in men only. Our findings can be used by endocrinologists and other clinicians managing patients with CMD, particularly in younger men, to support monitoring bone health (eg, bone turnover markers, CT imaging) in this population. This cross-sectional study served as the foundation for a future longitudinal study evaluating how CMD is associated with changes in bone health and vertebral fractures. Future analyses evaluating the causal mechanisms to our observed findings could lead to the addition of CMD or related biomarkers in fracture risk prediction models.

An extract from the source publication. Read it in context.

At a glance

Abstract: What objectives does the paper state?

Evaluate the relationship between CMD and vertebral bone measures.

Source: Objective

Study methods

What methods does the paper describe?

Study design and setting

We conducted a cross-sectional study using data from the Coronary Artery Risk Development in Young Adults (CARDIA) study. CARDIA is a biracial, community-based cohort of Black and White US adults, established in 1985-1986 to examine cardiovascular disease development from young adulthood (ages 18-30 years at baseline) through midlife (43, 44). CARDIA participants were examined approximately every 5 years, completing a battery of questionnaires, providing blood specimens, and completing imaging, including chest and abdominal CT at some exams.

The protocols were approved by Institutional Review Boards of the University of Alabama at Birmingham, Kaiser Permanente of Oakland, California; University of Minnesota; and Northwestern University. Signed informed consent was obtained from all participants at every examination at their respective field centers (Birmingham, AL; Chicago, IL; Minneapolis, MN; Oakland, CA).

This cross-sectional component of the CARDIA-Bone ancillary study, approved by the Tulane University Institutional Review Board, included 1000 men and 1777 women with year 25 (2010-2011) abdominal CT scans, encompassing a random selection of participants within each race and sex group (Fig. 1). Our sample had an equal distribution of Black and White participants within each sex group as in the parent study but included more women (∼2/3) than men by design. We excluded participants with an estimated glomerular filtration rate of <30 mL/min/1.73 m2, calculated using the Chronic Kidney Disease Epidemiology Collaboration equation without race, due to the possibility of chronic kidney disease: mineral bone disorder.

Figure 1

Participant selection flowchart.

Figure image at the original source: open the original paper. This file is not included in this text import.

Exposure: cardiometabolic disease

Our CMD exposure included the following 4 conditions: abdominal obesity, diabetes mellitus (DM), dyslipidemia, and HTN. Abdominal obesity was based on waist circumference, measured via standard protocols at the year 25 visit, and categorized as a binary variable using the thresholds that define metabolic syndrome (≥102 cm in men; ≥88 cm in women) (45). DM was identified in those with a fasting plasma glucose level ≥126 mg/dL, 2-hour oral glucose tolerance test ≥200 mg/dL, HbA1c ≥6.5%, self-reported use of DM, or self-reported DM medications (46).

Dyslipidemia was identified in participants with triglycerides ≥150 mg/dL, HDL-C <40 mg/dL, or self-reported use of lipid-lowering medications. Triglycerides and HDL-C were chosen over other lipid measures as they more quantitatively reflect metabolic dyslipidemia (47). Given that statins, the predominant lipid-lowering medications used, mostly target low-density lipoprotein cholesterol, we created a second dyslipidemia variable for a sensitivity analysis that excluded those meeting the dyslipidemia definition only via self-report of lipid-lowering medications. HTN was identified as a mean systolic blood pressure (BP) ≥130 mm Hg (48, 49) or mean diastolic BP ≥80 mm Hg (an average of the second and third BP measurement) (48, 49), or self-reported use of antihypertensive medications.

We created 2 CMD exposure variables: (1) CMD status (presence of at least 1 CMD) as a binary variable and (2) CMD burden as a categorical variable (0, 1, 2, or ≥3).

Outcome: BCT-based vertebral strength, mineral density, and size

The CARDIA abdominal CT scans were acquired using 64-channel multidetector GE (GE Healthcare Milwaukee, WI, USA; Birmingham and Oakland centers) and Siemens (Siemens, Erlangen, Germany; Chicago and Minneapolis centers) scanners (50). A standardized helical scan protocol limiting coverage to the lower lumbar vertebrae was used to capture images of L3 to L5. Technical parameters were standardized across the CT systems (120 kVp, 1-mm or 1.25-mm helical slices, 0.984 pitch, 0.5-second rotation, 35-cm display field of view with 50-cm reconstruction to provide full subcutaneous adipose coverage).

Tube current was 320 mA for GE scanners and 180 mAs for Siemens scanners, and tube currents were increased by 25% for individuals weighing >100 kg (51, 52). Each participant was scanned lying on a quality control phantom with known quantities of calcium hydroxyapatite and fat (INTable Calibration Pad; Image Analysis, Inc., Columbia, KY, USA). Radiology technologists were trained and certified on the research protocol, and scans were centrally monitored in near real-time to ensure cross-site consistency.

CT images were anonymized and electronically transmitted to the central CT reading center located at Wake Forest University School of Medicine (Winston-Salem, NC, USA).

BCT analysis using the VirtuOst software (O.N. Diagnostics, Berkeley, CA, USA) was performed at L3 or L4 when L3 was not usable (n = 88 [3.2%] of scans) on the de-identified abdominal CT scans. Details of BCT analysis have been previously published, but in summary, BCT combines finite element analysis and biomechanical modeling to perform a virtual stress test to compute a measurement of bone strength, which is the force required to virtually break or fracture the patient’s spine in a standardized loading configuration (38, 39, 53-56).

In addition to vertebral strength, other data obtained from BCT analysis included vBMD and measurements of vertebral size (mass and vertebral body cross-sectional area). The strength, density, and size parameters were generated for the total vertebral body as well as the trabecular and cortical compartments. The average coefficients of variation re-analysis precision (different analyst, same scan) for vertebral strength, mid-vertebral trabecular vBMD, and total vBMD are 0.5%, 0.5%, and 0.3%, respectively (57). The respective short-term precision values (same analyst, repeat scans of the same patient) is estimated to be 1.5%, 1.3%, and 0.7% (data on file at O.N.

Diagnostics). Our vertebral BCT outcomes included total vertebral strength (newtons), total vBMD (mg/cm3), trabecular (mid-vertebral) vBMD, cortical vBMD, and total mass (g).

Covariates

We evaluated the following variables as potential confounders in the relationship between CMD and vertebral strength: sociodemographic (age and race), anthropometric (height and weight), social determinants of health (education level, family income, employment status), nutritional (calcium and vitamin D intake), lifestyle (smoking status, alcohol consumption, physical activity), other comorbidities that could be related to CMD or bone health (thyroid disease, digestive disease, liver disease, depression, multiple sclerosis, chronic obstructive pulmonary disease, cancer), and antidepressants and corticosteroid medications. In women, we also included menopausal status, polycystic ovary syndrome, and use of menopausal hormone therapy.

Except for height and weight, which were assessed using standardized protocols at clinic visits, all covariate data were self-reported. Calcium and vitamin D intake data were only assessed in year 20 using the CARDIA diet history (58). Given habitual food choices in adults (59), we carried forward the year 20 estimates for these analyses. Alcohol consumption was summarized as the number of drinks (wine, beer, or liquor) per week. Physical activity was calculated as total intensity score in exercise units, with 300 exercise units approximately equivalent to 150 minutes per week of moderate-to-vigorous physical activity (60).

Corticosteroids included both oral and inhaled medications.

Statistical analysis

We described the sample characteristics overall and by CMD status (≥1 CMD variable). Given known differences in BMD by sex (61, 62), all analyses were stratified by sex. Differences in characteristics were identified using t tests for continuous variables and χ² tests for categorical variables. For each of the 2 exposure variables, multivariable linear regression models were used to estimate associations between CMD and each vertebral BCT outcome variable. We first adjusted for age, race, height, and weight given the high correlation of the BCT variables with age and body size (63, 64).

The fully adjusted model included significant social determinants of health, nutritional, lifestyle, comorbidities, and medication variables. For the categorical CMD burden variable, we also examined dose: response relationships using linear trend tests. Lastly, we evaluated whether race, age (≤50 vs >50 years), and menopausal status and hormone therapy use in women modified the association between CMD and the vertebral BCT variables using interaction terms (eg, CMD × Race). We stratified the results for all significant interaction terms (P < .05).

All analyses were conducted in Stata v18.5 (StataCorp, College Station, TX, USA).

Source: Methods Source: Methods Source: Methods Source: Methods Source: Methods

Abstract: Context

Cardiometabolic disease (CMD) may increase vertebral fracture risk through deterioration in bone structure and/or tissue properties, compromising bone strength.

Source: Context

Results: Characteristics of the sample

The analytic sample comprised 2728 participants (Fig. 1), with ages ranging from 42 to 59 years; 63.8% were women, and 50.4% were Black. At least 1 CMD was identified in 714 (72.3%) of men and 1226 (70.5%) of women (Table 1). Men were more likely than women to have dyslipidemia (44.8% vs 26.9%), whereas women were more likely to have abdominal obesity than men (53.5% vs 37.7%). The CMD burden was substantial, with 31.0% of men and 29.5% of women having ≥3 CMDs.

Men with at least 1 CMD were more likely to be older, have a higher mean weight, have lower physical activity scores, and be more likely to report cancer and antidepressant use. Women with at least 1 CMD were more likely to be Black, have a higher mean weight, have lower education and household income, have lower calcium and vitamin D intake, drink fewer alcoholic drinks per week, have lower physical activity scores, be current smokers, and greater report chronic obstructive pulmonary disease and antidepressant use.

Table 1

Characteristics of CARDIA participants by cardiometabolic disease status

Men
(n = 988)
Women
(n = 1740)
CMD yes
(n = 714)
CMD no
(n = 274)
P valueaCMD yes
(n = 1226)
CMD no
(n = 514)
P valuea
Sociodemographic
Age, years, mean (SD)50.2 (3.7)49.6 (3.7).02250.2 (3.7)50.1 (3.7).436
Race, n (%).074 <.001
 Black371 (52.0%)125 (45.6%)746 (60.8%)134 (26.1%)
 White343 (48.0%)149 (54.4%)480 (39.2%)380 (73.9%)
Height, mean (SD)177.9 (6.8)177.5 (7.0).408164.3 (6.8)165.0 (6.3).053
Weight, mean (SD)217.8 (44.5)178.3 (24.7) <.001200.6 (47.4)143.8 (20.6) <.001
Social determinants of health, n (%)
 Education level.458 <.001
  ≤High school295 (41.3)110 (40.1)470 (38.3)110 (21.4)
  Some college133 (18.6)41 (15.0)270 (22.0)74 (14.4)
  College graduate159 (22.3)70 (25.5)296 (24.1)183 (35.6)
  Postgraduate126 (17.6)51 (18.6)186 (15.2)145 (28.2)
 Family income.823 <.001
  <$35 000184 (25.8)65 (23.7)363 (29.6)79 (15.4)
  $35 000-<$50 00062 (8.7)25 (9.1)156 (12.7)49 (9.5)
  $50 000-<$75 000104 (14.6)45 (16.4)232 (18.9)96 (18.7)
  ≥$75 000359 (50.3)133 (48.)456 (37.2)279 (54.3)
 Employment status.247.067
  Employed for wages562 (78.7)230 (83.9)918 (74.9)412 (80.2)
  Homemaker100 (14.0)30 (10.9)259 (21.1)91 (17.7)
  Unemployed42 (5.9)11 (4.0)35 (2.9)8 (1.6)
 Nutritional, mean (SD)
  Calcium intake, mg929.1 (747.6)987.3 (859.8).294870.2 (750.9)1086.0 (867.0) <.001
  Vitamin D intake, mcg7.9 (8.9)7.8 (8.0).8677.4 (10.9)9.6 (16.9).002
Lifestyle
 Alcohol drinks/week, mean (SD)7.2 (13.5)7.4 (15.2).8702.8 (5.7)3.9 (5.5) <.001
 Smoking status, n (%).728.013
  Never417 (58.4)167 (60.9)718 (58.6)334 (65.0)
  Past142 (19.9)49 (17.9)278 (22.7)114 (22.2)
  Current148 (20.7)56 (20.4)206 (16.8)60 (11.7)
 Physical activity, mean (SD)
  Total intensity score, EU377.7 (289.4)443.7 (301.6).002245.2 (226.3)370.9 (265.5) <.001
Comorbidities, n (%)
 Thyroid disease18 (2.5)7 (2.6).825160 (13.1)65 (12.6).933
 Digestive disease34 (4.8)10 (3.6).719106 (8.6)36 (7.0).399
 Liver disease26 (3.6)6 (2.2).28640 (3.3)11 (2.1).293
 Depression108 (15.1)32 (11.7).205343 (28.0)133 (25.9).434
 MS0 (0.0)0 (0.0)Not reported7 (0.6)3 (0.6).979
 COPD10 (1.4)0 (0.0).08024 (2.0)2 (0.4).046
 Cancer40 (5.6)6 (2.2).03397 (7.9)60 (11.7).031
Medications, n (%)
 Antidepressants64 (9.0)12 (4.4).047210 (17.1)64 (12.5).019
 Corticosteroids4 (0.6)4 (1.5).15829 (2.4)8 (1.6).286
Women-specific reproductive factors, n (%)
 Postmenopausal521 (42.5)204 (39.7).279
 Hormone therapy96 (7.8)45 (8.8).519
 PCOS53 (4.3)14 (2.7).114
CMD details, n (%)
 Abdominal obesity372 (37.7)930 (53.5)
 Diabetes mellitus149 (15.1)301 (17.3)
 Dyslipidemia443 (44.8)467 (26.9)
 Hypertension476 (48.2)759 (43.7)
CMD burden, n (%)
 1279 (39.0)465 (37.9)
 2214 (30.0)399 (30.6)
 3+221 (31.0)362 (29.5)

Missing data: height: n = 2 (0.1%); weight: n = 3 (0.1%); education level: n = 9 (0.3%); family income: n = 41 (1.5%); alcohol consumption: n = 25 (0.9%); smoking status: n = 39 (1.4%); physical activity total intensity score: n = 23 (0.8%); thyroid disease: n = 9 (0.3%); digestive disease: n = 8 (0.3%); liver disease: n = 8 (0.3%); depression: n = 9 (0.3%); MS: n = 8 (0.3%); COPD: n = 8 (0.3%); cancer: n = 8 (0.3%); antidepressants: n = 12 (0.4%).

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease (status assessed as a binary variable); COPD, chronic obstructive pulmonary disease; EU, exercise units; MS, multiple sclerosis; PCOS, polycystic ovarian syndrome.

a P values for continuous variables are based on t test and χ² test for categorical variables.

Source: Characteristics of the sample Source: Characteristics of the sample

The association between CMD and vertebral BCT variables

Results: The association between CMD and vertebral BCT variables

Men

The mean total strength, total vBMD, and total mass did not differ by overall CMD status; however, men with at least 1 CMD had lower mean cortical vBMD than men without CMD (Mean [SD], 318 [40] vs 325 [40] mg/cm3; P = .016) (Table 2). In the crude regression analysis, CMD status was associated with significantly lower cortical vBMD (β, −6.92 mg/cm3; P = .016) (Table 3). After adjusting for age, race, height, and weight, CMD was marginally associated with lower total vertebral strength (β, −360.56 [186.74] newtons; P = .054), and significantly lower total, trabecular, and cortical vBMD (Table 3, Model 1) but was not associated with total mass (β, 0.16 [0.13] g; P = .215).

In the fully adjusted model, CMD was associated with significantly lower total (β, −6.32 [2.68] mg/cm3; P = .018), trabecular (β, −4.53 [2.29] mg/cm3; P = .048), and cortical (β, −6.63 [3.12] mg/cm3; P = .034) vBMD (Table 3, Model 2). Similarly, there was a linear trend showing that greater CMD burden was associated with lower strength (Table 3, Model 1), and lower total and trabecular vBMD (Table 3, Model 2).

Results were similar when dyslipidemia was defined based on lipid values alone compared with lipid values and self-reported medication use (data not shown).

Table 2

Mean (SD) vertebral bone strength, mineral density, and mass by cardiometabolic disease status

CMD yesCMD no P valuea
Mean (SD)Mean (SD)
Menn = 714n = 274
 Total strength, newtons11 320 (2621)11 185 (2514).465
 Total vBMD, mg/cm3228 (38)231 (38).379
 Midvertebral vBMD, mg/cm3145 (34)145 (37).847
 Cortical vBMD, mg/cm3 318 (40) 325 (40).016
 Total mass, g9.98 (1.86)9.82 (1.84).215
Womenn = 1226n = 514
 Total strength, newtons9258 (2332)9052 (2141).085
 Total vBMD, mg/cm3234 (43)234 (40).968
 Midvertebral vBMD, mg/cm3 154 (39) 147 (37) <.001
 Cortical vBMD, mg/cm3 322 (47) 334 (43) <.001
 Total mass, g7.67 (1.53)7.75 (1.42).312

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease (status assessed as binary variable); vBMD, volumetric bone mineral density.

a P values from t tests comparing unadjusted means by overall CMD status.

Table 3

The association between cardiometabolic disease and vertebral bone strength, mineral density, and mass in men

CrudeModel 1aModel 2b
β (SE) P valueβ (SE) P valueβ (SE) P value
Total strength, newtons
 CMD status, yes vs no134.57 (184.19).465−360.56 (186.74).054−316.32 (190.00).096
 CMD burden
  0Ref.Ref.Ref.
  1−0.09 (220.28)1.000−273.42 (204.65).182−270.00 (208.27).195
  231.08 (236.27).895−458.24 (235.73).052−312.73 (242.78).198
  3+404.78 (234.16).084 −549.58 (268.24).041−426.26 (285.02).135
 Linear trend.102.030.091
Total vBMD, mg/cm3
 CMD status, yes vs no−2.35 (2.67).379 −6.67 (2.64).012 −6.32 (2.68).018
 CMD burden
 0Ref.Ref.Ref.
  1−1.40 (3.20).662−5.03 (2.89).082−5.53 (2.92).058
  2−4.71 (3.43).170 −8.80 (3.33).008 −7.24 (3.38).032
  3+−1.26 (3.40).711 −9.43 (3.79).013 −8.20 (3.87).034
 Linear trend.486.007.028
Trabecular vBMD, mg/cm3
 CMD status, yes vs no0.48 (2.48).847 −4.80 (2.27).035 −4.53 (2.29).048
 CMD burden
  0Ref.Ref.Ref.
  10.71 (2.97).812−3.43 (2.49).168−3.73 (2.50).136
  2−1.71 (3.18).591 −6.49 (2.86).024−5.42 (2.89).061
  3+2.31 (3.15).464 −7.40 (3.26).023 −6.58 (3.31).047
 Linear trend.668.014.038
Cortical vBMD, mg/cm3
 CMD status, yes vs no −6.92 (2.86).016 −6.43 (3.07).036 −6.63 (3.12).034
 CMD burden
  0Ref.Ref.Ref.
  1−4.19 (3.43).222−5.28 (3.36).117−6.19 (3.40).069
  2 −8.22 (3.68).026 −7.73 (3.87).046−6.81 (3.93).084
  3+ −9.11 (3.64).013 −8.93 (4.40).043−8.57 (4.51).057
 Linear trend.006.033.058
Total mass, g
 CMD status, yes vs no0.16 (0.13).215−0.20 (0.13).125−0.18 (0.13).183
 CMD burden
  0Ref.Ref.Ref.
  1−0.03 (0.16).863−0.16 (0.14).258−0.16 (0.14).267
  20.12 (0.17).469−0.25 (0.16).130−0.18 (0.17).272
  3+ 0.44 (0.17).008−0.27 (0.19).152−0.25 (0.19).201
 Linear trend.006.123.202

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; Ref., reference; vBMD, volumetric bone mineral density.

a Model 1 = age, race, height, weight.

b Model 2 = Model 1 + education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, COPD, cancer, and antidepressants.

Women

The mean total strength, total vBMD, and total mass did not differ by overall CMD status in women. Subcompartment differences were observed; women with ≥1 CMD had higher trabecular vBMD (Mean [SD], 154 (39) vs 147 (37) mg/cm3; P < .001) but lower cortical vBMD (322 [47] vs 334 [43] mg/cm3; P < .001) (Table 2). In crude regression analyses, CMD status was significantly associated with higher trabecular vBMD and lower cortical vBMD (Table 4).

However, after adjusting for age, race, height, and weight (Model 1) and all confounders (Model 2), there was no association between CMD status or CMD burden and total strength, vBMD, and mass (Table 4). Again, results were similar when dyslipidemia was defined based on lipid values alone compared with lipid values and self-reported medication use (data not shown).

Table 4

The association between cardiometabolic disease and vertebral bone strength, mineral density, and mass in women

CrudeModel 1aModel 2b
β (SE) P value β (SE) P value β (SE) P value
Total strength, newtons
 CMD status, yes vs no206.36 (119.64).085−100.66 (130.94).442−65.22 (135.68).631
 CMD burden
  0Ref.Ref.Ref.
  1 355.43 (145.64).01528.89 (140.39).83754.23 (145.00).708
  252.70 (151.84).729 −382.78 (162.24).018−322.69 (167.65).054
  3+184.22 (156.14).238−112.76 (177.11).524−83.49 (186.49).654
 Linear trend.533.196.292
Total vBMD, mg/cm3
 CMD status, yes vs no−0.09 (2.23).968−1.71 (2.40).476−1.16 (2.48).641
 CMD burden
  0Ref.Ref.Ref.
  13.92 (2.72).1490.38 (2.57).8810.76 (2.65).775
  2−2.95 (2.83).298 −6.43 (2.97).030−5.30 (3.06).084
  3+−2.09 (2.91).472−1.54 (3.24).636−1.39 (3.41).683
 Linear trend.178.285.347
Trabecular vBMD, mg/cm3
 CMD status (yes vs no) 7.33 (2.02) <.001−1.45 (2.11).491−1.02 (2.18).641
 CMD burden
  0Ref.Ref.Ref.
  1 6.99 (2.46).004−0.63 (2.26).781−0.43 (2.34).853
  2 6.13 (2.56).017−4.30 (2.61).100−3.18 (2.70).239
  3+ 9.08 (2.64).0010.75 (2.85).7941.24 (3.01).679
 Linear trend.001.907.913
Cortical vBMD, mg/cm3
 CMD status (yes vs no) −11.51 (2.42) <.001−1.43 (2.70).596−1.03 (2.79).711
 CMD burden
  0Ref.Ref.Ref.
  1−3.04 (2.93).2981.07 (2.90).7121.19 (2.99).690
  2 −15.23 (3.05) <.001−6.24 (3.35).063−5.18 (3.45).134
  3+ −18.26 (3.14) <.001−3.05 (3.66).405−3.07 (3.84).425
 Linear trend <.001.164.204
Total mass, g
 CMD status, yes vs no−0.08 (0.08).312−0.07 (0.08).388−0.06 (0.09).513
 CMD burden
  0Ref.Ref.Ref.
  10.07 (0.10).4840.01 (0.09).9160.02 (0.09).821
  2−0.17 (0.10).098 −0.22 (0.10).035−0.20 (0.11).063
  3+−0.17 (0.10).091−0.15 (0.11).197−0.13 (0.12).264
Linear trend.023.068.104

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density.

a Model 1 = age, race, height, weight.

b Model 2 = Model 1 + education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.

Source: The association between CMD and vertebral BCT variables Source: The association between CMD and vertebral BCT variables

Results: Effect modification

There were many significant interactions between CMD and age in men (Table 5). CMD status was associated with lower total, trabecular, and cortical vBMD and lower total mass in younger men ≤50 years, whereas there was no significant association between CMD and the BCT measures in men >50 years of age (Fig. 2A-D). Similarly, greater CMD burden was associated with lower total (Fig. 2A) and trabecular vBMD (Fig. 2B) in younger men, whereas there was no association between CMD burden and BCT variables in older men.

There were no significant interactions between CMD and race in both men and women, nor between age, menopausal status, or hormone therapy use in women (Table 5).

Table 5

Race, age, menopausal status, and hormone therapy use interaction term significance level

MenWomen
BinaryCategoricalBinaryCategorical
P valuea P valueb P valuea P valueb P valuea P valueb P valuea P valueb
Race
 Total strength, newtons.529.933.255.212.644.777.984.935
 Total vBMD, mg/cm3.814.483.338.193.759.838.941.940
 Trabecular vBMD, mg/cm3.638.187.464.168.866.713.994.918
 Cortical vBMD, mg/cm3.894.533.380.256.537.489.969.923
 Total mass, g.564.999.189.165.471.483.673.667
Age
 Total strength, newtons.034.058.044.114.614.959.788.867
 Total vBMD, mg/cm3.011.019.013.038.414.666.839.955
 Trabecular vBMD, mg/cm3.007.010.008.014.705.954.962.980
 Cortical vBMD, mg/cm3.013.020.041.089.175.248.426.592
 Total mass, g.015.038.041.116.443.650.563.736
Menopausal status
 Total strength, newtons Not reported Not reported Not reported Not reported.685.944.719.832
 Total vBMD, mg/cm3 Not reported Not reported Not reported Not reported.193.409.373.694
 Trabecular vBMD, mg/cm3 Not reported Not reported Not reported Not reported.473.818.754.965
 Cortical vBMD, mg/cm3 Not reported Not reported Not reported Not reported.093.185.171.387
 Total mass, g Not reported Not reported Not reported Not reported.822.930.610.694
Hormone therapy use
 Total strength, newtons Not reported Not reported Not reported Not reported .737.833.985.968
 Total vBMD, mg/cm3 Not reported Not reported Not reported Not reported .876.785.971.996
 Trabecular vBMD, mg/cm3 Not reported Not reported Not reported Not reported .435.490.872.894
 Cortical vBMD, mg/cm3 Not reported Not reported Not reported Not reported .549.450.900.908
 Total mass, g Not reported Not reported Not reported Not reported .880.933.995.961

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density.

a Biomechinal computed tomograpy= CMD adjusted for age, race, height, and weight.

b Biomechinal computed tomograpy = CMD adjusted for age, race, height, weight, education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.

Figure 2

The association between cardiometabolic disease and vertebral bone mineral density stratified by age group in men. (A) Total vBMD; (B) trabecular vBMD; (C) cortical vBMD; and (D) total mass. CMD status assessed as a binary variable and categorical burden variable. Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density. aModels adjusted for race, height, weight, education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.bInteraction term P values: Age × Total vBMD = .019; Age × Trabecular vBMD = .010; Age × Cortical vBMD = .020; Age × Total Mass = .038.

Figure image at the original source: open the original paper. This file is not included in this text import.
Source: Effect modification Source: Effect modification Source: Effect modification

Discussion

In this cohort of middle-aged, Black and White adults, we found that having at least 1 of the CMDs of interest was associated with lower total and cortical vertebral vBMD in men, whereas there was no relationship between any CMD and any BCT variable in women. Similarly, greater CMD burden was associated with lower vBMD in men, with the strongest associations seen in men with ≥3 CMD conditions. The inverse relationship between CMD (status and burden) and vBMD was statistically significant in younger (≤50 years) compared with no association in older (>50 years) men.

Race did not modify the relationship between CMD and bone strength in either men or women. Age, menopausal status, and hormone therapy use did not modify the association between CMD and vertebral bone strength, density, or mass in women.

To our knowledge, this is the first study evaluating the association between CMD and BCT bone properties other than DXA-based BMD. Previous studies evaluating the relationship of CMD expressed as metabolic syndrome or as the individual conditions used trabecular bone score (TBS) or high-resolution peripheral quantitative CT (HRpQCT) to measure bone structural properties. TBS quantifies gray-scale variations in the DXA image reflecting bone texture, providing an indirect assessment of bone microarchitecture overall (65-67). Earlier studies showed lower TBS (ie, worse microarchitecture) in those with metabolic syndrome or T2DM despite higher DXA BMD (68).

However, more recent studies showed that this association was driven by soft tissue and not directly attributed to T2DM (33, 69-71). HRpQCT directly measures microarchitecture in trabecular and cortical bone at the radius and/or tibia instead of central sites (femoral neck and lumbar spine) used in DXA. Most of the HRpQCT studies are limited to T2DM, which have consistently reported no difference or sometimes greater trabecular BMD and microarchitecture (eg, trabecular number and thickness) in those with compared to those without T2DM (33).

The relationship between T2DM and cortical bone has been mixed, with some showing greater cortical porosity, especially in those with microvascular disease (72, 73), whereas others reported no significant difference in cortical parameters at the tibia (74). Our findings in men only confirm the greater body of HRpQCT evidence showing an adverse relationship with cortical bone. The differences in findings could be explained by our composite CMD variable instead of focusing on T2DM specifically.

There is limited information in population-based studies regarding bone strength and the other conditions included in our CMD exposure (75, 76). One small study on dyslipidemia and bone strength evaluated the relationship between atherogenic index of plasma (a ratio of triglycerides to HDL-C) and TBS and reported higher atherogenic index of plasma was significantly associated with 60% higher odds of low TBS (< 1.230), independent of age, body mass index, smoking, previous fracture, kidney function, and lumbar osteoporosis assessed via DXA in postmenopausal women (77).

On the cardiovascular disease side, a previous study in older adults from the United Kingdom reported that ischemic heart disease was associated with lower cortical vBMD and higher cortical porosity at the distal radius, and similar to our study, this association between ischemic heart disease and cortical vBMD was only significant in men (78).

Numerous studies have evaluated the possible mechanisms through which CMD is associated with bone, including dyslipidemia impacting osteocyte function and increasing bone resorption (79); hyperglycemia and visceral obesity impacting osteoblast function, disrupting bone formation (80); advanced glycation end-products reducing bone elasticity (81); and chronic inflammation associated with most of these conditions enhancing osteoclast activity (80, 82, 83). This lower formation and higher resorption can lead to low bone strength, density, and mass.

These mechanisms of action ideally should be the same in men and women, but we only found a significant association in men. The men in our study had a higher prevalence of dyslipidemia, so it is possible that factors related to dyslipidemia are driving these sex differences. Similarly, the women had a higher prevalence of abdominal obesity. Adipose tissue and estrogen have a complex relationship, with both positive and negative effects on bone (84), so a deeper exploration into these pathways could further elucidate information on the relationship between CMD and vertebral strength properties in women.

Lastly, more research is needed to explore the role sex hormones play in the relationship between CMD and bone health.

This study has many strengths, such as the inclusion of novel vertebral bone measures in Black and White men and women from a community-based cohort study. Given CARDIA’s standardized protocols and extensive data collection, our CMD exposure is highly valid. Likewise, the CARDIA imaging protocols yielded high-quality CT scans, minimizing missingness in the BCT data. The use of CT-based BMD measures reduces the high variability and limited reproducibility of BMD measurements associated with fat layering seen in DXA (85).

Despite these strengths, our study is not without limitations, including its cross-sectional design and limited age range of participants. The clinical definitions for the CMDs have changed throughout the course of the CARDIA study. We used the most current definitions to be consistent with today’s clinical practice, even if definitions were different at year 25 (2010-2011). CARDIA did not measure biomarkers specific to type 1 diabetes mellitus, which affects bone differently than T2DM. Misclassification of diabetes type is possible, but T2DM represents 90% to 95% of people with DM (86), so we anticipated minimal misclassification.

Our abdominal obesity definition was based on waist circumference, which is correlated with body weight. However, by retaining weight as a covariate, we were able to explicitly evaluate abdominal obesity and not general weight-based obesity. We also did not adjust for antiobesity medications. The first-generation antiobesity medications (eg, orlistat and phentermine) available in the study period (2010-2011) did not advance expected weight loss much above the 3% to 8% observed from the intensive lifestyle interventions in the Diabetes Prevention Program (87).

There are also no data to show that these older medications are associated with improved glucose control, insulin resistance, or reduced cardiovascular disease, thus limiting the effect of these medications on bone. Our CMD exposure was a composite of 4 variables, of which 3 had multiple avenues in which to meet the definition, leading to heterogeneity within each condition. Our future analyses will parse out this heterogeneity, along with identifying which condition or group of conditions are driving the associations observed.

Outside of height and weight, all potential confounders were self-reported. Self-reported data are prone to information bias, which could lead to misclassification and increase the type 1 error rate. Our nutrition data were carried forward from year 20. As previously mentioned, dietary habits in adults are relatively habitual. Those newly diagnosed with 1 of the CMDs between years 20 and 25 may have undergone dietary changes. Dietary changes for cardiovascular and metabolic health involve reducing refined carbohydrates and increasing fruit and vegetable consumption, so there could have been an increase in calcium and vitamin D intake, which would have a benefit on bone.

Given the potential for incorrect nutrition data, our observed findings could be an over: or underestimate of the true relationship between CMD and bone outcomes. Our study included 5 outcomes and 2 exposures, with additional effect modification analyses, potentially warranting adjustment for multiple comparisons to minimize the family-wise error rate (88). However, our 5 outcomes were not all independent of each other. Vertebral strength is highly correlated with density and moderately associated with size. Likewise, total vBMD is essentially the combination of trabecular and cortical vBMD.

Given the correlations between our outcomes, we did not formally adjust for multiple comparisons. Lastly, though we evaluated and adjusted for many potential confounders, as with any epidemiologic study, residual confounding could remain.

Source: Discussion Source: Discussion Source: Discussion Source: Discussion Source: Discussion Source: Discussion Source: Discussion

Contributor Information

Nicole C Wright, Email: nwright6@tulane.edu, Center for Health Outcomes, Implementation, and Community-Engaged Sciences, Tulane University, New Orleans, LA 70112, USA.

Rafeka Hossain, Center for Health Outcomes, Implementation, and Community-Engaged Sciences, Tulane University, New Orleans, LA 70112, USA.

Mary L Bouxsein, Center for Advanced Orthopaedic Studies, Beth Israel Deaconess Medical Center, Harvard University, Boston, MA 02215, USA.

Tony M Keaveny, O.N. Diagnostics, Berkeley, CA 94704, USA.

David C Lee, O.N. Diagnostics, Berkeley, CA 94704, USA.

Gerhard Hellemann, Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, Birmingham, AL 35233, USA.

John Jeffrey Carr, Department of Radiology, Vanderbilt University Medical Center, Nashville, TN 37232, USA.

James G Terry, Department of Radiology, Vanderbilt University Medical Center, Nashville, TN 37232, USA.

Amy H Warriner, Division of Endocrinology, Diabetes, and Metabolism, School of Medicine, University of Alabama at Birmingham, Birmingham, AL 35233, USA.

Cora E Lewis, Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL 35233, USA.

Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information Source: Contributor Information

Results

Characteristics of the sample

The analytic sample comprised 2728 participants (Fig. 1), with ages ranging from 42 to 59 years; 63.8% were women, and 50.4% were Black. At least 1 CMD was identified in 714 (72.3%) of men and 1226 (70.5%) of women (Table 1). Men were more likely than women to have dyslipidemia (44.8% vs 26.9%), whereas women were more likely to have abdominal obesity than men (53.5% vs 37.7%). The CMD burden was substantial, with 31.0% of men and 29.5% of women having ≥3 CMDs.

Men with at least 1 CMD were more likely to be older, have a higher mean weight, have lower physical activity scores, and be more likely to report cancer and antidepressant use. Women with at least 1 CMD were more likely to be Black, have a higher mean weight, have lower education and household income, have lower calcium and vitamin D intake, drink fewer alcoholic drinks per week, have lower physical activity scores, be current smokers, and greater report chronic obstructive pulmonary disease and antidepressant use.

Table 1

Characteristics of CARDIA participants by cardiometabolic disease status

Men
(n = 988)
Women
(n = 1740)
CMD yes
(n = 714)
CMD no
(n = 274)
P valueaCMD yes
(n = 1226)
CMD no
(n = 514)
P valuea
Sociodemographic
Age, years, mean (SD)50.2 (3.7)49.6 (3.7).02250.2 (3.7)50.1 (3.7).436
Race, n (%).074 <.001
 Black371 (52.0%)125 (45.6%)746 (60.8%)134 (26.1%)
 White343 (48.0%)149 (54.4%)480 (39.2%)380 (73.9%)
Height, mean (SD)177.9 (6.8)177.5 (7.0).408164.3 (6.8)165.0 (6.3).053
Weight, mean (SD)217.8 (44.5)178.3 (24.7) <.001200.6 (47.4)143.8 (20.6) <.001
Social determinants of health, n (%)
 Education level.458 <.001
  ≤High school295 (41.3)110 (40.1)470 (38.3)110 (21.4)
  Some college133 (18.6)41 (15.0)270 (22.0)74 (14.4)
  College graduate159 (22.3)70 (25.5)296 (24.1)183 (35.6)
  Postgraduate126 (17.6)51 (18.6)186 (15.2)145 (28.2)
 Family income.823 <.001
  <$35 000184 (25.8)65 (23.7)363 (29.6)79 (15.4)
  $35 000-<$50 00062 (8.7)25 (9.1)156 (12.7)49 (9.5)
  $50 000-<$75 000104 (14.6)45 (16.4)232 (18.9)96 (18.7)
  ≥$75 000359 (50.3)133 (48.)456 (37.2)279 (54.3)
 Employment status.247.067
  Employed for wages562 (78.7)230 (83.9)918 (74.9)412 (80.2)
  Homemaker100 (14.0)30 (10.9)259 (21.1)91 (17.7)
  Unemployed42 (5.9)11 (4.0)35 (2.9)8 (1.6)
 Nutritional, mean (SD)
  Calcium intake, mg929.1 (747.6)987.3 (859.8).294870.2 (750.9)1086.0 (867.0) <.001
  Vitamin D intake, mcg7.9 (8.9)7.8 (8.0).8677.4 (10.9)9.6 (16.9).002
Lifestyle
 Alcohol drinks/week, mean (SD)7.2 (13.5)7.4 (15.2).8702.8 (5.7)3.9 (5.5) <.001
 Smoking status, n (%).728.013
  Never417 (58.4)167 (60.9)718 (58.6)334 (65.0)
  Past142 (19.9)49 (17.9)278 (22.7)114 (22.2)
  Current148 (20.7)56 (20.4)206 (16.8)60 (11.7)
 Physical activity, mean (SD)
  Total intensity score, EU377.7 (289.4)443.7 (301.6).002245.2 (226.3)370.9 (265.5) <.001
Comorbidities, n (%)
 Thyroid disease18 (2.5)7 (2.6).825160 (13.1)65 (12.6).933
 Digestive disease34 (4.8)10 (3.6).719106 (8.6)36 (7.0).399
 Liver disease26 (3.6)6 (2.2).28640 (3.3)11 (2.1).293
 Depression108 (15.1)32 (11.7).205343 (28.0)133 (25.9).434
 MS0 (0.0)0 (0.0)Not reported7 (0.6)3 (0.6).979
 COPD10 (1.4)0 (0.0).08024 (2.0)2 (0.4).046
 Cancer40 (5.6)6 (2.2).03397 (7.9)60 (11.7).031
Medications, n (%)
 Antidepressants64 (9.0)12 (4.4).047210 (17.1)64 (12.5).019
 Corticosteroids4 (0.6)4 (1.5).15829 (2.4)8 (1.6).286
Women-specific reproductive factors, n (%)
 Postmenopausal521 (42.5)204 (39.7).279
 Hormone therapy96 (7.8)45 (8.8).519
 PCOS53 (4.3)14 (2.7).114
CMD details, n (%)
 Abdominal obesity372 (37.7)930 (53.5)
 Diabetes mellitus149 (15.1)301 (17.3)
 Dyslipidemia443 (44.8)467 (26.9)
 Hypertension476 (48.2)759 (43.7)
CMD burden, n (%)
 1279 (39.0)465 (37.9)
 2214 (30.0)399 (30.6)
 3+221 (31.0)362 (29.5)

Missing data: height: n = 2 (0.1%); weight: n = 3 (0.1%); education level: n = 9 (0.3%); family income: n = 41 (1.5%); alcohol consumption: n = 25 (0.9%); smoking status: n = 39 (1.4%); physical activity total intensity score: n = 23 (0.8%); thyroid disease: n = 9 (0.3%); digestive disease: n = 8 (0.3%); liver disease: n = 8 (0.3%); depression: n = 9 (0.3%); MS: n = 8 (0.3%); COPD: n = 8 (0.3%); cancer: n = 8 (0.3%); antidepressants: n = 12 (0.4%).

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease (status assessed as a binary variable); COPD, chronic obstructive pulmonary disease; EU, exercise units; MS, multiple sclerosis; PCOS, polycystic ovarian syndrome.

a P values for continuous variables are based on t test and χ² test for categorical variables.

The association between CMD and vertebral BCT variables

Men

The mean total strength, total vBMD, and total mass did not differ by overall CMD status; however, men with at least 1 CMD had lower mean cortical vBMD than men without CMD (Mean [SD], 318 [40] vs 325 [40] mg/cm3; P = .016) (Table 2). In the crude regression analysis, CMD status was associated with significantly lower cortical vBMD (β, −6.92 mg/cm3; P = .016) (Table 3). After adjusting for age, race, height, and weight, CMD was marginally associated with lower total vertebral strength (β, −360.56 [186.74] newtons; P = .054), and significantly lower total, trabecular, and cortical vBMD (Table 3, Model 1) but was not associated with total mass (β, 0.16 [0.13] g; P = .215).

In the fully adjusted model, CMD was associated with significantly lower total (β, −6.32 [2.68] mg/cm3; P = .018), trabecular (β, −4.53 [2.29] mg/cm3; P = .048), and cortical (β, −6.63 [3.12] mg/cm3; P = .034) vBMD (Table 3, Model 2). Similarly, there was a linear trend showing that greater CMD burden was associated with lower strength (Table 3, Model 1), and lower total and trabecular vBMD (Table 3, Model 2).

Results were similar when dyslipidemia was defined based on lipid values alone compared with lipid values and self-reported medication use (data not shown).

Table 2

Mean (SD) vertebral bone strength, mineral density, and mass by cardiometabolic disease status

CMD yesCMD no P valuea
Mean (SD)Mean (SD)
Menn = 714n = 274
 Total strength, newtons11 320 (2621)11 185 (2514).465
 Total vBMD, mg/cm3228 (38)231 (38).379
 Midvertebral vBMD, mg/cm3145 (34)145 (37).847
 Cortical vBMD, mg/cm3 318 (40) 325 (40).016
 Total mass, g9.98 (1.86)9.82 (1.84).215
Womenn = 1226n = 514
 Total strength, newtons9258 (2332)9052 (2141).085
 Total vBMD, mg/cm3234 (43)234 (40).968
 Midvertebral vBMD, mg/cm3 154 (39) 147 (37) <.001
 Cortical vBMD, mg/cm3 322 (47) 334 (43) <.001
 Total mass, g7.67 (1.53)7.75 (1.42).312

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease (status assessed as binary variable); vBMD, volumetric bone mineral density.

a P values from t tests comparing unadjusted means by overall CMD status.

Table 3

The association between cardiometabolic disease and vertebral bone strength, mineral density, and mass in men

CrudeModel 1aModel 2b
β (SE) P valueβ (SE) P valueβ (SE) P value
Total strength, newtons
 CMD status, yes vs no134.57 (184.19).465−360.56 (186.74).054−316.32 (190.00).096
 CMD burden
  0Ref.Ref.Ref.
  1−0.09 (220.28)1.000−273.42 (204.65).182−270.00 (208.27).195
  231.08 (236.27).895−458.24 (235.73).052−312.73 (242.78).198
  3+404.78 (234.16).084 −549.58 (268.24).041−426.26 (285.02).135
 Linear trend.102.030.091
Total vBMD, mg/cm3
 CMD status, yes vs no−2.35 (2.67).379 −6.67 (2.64).012 −6.32 (2.68).018
 CMD burden
 0Ref.Ref.Ref.
  1−1.40 (3.20).662−5.03 (2.89).082−5.53 (2.92).058
  2−4.71 (3.43).170 −8.80 (3.33).008 −7.24 (3.38).032
  3+−1.26 (3.40).711 −9.43 (3.79).013 −8.20 (3.87).034
 Linear trend.486.007.028
Trabecular vBMD, mg/cm3
 CMD status, yes vs no0.48 (2.48).847 −4.80 (2.27).035 −4.53 (2.29).048
 CMD burden
  0Ref.Ref.Ref.
  10.71 (2.97).812−3.43 (2.49).168−3.73 (2.50).136
  2−1.71 (3.18).591 −6.49 (2.86).024−5.42 (2.89).061
  3+2.31 (3.15).464 −7.40 (3.26).023 −6.58 (3.31).047
 Linear trend.668.014.038
Cortical vBMD, mg/cm3
 CMD status, yes vs no −6.92 (2.86).016 −6.43 (3.07).036 −6.63 (3.12).034
 CMD burden
  0Ref.Ref.Ref.
  1−4.19 (3.43).222−5.28 (3.36).117−6.19 (3.40).069
  2 −8.22 (3.68).026 −7.73 (3.87).046−6.81 (3.93).084
  3+ −9.11 (3.64).013 −8.93 (4.40).043−8.57 (4.51).057
 Linear trend.006.033.058
Total mass, g
 CMD status, yes vs no0.16 (0.13).215−0.20 (0.13).125−0.18 (0.13).183
 CMD burden
  0Ref.Ref.Ref.
  1−0.03 (0.16).863−0.16 (0.14).258−0.16 (0.14).267
  20.12 (0.17).469−0.25 (0.16).130−0.18 (0.17).272
  3+ 0.44 (0.17).008−0.27 (0.19).152−0.25 (0.19).201
 Linear trend.006.123.202

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; Ref., reference; vBMD, volumetric bone mineral density.

a Model 1 = age, race, height, weight.

b Model 2 = Model 1 + education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, COPD, cancer, and antidepressants.

Women

The mean total strength, total vBMD, and total mass did not differ by overall CMD status in women. Subcompartment differences were observed; women with ≥1 CMD had higher trabecular vBMD (Mean [SD], 154 (39) vs 147 (37) mg/cm3; P < .001) but lower cortical vBMD (322 [47] vs 334 [43] mg/cm3; P < .001) (Table 2). In crude regression analyses, CMD status was significantly associated with higher trabecular vBMD and lower cortical vBMD (Table 4).

However, after adjusting for age, race, height, and weight (Model 1) and all confounders (Model 2), there was no association between CMD status or CMD burden and total strength, vBMD, and mass (Table 4). Again, results were similar when dyslipidemia was defined based on lipid values alone compared with lipid values and self-reported medication use (data not shown).

Table 4

The association between cardiometabolic disease and vertebral bone strength, mineral density, and mass in women

CrudeModel 1aModel 2b
β (SE) P value β (SE) P value β (SE) P value
Total strength, newtons
 CMD status, yes vs no206.36 (119.64).085−100.66 (130.94).442−65.22 (135.68).631
 CMD burden
  0Ref.Ref.Ref.
  1 355.43 (145.64).01528.89 (140.39).83754.23 (145.00).708
  252.70 (151.84).729 −382.78 (162.24).018−322.69 (167.65).054
  3+184.22 (156.14).238−112.76 (177.11).524−83.49 (186.49).654
 Linear trend.533.196.292
Total vBMD, mg/cm3
 CMD status, yes vs no−0.09 (2.23).968−1.71 (2.40).476−1.16 (2.48).641
 CMD burden
  0Ref.Ref.Ref.
  13.92 (2.72).1490.38 (2.57).8810.76 (2.65).775
  2−2.95 (2.83).298 −6.43 (2.97).030−5.30 (3.06).084
  3+−2.09 (2.91).472−1.54 (3.24).636−1.39 (3.41).683
 Linear trend.178.285.347
Trabecular vBMD, mg/cm3
 CMD status (yes vs no) 7.33 (2.02) <.001−1.45 (2.11).491−1.02 (2.18).641
 CMD burden
  0Ref.Ref.Ref.
  1 6.99 (2.46).004−0.63 (2.26).781−0.43 (2.34).853
  2 6.13 (2.56).017−4.30 (2.61).100−3.18 (2.70).239
  3+ 9.08 (2.64).0010.75 (2.85).7941.24 (3.01).679
 Linear trend.001.907.913
Cortical vBMD, mg/cm3
 CMD status (yes vs no) −11.51 (2.42) <.001−1.43 (2.70).596−1.03 (2.79).711
 CMD burden
  0Ref.Ref.Ref.
  1−3.04 (2.93).2981.07 (2.90).7121.19 (2.99).690
  2 −15.23 (3.05) <.001−6.24 (3.35).063−5.18 (3.45).134
  3+ −18.26 (3.14) <.001−3.05 (3.66).405−3.07 (3.84).425
 Linear trend <.001.164.204
Total mass, g
 CMD status, yes vs no−0.08 (0.08).312−0.07 (0.08).388−0.06 (0.09).513
 CMD burden
  0Ref.Ref.Ref.
  10.07 (0.10).4840.01 (0.09).9160.02 (0.09).821
  2−0.17 (0.10).098 −0.22 (0.10).035−0.20 (0.11).063
  3+−0.17 (0.10).091−0.15 (0.11).197−0.13 (0.12).264
Linear trend.023.068.104

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density.

a Model 1 = age, race, height, weight.

b Model 2 = Model 1 + education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.

Effect modification

There were many significant interactions between CMD and age in men (Table 5). CMD status was associated with lower total, trabecular, and cortical vBMD and lower total mass in younger men ≤50 years, whereas there was no significant association between CMD and the BCT measures in men >50 years of age (Fig. 2A-D). Similarly, greater CMD burden was associated with lower total (Fig. 2A) and trabecular vBMD (Fig. 2B) in younger men, whereas there was no association between CMD burden and BCT variables in older men.

There were no significant interactions between CMD and race in both men and women, nor between age, menopausal status, or hormone therapy use in women (Table 5).

Table 5

Race, age, menopausal status, and hormone therapy use interaction term significance level

MenWomen
BinaryCategoricalBinaryCategorical
P valuea P valueb P valuea P valueb P valuea P valueb P valuea P valueb
Race
 Total strength, newtons.529.933.255.212.644.777.984.935
 Total vBMD, mg/cm3.814.483.338.193.759.838.941.940
 Trabecular vBMD, mg/cm3.638.187.464.168.866.713.994.918
 Cortical vBMD, mg/cm3.894.533.380.256.537.489.969.923
 Total mass, g.564.999.189.165.471.483.673.667
Age
 Total strength, newtons.034.058.044.114.614.959.788.867
 Total vBMD, mg/cm3.011.019.013.038.414.666.839.955
 Trabecular vBMD, mg/cm3.007.010.008.014.705.954.962.980
 Cortical vBMD, mg/cm3.013.020.041.089.175.248.426.592
 Total mass, g.015.038.041.116.443.650.563.736
Menopausal status
 Total strength, newtons Not reported Not reported Not reported Not reported.685.944.719.832
 Total vBMD, mg/cm3 Not reported Not reported Not reported Not reported.193.409.373.694
 Trabecular vBMD, mg/cm3 Not reported Not reported Not reported Not reported.473.818.754.965
 Cortical vBMD, mg/cm3 Not reported Not reported Not reported Not reported.093.185.171.387
 Total mass, g Not reported Not reported Not reported Not reported.822.930.610.694
Hormone therapy use
 Total strength, newtons Not reported Not reported Not reported Not reported .737.833.985.968
 Total vBMD, mg/cm3 Not reported Not reported Not reported Not reported .876.785.971.996
 Trabecular vBMD, mg/cm3 Not reported Not reported Not reported Not reported .435.490.872.894
 Cortical vBMD, mg/cm3 Not reported Not reported Not reported Not reported .549.450.900.908
 Total mass, g Not reported Not reported Not reported Not reported .880.933.995.961

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density.

a Biomechinal computed tomograpy= CMD adjusted for age, race, height, and weight.

b Biomechinal computed tomograpy = CMD adjusted for age, race, height, weight, education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.

Figure 2

The association between cardiometabolic disease and vertebral bone mineral density stratified by age group in men. (A) Total vBMD; (B) trabecular vBMD; (C) cortical vBMD; and (D) total mass. CMD status assessed as a binary variable and categorical burden variable. Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density. aModels adjusted for race, height, weight, education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.bInteraction term P values: Age × Total vBMD = .019; Age × Trabecular vBMD = .010; Age × Cortical vBMD = .020; Age × Total Mass = .038.

Figure image at the original source: open the original paper. This file is not included in this text import.
Source: Results Source: Results Source: Results

Conclusion

In this diverse cohort of US adults, CMD was significantly associated with lower overall and subcompartment vertebral vBMD in men only. Our findings can be used by endocrinologists and other clinicians managing patients with CMD, particularly in younger men, to support monitoring bone health (eg, bone turnover markers, CT imaging) in this population. This cross-sectional study served as the foundation for a future longitudinal study evaluating how CMD is associated with changes in bone health and vertebral fractures. Future analyses evaluating the causal mechanisms to our observed findings could lead to the addition of CMD or related biomarkers in fracture risk prediction models.

Source: Conclusion
Authors’ abstract

Abstract

Context

Cardiometabolic disease (CMD) may increase vertebral fracture risk through deterioration in bone structure and/or tissue properties, compromising bone strength.

Objective

Evaluate the relationship between CMD and vertebral bone measures.

Design

Cross-sectional.

Participants

Coronary Artery Risk Development in Young Adults (CARDIA) study participants with year 25 (2010-2011) abdominal computed tomography (CT) scans (n = 2728).

Exposure

CMD (abdominal obesity, diabetes mellitus, dyslipidemia, or hypertension) status (≥1 vs 0) and burden (0, 1, 2, ≥3).

Outcome Measure(s)

Vertebral strength, volumetric bone mineral density (vBMD), and bone mass, obtained via biomechanical CT analysis.

Methods

Multivariable linear regression evaluated the association between CMD and vertebral bone measures. Sex-specific models were adjusted for age, race, height, weight, and additional social determinants of health, lifestyle factors, and comorbidities. Effect modification by race, age, and menopausal status was evaluated.

Results

Overall, at least 1 CMD was identified in 72.3% of men and 70.5% of women. In the fully adjusted model, CMD status was significantly associated with lower total (β = −6.32 mg/cm3; P = .018), trabecular (β = −4.53 mg/cm3; P = .048), and cortical (β = −6.63 mg/cm3; P = .034) vBMD in men. Greater CMD burden was associated with lower total and trabecular vBMD. The negative association between CMD and vBMD was stronger in men ≤50 years.

There was no association in women.

Conclusion

In this biracial cohort of Black and White adults from the United States, CMD was associated with significantly lower total and subcompartment vertebral vBMD in men.

Keywords:

cardiometabolic disease, bone strength, biomechanical computed tomography, epidemiology

Read the full paper

Cardiometabolic diseases (CMDs) are an interconnected group of conditions with shared pathophysiology, including obesity, insulin resistance, dyslipidemia, hypertension (HTN), and inflammation (1). CMDs account for ∼20 million deaths globally (2). In addition to mortality, CMD increases the risk for other health conditions, including osteoporotic fractures.

Osteoporotic fractures are highly prevalent (3-5), have a significant economic burden (3, 6, 7), and are associated with many adverse outcomes, including death (8, 9). Vertebral fractures are the most common osteoporotic fractures (10). Regarding CMD and vertebral fractures, the most is known about type 2 diabetes mellitus (T2DM). Individual studies and meta-analyses have highlighted that T2DM increases the risk for vertebral fractures by ∼40% (11-16). Abdominal obesity has been shown to be associated with an 18% higher risk of vertebral fractures in a meta-analysis (17), with more recent studies reporting a higher risk of vertebral fractures in those with abdominal obesity based on waist circumference (18) even in those with lower body fat percentage (19) or those with normal body mass index (20).

The relationship between dyslipidemia and vertebral fractures is mixed, with studies showing lower triglycerides in those with incident fractures (21, 22), while others have shown higher triglycerides and high-density lipoprotein cholesterol (HDL-C) were associated with vertebral fractures (23, 24). Similarly, studies specifically evaluating HTN and vertebral fracture are mixed, with a study of Japanese postmenopausal women reporting HTN being associated with a 76% increased risk (25), whereas a Danish study reported HTN was associated with vertebral fracture only in those with a shorter duration of HTN (26).

Lastly, medications used in the treatment of CMDs, such as thiazolidinediones (27), insulin (28, 29), sodium-glucose cotransporter 2 inhibitors (particularly canagliflozin) (30, 31), diuretics (both thiazide and loop), and calcium channel blockers (32), are associated with increased risk of vertebral fractures. Though studies have shown a relationship between the individual CMDs and fracture, many previous studies did not account for the CMD clustering effect, which is particularly important given the high multimorbidity within these conditions.

Such clustering could have an additive or multiplicative effect on the skeleton.

Bone mineral density (BMD), measured via dual energy X-ray absorptiometry (DXA), is 1 of the strongest risk factors for fracture. However, paradoxically, BMD has been shown to be higher in those with CMDs (33), particularly T2DM and abdominal obesity. This suggests that 1 mechanism through which CMDs are associated with higher fracture risk involves deterioration in bone structure and/or bone tissue properties, compromising bone strength. Unfortunately, DXA, the gold standard for assessing BMD, does not directly measure bone strength.

Advanced analysis of clinical computed tomography (CT) scans, coined biomechanical computed tomography (BCT), uses finite element analysis to estimate bone strength from the measured volumetric BMD (vBMD) and vertebral geometry (34). Studies have shown that BCT-based vertebral strength predicts fracture better than or as well as DXA-BMD (35-39). However, to our knowledge, the potential relationship between CMD and BCT bone strength measures has not been studied. Given that recent data show that only 6.8% of adults in the United States (US) have optimal cardiometabolic markers (eg, adiposity and glucose) (40), and the projected increase in the prevalence of CMDs (41, 42), studying the relationship between CMD and vertebral strength measures will provide novel information on how these conditions as a collective could impact fracture risk.

Methods

Study design and setting

We conducted a cross-sectional study using data from the Coronary Artery Risk Development in Young Adults (CARDIA) study. CARDIA is a biracial, community-based cohort of Black and White US adults, established in 1985-1986 to examine cardiovascular disease development from young adulthood (ages 18-30 years at baseline) through midlife (43, 44). CARDIA participants were examined approximately every 5 years, completing a battery of questionnaires, providing blood specimens, and completing imaging, including chest and abdominal CT at some exams.

The protocols were approved by Institutional Review Boards of the University of Alabama at Birmingham, Kaiser Permanente of Oakland, California; University of Minnesota; and Northwestern University. Signed informed consent was obtained from all participants at every examination at their respective field centers (Birmingham, AL; Chicago, IL; Minneapolis, MN; Oakland, CA).

This cross-sectional component of the CARDIA-Bone ancillary study, approved by the Tulane University Institutional Review Board, included 1000 men and 1777 women with year 25 (2010-2011) abdominal CT scans, encompassing a random selection of participants within each race and sex group (Fig. 1). Our sample had an equal distribution of Black and White participants within each sex group as in the parent study but included more women (∼2/3) than men by design. We excluded participants with an estimated glomerular filtration rate of <30 mL/min/1.73 m2, calculated using the Chronic Kidney Disease Epidemiology Collaboration equation without race, due to the possibility of chronic kidney disease: mineral bone disorder.

Figure 1

Participant selection flowchart.

Figure image at the original source: open the original paper. This file is not included in this text import.

Exposure: cardiometabolic disease

Our CMD exposure included the following 4 conditions: abdominal obesity, diabetes mellitus (DM), dyslipidemia, and HTN. Abdominal obesity was based on waist circumference, measured via standard protocols at the year 25 visit, and categorized as a binary variable using the thresholds that define metabolic syndrome (≥102 cm in men; ≥88 cm in women) (45). DM was identified in those with a fasting plasma glucose level ≥126 mg/dL, 2-hour oral glucose tolerance test ≥200 mg/dL, HbA1c ≥6.5%, self-reported use of DM, or self-reported DM medications (46).

Dyslipidemia was identified in participants with triglycerides ≥150 mg/dL, HDL-C <40 mg/dL, or self-reported use of lipid-lowering medications. Triglycerides and HDL-C were chosen over other lipid measures as they more quantitatively reflect metabolic dyslipidemia (47). Given that statins, the predominant lipid-lowering medications used, mostly target low-density lipoprotein cholesterol, we created a second dyslipidemia variable for a sensitivity analysis that excluded those meeting the dyslipidemia definition only via self-report of lipid-lowering medications. HTN was identified as a mean systolic blood pressure (BP) ≥130 mm Hg (48, 49) or mean diastolic BP ≥80 mm Hg (an average of the second and third BP measurement) (48, 49), or self-reported use of antihypertensive medications.

We created 2 CMD exposure variables: (1) CMD status (presence of at least 1 CMD) as a binary variable and (2) CMD burden as a categorical variable (0, 1, 2, or ≥3).

Outcome: BCT-based vertebral strength, mineral density, and size

The CARDIA abdominal CT scans were acquired using 64-channel multidetector GE (GE Healthcare Milwaukee, WI, USA; Birmingham and Oakland centers) and Siemens (Siemens, Erlangen, Germany; Chicago and Minneapolis centers) scanners (50). A standardized helical scan protocol limiting coverage to the lower lumbar vertebrae was used to capture images of L3 to L5. Technical parameters were standardized across the CT systems (120 kVp, 1-mm or 1.25-mm helical slices, 0.984 pitch, 0.5-second rotation, 35-cm display field of view with 50-cm reconstruction to provide full subcutaneous adipose coverage).

Tube current was 320 mA for GE scanners and 180 mAs for Siemens scanners, and tube currents were increased by 25% for individuals weighing >100 kg (51, 52). Each participant was scanned lying on a quality control phantom with known quantities of calcium hydroxyapatite and fat (INTable Calibration Pad; Image Analysis, Inc., Columbia, KY, USA). Radiology technologists were trained and certified on the research protocol, and scans were centrally monitored in near real-time to ensure cross-site consistency.

CT images were anonymized and electronically transmitted to the central CT reading center located at Wake Forest University School of Medicine (Winston-Salem, NC, USA).

BCT analysis using the VirtuOst software (O.N. Diagnostics, Berkeley, CA, USA) was performed at L3 or L4 when L3 was not usable (n = 88 [3.2%] of scans) on the de-identified abdominal CT scans. Details of BCT analysis have been previously published, but in summary, BCT combines finite element analysis and biomechanical modeling to perform a virtual stress test to compute a measurement of bone strength, which is the force required to virtually break or fracture the patient’s spine in a standardized loading configuration (38, 39, 53-56).

In addition to vertebral strength, other data obtained from BCT analysis included vBMD and measurements of vertebral size (mass and vertebral body cross-sectional area). The strength, density, and size parameters were generated for the total vertebral body as well as the trabecular and cortical compartments. The average coefficients of variation re-analysis precision (different analyst, same scan) for vertebral strength, mid-vertebral trabecular vBMD, and total vBMD are 0.5%, 0.5%, and 0.3%, respectively (57). The respective short-term precision values (same analyst, repeat scans of the same patient) is estimated to be 1.5%, 1.3%, and 0.7% (data on file at O.N.

Diagnostics). Our vertebral BCT outcomes included total vertebral strength (newtons), total vBMD (mg/cm3), trabecular (mid-vertebral) vBMD, cortical vBMD, and total mass (g).

Covariates

We evaluated the following variables as potential confounders in the relationship between CMD and vertebral strength: sociodemographic (age and race), anthropometric (height and weight), social determinants of health (education level, family income, employment status), nutritional (calcium and vitamin D intake), lifestyle (smoking status, alcohol consumption, physical activity), other comorbidities that could be related to CMD or bone health (thyroid disease, digestive disease, liver disease, depression, multiple sclerosis, chronic obstructive pulmonary disease, cancer), and antidepressants and corticosteroid medications. In women, we also included menopausal status, polycystic ovary syndrome, and use of menopausal hormone therapy.

Except for height and weight, which were assessed using standardized protocols at clinic visits, all covariate data were self-reported. Calcium and vitamin D intake data were only assessed in year 20 using the CARDIA diet history (58). Given habitual food choices in adults (59), we carried forward the year 20 estimates for these analyses. Alcohol consumption was summarized as the number of drinks (wine, beer, or liquor) per week. Physical activity was calculated as total intensity score in exercise units, with 300 exercise units approximately equivalent to 150 minutes per week of moderate-to-vigorous physical activity (60).

Corticosteroids included both oral and inhaled medications.

Statistical analysis

We described the sample characteristics overall and by CMD status (≥1 CMD variable). Given known differences in BMD by sex (61, 62), all analyses were stratified by sex. Differences in characteristics were identified using t tests for continuous variables and χ² tests for categorical variables. For each of the 2 exposure variables, multivariable linear regression models were used to estimate associations between CMD and each vertebral BCT outcome variable. We first adjusted for age, race, height, and weight given the high correlation of the BCT variables with age and body size (63, 64).

The fully adjusted model included significant social determinants of health, nutritional, lifestyle, comorbidities, and medication variables. For the categorical CMD burden variable, we also examined dose: response relationships using linear trend tests. Lastly, we evaluated whether race, age (≤50 vs >50 years), and menopausal status and hormone therapy use in women modified the association between CMD and the vertebral BCT variables using interaction terms (eg, CMD × Race). We stratified the results for all significant interaction terms (P < .05).

All analyses were conducted in Stata v18.5 (StataCorp, College Station, TX, USA).

Results

Characteristics of the sample

The analytic sample comprised 2728 participants (Fig. 1), with ages ranging from 42 to 59 years; 63.8% were women, and 50.4% were Black. At least 1 CMD was identified in 714 (72.3%) of men and 1226 (70.5%) of women (Table 1). Men were more likely than women to have dyslipidemia (44.8% vs 26.9%), whereas women were more likely to have abdominal obesity than men (53.5% vs 37.7%). The CMD burden was substantial, with 31.0% of men and 29.5% of women having ≥3 CMDs.

Men with at least 1 CMD were more likely to be older, have a higher mean weight, have lower physical activity scores, and be more likely to report cancer and antidepressant use. Women with at least 1 CMD were more likely to be Black, have a higher mean weight, have lower education and household income, have lower calcium and vitamin D intake, drink fewer alcoholic drinks per week, have lower physical activity scores, be current smokers, and greater report chronic obstructive pulmonary disease and antidepressant use.

Table 1

Characteristics of CARDIA participants by cardiometabolic disease status

Men
(n = 988)
Women
(n = 1740)
CMD yes
(n = 714)
CMD no
(n = 274)
P valueaCMD yes
(n = 1226)
CMD no
(n = 514)
P valuea
Sociodemographic
Age, years, mean (SD)50.2 (3.7)49.6 (3.7).02250.2 (3.7)50.1 (3.7).436
Race, n (%).074 <.001
 Black371 (52.0%)125 (45.6%)746 (60.8%)134 (26.1%)
 White343 (48.0%)149 (54.4%)480 (39.2%)380 (73.9%)
Height, mean (SD)177.9 (6.8)177.5 (7.0).408164.3 (6.8)165.0 (6.3).053
Weight, mean (SD)217.8 (44.5)178.3 (24.7) <.001200.6 (47.4)143.8 (20.6) <.001
Social determinants of health, n (%)
 Education level.458 <.001
  ≤High school295 (41.3)110 (40.1)470 (38.3)110 (21.4)
  Some college133 (18.6)41 (15.0)270 (22.0)74 (14.4)
  College graduate159 (22.3)70 (25.5)296 (24.1)183 (35.6)
  Postgraduate126 (17.6)51 (18.6)186 (15.2)145 (28.2)
 Family income.823 <.001
  <$35 000184 (25.8)65 (23.7)363 (29.6)79 (15.4)
  $35 000-<$50 00062 (8.7)25 (9.1)156 (12.7)49 (9.5)
  $50 000-<$75 000104 (14.6)45 (16.4)232 (18.9)96 (18.7)
  ≥$75 000359 (50.3)133 (48.)456 (37.2)279 (54.3)
 Employment status.247.067
  Employed for wages562 (78.7)230 (83.9)918 (74.9)412 (80.2)
  Homemaker100 (14.0)30 (10.9)259 (21.1)91 (17.7)
  Unemployed42 (5.9)11 (4.0)35 (2.9)8 (1.6)
 Nutritional, mean (SD)
  Calcium intake, mg929.1 (747.6)987.3 (859.8).294870.2 (750.9)1086.0 (867.0) <.001
  Vitamin D intake, mcg7.9 (8.9)7.8 (8.0).8677.4 (10.9)9.6 (16.9).002
Lifestyle
 Alcohol drinks/week, mean (SD)7.2 (13.5)7.4 (15.2).8702.8 (5.7)3.9 (5.5) <.001
 Smoking status, n (%).728.013
  Never417 (58.4)167 (60.9)718 (58.6)334 (65.0)
  Past142 (19.9)49 (17.9)278 (22.7)114 (22.2)
  Current148 (20.7)56 (20.4)206 (16.8)60 (11.7)
 Physical activity, mean (SD)
  Total intensity score, EU377.7 (289.4)443.7 (301.6).002245.2 (226.3)370.9 (265.5) <.001
Comorbidities, n (%)
 Thyroid disease18 (2.5)7 (2.6).825160 (13.1)65 (12.6).933
 Digestive disease34 (4.8)10 (3.6).719106 (8.6)36 (7.0).399
 Liver disease26 (3.6)6 (2.2).28640 (3.3)11 (2.1).293
 Depression108 (15.1)32 (11.7).205343 (28.0)133 (25.9).434
 MS0 (0.0)0 (0.0)Not reported7 (0.6)3 (0.6).979
 COPD10 (1.4)0 (0.0).08024 (2.0)2 (0.4).046
 Cancer40 (5.6)6 (2.2).03397 (7.9)60 (11.7).031
Medications, n (%)
 Antidepressants64 (9.0)12 (4.4).047210 (17.1)64 (12.5).019
 Corticosteroids4 (0.6)4 (1.5).15829 (2.4)8 (1.6).286
Women-specific reproductive factors, n (%)
 Postmenopausal521 (42.5)204 (39.7).279
 Hormone therapy96 (7.8)45 (8.8).519
 PCOS53 (4.3)14 (2.7).114
CMD details, n (%)
 Abdominal obesity372 (37.7)930 (53.5)
 Diabetes mellitus149 (15.1)301 (17.3)
 Dyslipidemia443 (44.8)467 (26.9)
 Hypertension476 (48.2)759 (43.7)
CMD burden, n (%)
 1279 (39.0)465 (37.9)
 2214 (30.0)399 (30.6)
 3+221 (31.0)362 (29.5)

Missing data: height: n = 2 (0.1%); weight: n = 3 (0.1%); education level: n = 9 (0.3%); family income: n = 41 (1.5%); alcohol consumption: n = 25 (0.9%); smoking status: n = 39 (1.4%); physical activity total intensity score: n = 23 (0.8%); thyroid disease: n = 9 (0.3%); digestive disease: n = 8 (0.3%); liver disease: n = 8 (0.3%); depression: n = 9 (0.3%); MS: n = 8 (0.3%); COPD: n = 8 (0.3%); cancer: n = 8 (0.3%); antidepressants: n = 12 (0.4%).

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease (status assessed as a binary variable); COPD, chronic obstructive pulmonary disease; EU, exercise units; MS, multiple sclerosis; PCOS, polycystic ovarian syndrome.

a P values for continuous variables are based on t test and χ² test for categorical variables.

The association between CMD and vertebral BCT variables

Men

The mean total strength, total vBMD, and total mass did not differ by overall CMD status; however, men with at least 1 CMD had lower mean cortical vBMD than men without CMD (Mean [SD], 318 [40] vs 325 [40] mg/cm3; P = .016) (Table 2). In the crude regression analysis, CMD status was associated with significantly lower cortical vBMD (β, −6.92 mg/cm3; P = .016) (Table 3). After adjusting for age, race, height, and weight, CMD was marginally associated with lower total vertebral strength (β, −360.56 [186.74] newtons; P = .054), and significantly lower total, trabecular, and cortical vBMD (Table 3, Model 1) but was not associated with total mass (β, 0.16 [0.13] g; P = .215).

In the fully adjusted model, CMD was associated with significantly lower total (β, −6.32 [2.68] mg/cm3; P = .018), trabecular (β, −4.53 [2.29] mg/cm3; P = .048), and cortical (β, −6.63 [3.12] mg/cm3; P = .034) vBMD (Table 3, Model 2). Similarly, there was a linear trend showing that greater CMD burden was associated with lower strength (Table 3, Model 1), and lower total and trabecular vBMD (Table 3, Model 2).

Results were similar when dyslipidemia was defined based on lipid values alone compared with lipid values and self-reported medication use (data not shown).

Table 2

Mean (SD) vertebral bone strength, mineral density, and mass by cardiometabolic disease status

CMD yesCMD no P valuea
Mean (SD)Mean (SD)
Menn = 714n = 274
 Total strength, newtons11 320 (2621)11 185 (2514).465
 Total vBMD, mg/cm3228 (38)231 (38).379
 Midvertebral vBMD, mg/cm3145 (34)145 (37).847
 Cortical vBMD, mg/cm3 318 (40) 325 (40).016
 Total mass, g9.98 (1.86)9.82 (1.84).215
Womenn = 1226n = 514
 Total strength, newtons9258 (2332)9052 (2141).085
 Total vBMD, mg/cm3234 (43)234 (40).968
 Midvertebral vBMD, mg/cm3 154 (39) 147 (37) <.001
 Cortical vBMD, mg/cm3 322 (47) 334 (43) <.001
 Total mass, g7.67 (1.53)7.75 (1.42).312

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease (status assessed as binary variable); vBMD, volumetric bone mineral density.

a P values from t tests comparing unadjusted means by overall CMD status.

Table 3

The association between cardiometabolic disease and vertebral bone strength, mineral density, and mass in men

CrudeModel 1aModel 2b
β (SE) P valueβ (SE) P valueβ (SE) P value
Total strength, newtons
 CMD status, yes vs no134.57 (184.19).465−360.56 (186.74).054−316.32 (190.00).096
 CMD burden
  0Ref.Ref.Ref.
  1−0.09 (220.28)1.000−273.42 (204.65).182−270.00 (208.27).195
  231.08 (236.27).895−458.24 (235.73).052−312.73 (242.78).198
  3+404.78 (234.16).084 −549.58 (268.24).041−426.26 (285.02).135
 Linear trend.102.030.091
Total vBMD, mg/cm3
 CMD status, yes vs no−2.35 (2.67).379 −6.67 (2.64).012 −6.32 (2.68).018
 CMD burden
 0Ref.Ref.Ref.
  1−1.40 (3.20).662−5.03 (2.89).082−5.53 (2.92).058
  2−4.71 (3.43).170 −8.80 (3.33).008 −7.24 (3.38).032
  3+−1.26 (3.40).711 −9.43 (3.79).013 −8.20 (3.87).034
 Linear trend.486.007.028
Trabecular vBMD, mg/cm3
 CMD status, yes vs no0.48 (2.48).847 −4.80 (2.27).035 −4.53 (2.29).048
 CMD burden
  0Ref.Ref.Ref.
  10.71 (2.97).812−3.43 (2.49).168−3.73 (2.50).136
  2−1.71 (3.18).591 −6.49 (2.86).024−5.42 (2.89).061
  3+2.31 (3.15).464 −7.40 (3.26).023 −6.58 (3.31).047
 Linear trend.668.014.038
Cortical vBMD, mg/cm3
 CMD status, yes vs no −6.92 (2.86).016 −6.43 (3.07).036 −6.63 (3.12).034
 CMD burden
  0Ref.Ref.Ref.
  1−4.19 (3.43).222−5.28 (3.36).117−6.19 (3.40).069
  2 −8.22 (3.68).026 −7.73 (3.87).046−6.81 (3.93).084
  3+ −9.11 (3.64).013 −8.93 (4.40).043−8.57 (4.51).057
 Linear trend.006.033.058
Total mass, g
 CMD status, yes vs no0.16 (0.13).215−0.20 (0.13).125−0.18 (0.13).183
 CMD burden
  0Ref.Ref.Ref.
  1−0.03 (0.16).863−0.16 (0.14).258−0.16 (0.14).267
  20.12 (0.17).469−0.25 (0.16).130−0.18 (0.17).272
  3+ 0.44 (0.17).008−0.27 (0.19).152−0.25 (0.19).201
 Linear trend.006.123.202

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; Ref., reference; vBMD, volumetric bone mineral density.

a Model 1 = age, race, height, weight.

b Model 2 = Model 1 + education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, COPD, cancer, and antidepressants.

Women

The mean total strength, total vBMD, and total mass did not differ by overall CMD status in women. Subcompartment differences were observed; women with ≥1 CMD had higher trabecular vBMD (Mean [SD], 154 (39) vs 147 (37) mg/cm3; P < .001) but lower cortical vBMD (322 [47] vs 334 [43] mg/cm3; P < .001) (Table 2). In crude regression analyses, CMD status was significantly associated with higher trabecular vBMD and lower cortical vBMD (Table 4).

However, after adjusting for age, race, height, and weight (Model 1) and all confounders (Model 2), there was no association between CMD status or CMD burden and total strength, vBMD, and mass (Table 4). Again, results were similar when dyslipidemia was defined based on lipid values alone compared with lipid values and self-reported medication use (data not shown).

Table 4

The association between cardiometabolic disease and vertebral bone strength, mineral density, and mass in women

CrudeModel 1aModel 2b
β (SE) P value β (SE) P value β (SE) P value
Total strength, newtons
 CMD status, yes vs no206.36 (119.64).085−100.66 (130.94).442−65.22 (135.68).631
 CMD burden
  0Ref.Ref.Ref.
  1 355.43 (145.64).01528.89 (140.39).83754.23 (145.00).708
  252.70 (151.84).729 −382.78 (162.24).018−322.69 (167.65).054
  3+184.22 (156.14).238−112.76 (177.11).524−83.49 (186.49).654
 Linear trend.533.196.292
Total vBMD, mg/cm3
 CMD status, yes vs no−0.09 (2.23).968−1.71 (2.40).476−1.16 (2.48).641
 CMD burden
  0Ref.Ref.Ref.
  13.92 (2.72).1490.38 (2.57).8810.76 (2.65).775
  2−2.95 (2.83).298 −6.43 (2.97).030−5.30 (3.06).084
  3+−2.09 (2.91).472−1.54 (3.24).636−1.39 (3.41).683
 Linear trend.178.285.347
Trabecular vBMD, mg/cm3
 CMD status (yes vs no) 7.33 (2.02) <.001−1.45 (2.11).491−1.02 (2.18).641
 CMD burden
  0Ref.Ref.Ref.
  1 6.99 (2.46).004−0.63 (2.26).781−0.43 (2.34).853
  2 6.13 (2.56).017−4.30 (2.61).100−3.18 (2.70).239
  3+ 9.08 (2.64).0010.75 (2.85).7941.24 (3.01).679
 Linear trend.001.907.913
Cortical vBMD, mg/cm3
 CMD status (yes vs no) −11.51 (2.42) <.001−1.43 (2.70).596−1.03 (2.79).711
 CMD burden
  0Ref.Ref.Ref.
  1−3.04 (2.93).2981.07 (2.90).7121.19 (2.99).690
  2 −15.23 (3.05) <.001−6.24 (3.35).063−5.18 (3.45).134
  3+ −18.26 (3.14) <.001−3.05 (3.66).405−3.07 (3.84).425
 Linear trend <.001.164.204
Total mass, g
 CMD status, yes vs no−0.08 (0.08).312−0.07 (0.08).388−0.06 (0.09).513
 CMD burden
  0Ref.Ref.Ref.
  10.07 (0.10).4840.01 (0.09).9160.02 (0.09).821
  2−0.17 (0.10).098 −0.22 (0.10).035−0.20 (0.11).063
  3+−0.17 (0.10).091−0.15 (0.11).197−0.13 (0.12).264
Linear trend.023.068.104

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density.

a Model 1 = age, race, height, weight.

b Model 2 = Model 1 + education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.

Effect modification

There were many significant interactions between CMD and age in men (Table 5). CMD status was associated with lower total, trabecular, and cortical vBMD and lower total mass in younger men ≤50 years, whereas there was no significant association between CMD and the BCT measures in men >50 years of age (Fig. 2A-D). Similarly, greater CMD burden was associated with lower total (Fig. 2A) and trabecular vBMD (Fig. 2B) in younger men, whereas there was no association between CMD burden and BCT variables in older men.

There were no significant interactions between CMD and race in both men and women, nor between age, menopausal status, or hormone therapy use in women (Table 5).

Table 5

Race, age, menopausal status, and hormone therapy use interaction term significance level

MenWomen
BinaryCategoricalBinaryCategorical
P valuea P valueb P valuea P valueb P valuea P valueb P valuea P valueb
Race
 Total strength, newtons.529.933.255.212.644.777.984.935
 Total vBMD, mg/cm3.814.483.338.193.759.838.941.940
 Trabecular vBMD, mg/cm3.638.187.464.168.866.713.994.918
 Cortical vBMD, mg/cm3.894.533.380.256.537.489.969.923
 Total mass, g.564.999.189.165.471.483.673.667
Age
 Total strength, newtons.034.058.044.114.614.959.788.867
 Total vBMD, mg/cm3.011.019.013.038.414.666.839.955
 Trabecular vBMD, mg/cm3.007.010.008.014.705.954.962.980
 Cortical vBMD, mg/cm3.013.020.041.089.175.248.426.592
 Total mass, g.015.038.041.116.443.650.563.736
Menopausal status
 Total strength, newtons Not reported Not reported Not reported Not reported.685.944.719.832
 Total vBMD, mg/cm3 Not reported Not reported Not reported Not reported.193.409.373.694
 Trabecular vBMD, mg/cm3 Not reported Not reported Not reported Not reported.473.818.754.965
 Cortical vBMD, mg/cm3 Not reported Not reported Not reported Not reported.093.185.171.387
 Total mass, g Not reported Not reported Not reported Not reported.822.930.610.694
Hormone therapy use
 Total strength, newtons Not reported Not reported Not reported Not reported .737.833.985.968
 Total vBMD, mg/cm3 Not reported Not reported Not reported Not reported .876.785.971.996
 Trabecular vBMD, mg/cm3 Not reported Not reported Not reported Not reported .435.490.872.894
 Cortical vBMD, mg/cm3 Not reported Not reported Not reported Not reported .549.450.900.908
 Total mass, g Not reported Not reported Not reported Not reported .880.933.995.961

Bold indicates P < .05.

Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density.

a Biomechinal computed tomograpy= CMD adjusted for age, race, height, and weight.

b Biomechinal computed tomograpy = CMD adjusted for age, race, height, weight, education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.

Figure 2

The association between cardiometabolic disease and vertebral bone mineral density stratified by age group in men. (A) Total vBMD; (B) trabecular vBMD; (C) cortical vBMD; and (D) total mass. CMD status assessed as a binary variable and categorical burden variable. Abbreviations: CMD, cardiometabolic disease; vBMD, volumetric bone mineral density. aModels adjusted for race, height, weight, education status, family income, smoking, alcohol, physical activity, calcium intake, vitamin D intake, chronic obstructive pulmonary disease, cancer, and antidepressants.bInteraction term P values: Age × Total vBMD = .019; Age × Trabecular vBMD = .010; Age × Cortical vBMD = .020; Age × Total Mass = .038.

Figure image at the original source: open the original paper. This file is not included in this text import.

Discussion

In this cohort of middle-aged, Black and White adults, we found that having at least 1 of the CMDs of interest was associated with lower total and cortical vertebral vBMD in men, whereas there was no relationship between any CMD and any BCT variable in women. Similarly, greater CMD burden was associated with lower vBMD in men, with the strongest associations seen in men with ≥3 CMD conditions. The inverse relationship between CMD (status and burden) and vBMD was statistically significant in younger (≤50 years) compared with no association in older (>50 years) men.

Race did not modify the relationship between CMD and bone strength in either men or women. Age, menopausal status, and hormone therapy use did not modify the association between CMD and vertebral bone strength, density, or mass in women.

To our knowledge, this is the first study evaluating the association between CMD and BCT bone properties other than DXA-based BMD. Previous studies evaluating the relationship of CMD expressed as metabolic syndrome or as the individual conditions used trabecular bone score (TBS) or high-resolution peripheral quantitative CT (HRpQCT) to measure bone structural properties. TBS quantifies gray-scale variations in the DXA image reflecting bone texture, providing an indirect assessment of bone microarchitecture overall (65-67). Earlier studies showed lower TBS (ie, worse microarchitecture) in those with metabolic syndrome or T2DM despite higher DXA BMD (68).

However, more recent studies showed that this association was driven by soft tissue and not directly attributed to T2DM (33, 69-71). HRpQCT directly measures microarchitecture in trabecular and cortical bone at the radius and/or tibia instead of central sites (femoral neck and lumbar spine) used in DXA. Most of the HRpQCT studies are limited to T2DM, which have consistently reported no difference or sometimes greater trabecular BMD and microarchitecture (eg, trabecular number and thickness) in those with compared to those without T2DM (33).

The relationship between T2DM and cortical bone has been mixed, with some showing greater cortical porosity, especially in those with microvascular disease (72, 73), whereas others reported no significant difference in cortical parameters at the tibia (74). Our findings in men only confirm the greater body of HRpQCT evidence showing an adverse relationship with cortical bone. The differences in findings could be explained by our composite CMD variable instead of focusing on T2DM specifically.

There is limited information in population-based studies regarding bone strength and the other conditions included in our CMD exposure (75, 76). One small study on dyslipidemia and bone strength evaluated the relationship between atherogenic index of plasma (a ratio of triglycerides to HDL-C) and TBS and reported higher atherogenic index of plasma was significantly associated with 60% higher odds of low TBS (< 1.230), independent of age, body mass index, smoking, previous fracture, kidney function, and lumbar osteoporosis assessed via DXA in postmenopausal women (77).

On the cardiovascular disease side, a previous study in older adults from the United Kingdom reported that ischemic heart disease was associated with lower cortical vBMD and higher cortical porosity at the distal radius, and similar to our study, this association between ischemic heart disease and cortical vBMD was only significant in men (78).

Numerous studies have evaluated the possible mechanisms through which CMD is associated with bone, including dyslipidemia impacting osteocyte function and increasing bone resorption (79); hyperglycemia and visceral obesity impacting osteoblast function, disrupting bone formation (80); advanced glycation end-products reducing bone elasticity (81); and chronic inflammation associated with most of these conditions enhancing osteoclast activity (80, 82, 83). This lower formation and higher resorption can lead to low bone strength, density, and mass.

These mechanisms of action ideally should be the same in men and women, but we only found a significant association in men. The men in our study had a higher prevalence of dyslipidemia, so it is possible that factors related to dyslipidemia are driving these sex differences. Similarly, the women had a higher prevalence of abdominal obesity. Adipose tissue and estrogen have a complex relationship, with both positive and negative effects on bone (84), so a deeper exploration into these pathways could further elucidate information on the relationship between CMD and vertebral strength properties in women.

Lastly, more research is needed to explore the role sex hormones play in the relationship between CMD and bone health.

This study has many strengths, such as the inclusion of novel vertebral bone measures in Black and White men and women from a community-based cohort study. Given CARDIA’s standardized protocols and extensive data collection, our CMD exposure is highly valid. Likewise, the CARDIA imaging protocols yielded high-quality CT scans, minimizing missingness in the BCT data. The use of CT-based BMD measures reduces the high variability and limited reproducibility of BMD measurements associated with fat layering seen in DXA (85).

Despite these strengths, our study is not without limitations, including its cross-sectional design and limited age range of participants. The clinical definitions for the CMDs have changed throughout the course of the CARDIA study. We used the most current definitions to be consistent with today’s clinical practice, even if definitions were different at year 25 (2010-2011). CARDIA did not measure biomarkers specific to type 1 diabetes mellitus, which affects bone differently than T2DM. Misclassification of diabetes type is possible, but T2DM represents 90% to 95% of people with DM (86), so we anticipated minimal misclassification.

Our abdominal obesity definition was based on waist circumference, which is correlated with body weight. However, by retaining weight as a covariate, we were able to explicitly evaluate abdominal obesity and not general weight-based obesity. We also did not adjust for antiobesity medications. The first-generation antiobesity medications (eg, orlistat and phentermine) available in the study period (2010-2011) did not advance expected weight loss much above the 3% to 8% observed from the intensive lifestyle interventions in the Diabetes Prevention Program (87).

There are also no data to show that these older medications are associated with improved glucose control, insulin resistance, or reduced cardiovascular disease, thus limiting the effect of these medications on bone. Our CMD exposure was a composite of 4 variables, of which 3 had multiple avenues in which to meet the definition, leading to heterogeneity within each condition. Our future analyses will parse out this heterogeneity, along with identifying which condition or group of conditions are driving the associations observed.

Outside of height and weight, all potential confounders were self-reported. Self-reported data are prone to information bias, which could lead to misclassification and increase the type 1 error rate. Our nutrition data were carried forward from year 20. As previously mentioned, dietary habits in adults are relatively habitual. Those newly diagnosed with 1 of the CMDs between years 20 and 25 may have undergone dietary changes. Dietary changes for cardiovascular and metabolic health involve reducing refined carbohydrates and increasing fruit and vegetable consumption, so there could have been an increase in calcium and vitamin D intake, which would have a benefit on bone.

Given the potential for incorrect nutrition data, our observed findings could be an over: or underestimate of the true relationship between CMD and bone outcomes. Our study included 5 outcomes and 2 exposures, with additional effect modification analyses, potentially warranting adjustment for multiple comparisons to minimize the family-wise error rate (88). However, our 5 outcomes were not all independent of each other. Vertebral strength is highly correlated with density and moderately associated with size. Likewise, total vBMD is essentially the combination of trabecular and cortical vBMD.

Given the correlations between our outcomes, we did not formally adjust for multiple comparisons. Lastly, though we evaluated and adjusted for many potential confounders, as with any epidemiologic study, residual confounding could remain.

Conclusion

In this diverse cohort of US adults, CMD was significantly associated with lower overall and subcompartment vertebral vBMD in men only. Our findings can be used by endocrinologists and other clinicians managing patients with CMD, particularly in younger men, to support monitoring bone health (eg, bone turnover markers, CT imaging) in this population. This cross-sectional study served as the foundation for a future longitudinal study evaluating how CMD is associated with changes in bone health and vertebral fractures. Future analyses evaluating the causal mechanisms to our observed findings could lead to the addition of CMD or related biomarkers in fracture risk prediction models.

Acknowledgments

This manuscript has been reviewed by CARDIA for scientific content and consistency of data interpretation with previous CARDIA publications.

Contributor Information

Nicole C Wright, Email: nwright6@tulane.edu, Center for Health Outcomes, Implementation, and Community-Engaged Sciences, Tulane University, New Orleans, LA 70112, USA.

Rafeka Hossain, Center for Health Outcomes, Implementation, and Community-Engaged Sciences, Tulane University, New Orleans, LA 70112, USA.

Mary L Bouxsein, Center for Advanced Orthopaedic Studies, Beth Israel Deaconess Medical Center, Harvard University, Boston, MA 02215, USA.

Tony M Keaveny, O.N. Diagnostics, Berkeley, CA 94704, USA.

David C Lee, O.N. Diagnostics, Berkeley, CA 94704, USA.

Gerhard Hellemann, Department of Biostatistics, School of Public Health, University of Alabama at Birmingham, Birmingham, AL 35233, USA.

John Jeffrey Carr, Department of Radiology, Vanderbilt University Medical Center, Nashville, TN 37232, USA.

James G Terry, Department of Radiology, Vanderbilt University Medical Center, Nashville, TN 37232, USA.

Amy H Warriner, Division of Endocrinology, Diabetes, and Metabolism, School of Medicine, University of Alabama at Birmingham, Birmingham, AL 35233, USA.

Cora E Lewis, Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL 35233, USA.

Funding

This work was funded by the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) R01AR080868. The Coronary Artery Risk Development in Young Adults Study (CARDIA) is supported by contracts 75N92023D00002, 75N92023D00003, 75N92023D00004, 75N92023D00005, and 75N92023D00006 from the National Heart, Lung, and Blood Institute (NHLBI).

Disclosures

N.C.W., board of trustees, Bone Health & Osteoporosis Foundation. T.M.K., consultant, Amgen, Bone Health Technologies; employee, O.N. Diagnostics; equity, O.N., Diagnostics. D.C.L., employee, O.N. Diagnostics; equity, O.N. Diagnostics.

Data availability

CARDIA data are available upon reasonable request from the CARDIA Coordinating Center. CARDIA investigators are eager to collaborate with investigators interested in using CARDIA data. Please see the CARDIA website (https://sites.uab.edu/cardia/) for publication policies and for a list of CARDIA Representatives. CARDIA data are also publicly available on the NIH-supported BioLINCC and dbGaP platforms.

References

1. Eroglu  T, Capone  F, Schiattarella  GG. The evolving landscape of cardiometabolic diseases. EBioMedicine. 2024;109:105447.
2. Martin  SS, Aday  AW, Allen  NB, et al.  Heart disease and stroke statistics: a report of US and global data from the American Heart Association. Circulation. 2025;151(8):e41‐e660.
3. Burge  R, Dawson-Hughes  B, Solomon  DH, Wong  JB, King  A, Tosteson  A. Incidence and economic burden of osteoporosis-related fractures in the United States, 2005-2025. J Bone Miner Res. 2007;22(3):465‐475.
4. Johnell  O, Kanis  JA. An estimate of the worldwide prevalence and disability associated with osteoporotic fractures. Osteoporos Int. 2006;17(12):1726‐1733.
5. Shen  Y, Huang  X, Wu  J, et 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.
6. Hansen  DP, Pelizzari  P, Pyenson, B. Medicare costs of osteoporotic fractures: 2021 updated report. March 2021. Accessed October 7, 2025. https://www.milliman.com/en/insight/medicare-cost-of-osteoporotic-fractures-2021-updated-report
7. Moayyeri  A, Warden  J, Han  S, et al.  Estimating the economic burden of osteoporotic fractures in a multinational study: a real-world data perspective. Osteoporos Int. 2023;34(12):2121‐2132.
8. Cooper  C, Atkinson  EJ, Jacobsen  SJ, O’Fallon  WM, Melton  LJ  3rd. Population-based study of survival after osteoporotic fractures. Am J Epidemiol. 1993;137(9):1001‐1005.
9. Wright  NC, Chen  L, Saag  KG, Brown  CJ, Shikany  JM, Curtis  JR. Racial disparities exist in outcomes after Major fragility fractures. J Am Geriatr Soc. 2020;68(8):1803‐1810.
10. Lems  WF, Paccou  J, Zhang  J, et al.  Vertebral fracture: epidemiology, impact and use of DXA vertebral fracture assessment in fracture liaison services. Osteoporos Int. 2021;32(3):399‐411.
11. Bai  J, Gao  Q, Wang  C, Dai  J. Diabetes mellitus and risk of low-energy fracture: a meta-analysis. Aging Clin Exp Res. 2019;32(11):2173‐2186.
12. Janghorbani  M, Van Dam  RM, Willett  WC, Hu  FB. Systematic review of type 1 and type 2 diabetes mellitus and risk of fracture. Am J Epidemiol. 2007;166(5):495‐505.
13. Koromani  F, Ghatan  S, van Hoek  M, et al.  Type 2 diabetes mellitus and vertebral fracture risk. Curr Osteoporos Rep. 2021;19(1):50‐57.
14. Koromani  F, Oei  L, Shevroja  E, et al.  Vertebral fractures in individuals with type 2 diabetes: more than skeletal complications alone. Diabetes Care. 2020;43(1):137‐144.
15. Schwartz  AV, Sellmeyer  DE, Ensrud  KE, et al.  Older women with diabetes have an increased risk of fracture: a prospective study. J Clin Endocrinol Metab. 2001;86(1):32‐38.
16. Viegas  M, Costa  C, Lopes  A, Griz  L, Medeiro  MA, Bandeira  F. Prevalence of osteoporosis and vertebral fractures in postmenopausal women with type 2 diabetes mellitus and their relationship with duration of the disease and chronic complications. J Diabetes Complications. 2011;25(4):216‐221.
17. Zahedi  H, Atayie  F, Samii Kondrud  F, et al.  Associations of abdominal obesity with different types of bone fractures in adults: a systematic review and dose-response meta-analysis of prospective cohort studies. Crit Rev Food Sci Nutr. 2024;64(18):6239‐6250.
18. Ofir  O, Buch  A, Rouach  V, Goldsmith  R, Stern  N, Monsonego-Ornan  E. Association between abdominal obesity and fragility fractures among elderly Israeli women. Aging Clin Exp Res. 2020;32(8):1459‐1467.
19. Lin  YH, Teng  MMH. Central obesity is more strongly associated with vertebral fractures than general obesity: a cross-sectional study. BMC Geriatr. 2025;25(1):522.
20. Qi  W, Yu  W, Pang  Q, et al.  Associations between abdominal obesity and the prevalence of fractures among Chinese adults: insights from a nationwide cross-sectional study. Front Endocrinol (Lausanne). 2026;17:1759573.
21. Inose  H, Tanaka  S, Mori  S, Hagino  H, Soen  S. Predictors of vertebral fractures in patients with severe osteoporosis: a sub-analysis of the Japanese Osteoporosis Intervention Trial-05 (JOINT-05). Arch Osteoporos. 2025;21(1):12.
22. Long  Y, Kang  W, Yao  T, Lin  X, Xu  W. Association between baseline lipid levels and re-fracture after percutaneous vertebroplasty for osteoporotic vertebral compression fractures: a retrospective cohort study. J Orthop Surg Res. 2026;21(1):344.
23. Feng  Y, Tang  Q, Li  S, et al.  Characterizing the lipid profile in patients with vertebral or hip fragility fractures: a hospital-based descriptive study. J Clin Med. 2025;14(19):7029.
24. Lee  JW, Kim  J, Kim  Y, Kim  YS, Kim  J, Kang  HT. Higher high-density lipoprotein cholesterol levels increased vertebral osteoporotic fracture, but reduced hip fracture in men based on the National Health Insurance Service-National Health Screening Cohort. Spine J. 2024;24(10):1929‐1938.
25. Wada  H, Hirano  F, Kuroda  T, Shiraki  M. Breast arterial calcification and hypertension associated with vertebral fracture. Geriatr Gerontol Int. 2012;12(2):330‐335.
26. Vestergaard  P, Rejnmark  L, Mosekilde  L. Hypertension is a risk factor for fractures. Calcif Tissue Int. 2009;84(2):103‐111.
27. Bazelier  MT, de Vries  F, Vestergaard  P, et al.  Risk of fracture with thiazolidinediones: an individual patient data meta-analysis. Front Endocrinol (Lausanne). 2013;4:11.
28. Kanazawa  I, Yamaguchi  T, Yamamoto  M, Sugimoto  T. Relationship between treatments with insulin and oral hypoglycemic agents versus the presence of vertebral fractures in type 2 diabetes mellitus. J Bone Miner Metab. 2010;28(5):554‐560.
29. Zhang  Y, Chen  Q, Liang  Y, et al.  Insulin use and fracture risk in patients with type 2 diabetes: a meta-analysis of 138,690 patients. Exp Ther Med. 2019;17(5):3957‐3964.
30. Lee  S, Yu  MH, Hong  N, et al.  Association of sodium-glucose cotransporter 2 inhibitor use with risk of osteoporotic fracture among older women: a nationwide, population-based cohort study. Diabetes Res Clin Pract. 2024;213:111712.
31. Zhou  Z, Jardine  M, Perkovic  V, et al.  Canagliflozin and fracture risk in individuals with type 2 diabetes: results from the CANVAS program. Diabetologia. 2019;62(10):1854‐1867.
32. Hargrove  JL, Golightly  YM, Pate  V, et al.  Initiation of antihypertensive monotherapy and incident fractures among Medicare beneficiaries. Inj Epidemiol. 2017;4(1):27.
33. Ferrari  S, Akesson  KE, Al-Daghri  N, et al.  Bone microstructure and TBS in diabetes: what have we learned? A narrative review. Osteoporos Int. 2025;36(7):1115‐1128.
34. Keaveny  TM. Biomechanical computed tomography-noninvasive bone strength analysis using clinical computed tomography scans. Ann N Y Acad Sci. 2010;1192(1):57‐65.
35. Allaire  BT, Lu  D, Johannesdottir  F, et al.  Prediction of incident vertebral fracture using CT-based finite element analysis. Osteoporos Int. 2019;30(2):323‐331.
36. Anderson  DE, Demissie  S, Allaire  BT, et al.  The associations between QCT-based vertebral bone measurements and prevalent vertebral fractures depend on the spinal locations of both bone measurement and fracture. Osteoporos Int. 2014;25(2):559‐566.
37. Johannesdottir  F, Allaire  B, Kopperdahl  DL, et al.  Bone density and strength from thoracic and lumbar CT scans both predict incident vertebral fractures independently of fracture location. Osteoporos Int. 2021;32(2):261‐269.
38. Kopperdahl  DL, Aspelund  T, Hoffmann  PF, et al.  Assessment of incident spine and hip fractures in women and men using finite element analysis of CT scans. J Bone Miner Res. 2014;29(3):570‐580.
39. Wang  X, Sanyal  A, Cawthon  PM, et al.  Prediction of new clinical vertebral fractures in elderly men using finite element analysis of CT scans. J Bone Miner Res. 2012;27(4):808‐816.
40. O’Hearn  M, Lauren  BN, Wong  JB, Kim  DD, Mozaffarian  D. Trends and disparities in cardiometabolic health among U.S. adults, 1999-2018. J Am Coll Cardiol. 2022;80(2):138‐151.
41. Chong  B, Jayabaskaran  J, Jauhari  SM, et al.  The global syndemic of modifiable cardiovascular risk factors projected from 2025 to 2050. J Am Coll Cardiol. 2025;86(3):165‐177.
42. Mohebi  R, Chen  C, Ibrahim  NE, et al.  Cardiovascular disease projections in the United States based on the 2020 census estimates. J Am Coll Cardiol. 2022;80(6):565‐578.
43. Friedman  GD, Cutter  GR, Donahue  RP, et al.  CARDIA: study design, recruitment, and some characteristics of the examined subjects. J Clin Epidemiol. 1988;41(11):1105‐1116.
44. Lloyd-Jones  DM, Lewis  CE, Schreiner  PJ, Shikany  JM, Sidney  S, Reis  JP. The Coronary Artery Risk Development in Young Adults (CARDIA) study: JACC focus seminar 8/8. J Am Coll Cardiol. 2021;78(3):260‐277.
45. Grundy  SM, Cleeman  JI, Daniels  SR, et al.  Diagnosis and management of the metabolic syndrome: an American Heart Association/National Heart, Lung, and Blood Institute Scientific Statement. Circulation. 2005;112(17):2735‐2752.
46. Derived Variables from CARDIA Data. Accessed February 18, 2022. https://sites.uab.edu/cardia/for-researchers/study-information/derived-variables-from-cardia-data/
47. Blaton  V. How is the metabolic syndrome related to the dyslipidemia?  EJIFCC. 2007;18(1):15‐22.
48. Whelton  PK, Carey  RM, Aronow  WS, et al.  2017 ACC/AHA/AAPA/ABC/ACPM/AGS/APhA/ASH/ASPC/NMA/PCNA guideline for the prevention, detection, evaluation, and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol. 2018;71(19):e127‐e248.
49. Writing Committee Members, Jones  DW, Ferdinand  KC, et al.  2025 AHA/ACC/AANP/AAPA/ABC/ACCP/ACPM/AGS/AMA/ASPC/NMA/PCNA/SGIM guideline for the prevention, detection, evaluation and management of high blood pressure in adults: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. Hypertension. 2025;82(10):e212‐e316.
50. Carr  JJ, Jacobs  DR  Jr, Terry  JG, et al.  Association of coronary artery calcium in adults aged 32 to 46 years with incident coronary heart disease and death. JAMA Cardiol. 2017;2(4):391‐399.
51. Terry  JG, Shay  CM, Schreiner  PJ, et al.  Intermuscular adipose tissue and subclinical coronary artery calcification in midlife: the CARDIA study (Coronary Artery Risk Development in Young Adults). Arterioscler Thromb Vasc Biol. 2017;37(12):2370‐2378.
52. VanWagner  LB, Ning  H, Lewis  CE, et al.  Associations between nonalcoholic fatty liver disease and subclinical atherosclerosis in middle-aged adults: the Coronary Artery Risk Development in Young Adults Study. Atherosclerosis. 2014;235(2):599‐605.
53. Keaveny  TM, Crittenden  DB, Bolognese  MA, et al.  Greater gains in spine and hip strength for romosozumab compared with teriparatide in postmenopausal women with low bone mass. J Bone Miner Res. 2017;32(9):1956‐1962.
54. Keaveny  TM, Donley  DW, Hoffmann  PF, Mitlak  BH, Glass  EV, San Martin  JA. Effects of teriparatide and alendronate on vertebral strength as assessed by finite element modeling of QCT scans in women with osteoporosis. J Bone Miner Res. 2007;22(1):149‐157.
55. Keaveny  TM, McClung  MR, Genant  HK, et al.  Femoral and vertebral strength improvements in postmenopausal women with osteoporosis treated with denosumab. J Bone Miner Res. 2014;29(1):158‐165.
56. Orwoll  ES, Marshall  LM, Nielson  CM, et al.  Finite element analysis of the proximal femur and hip fracture risk in older men. J Bone Miner Res. 2009;24(3):475‐483.
57. Lee  DC, Hoffmann  PF, Kopperdahl  DL, Keaveny  TM. Phantomless calibration of CT scans for measurement of BMD and bone strength-inter-operator reanalysis precision. Bone. 2017;103:325‐333.
58. McDonald  A, Van Horn  L, Slattery  M, et al.  The CARDIA dietary history: development, implementation, and evaluation. J Am Diet Assoc. 1991;91(9):1104‐1112.
59. Whitelock  E, Ensaff  H. On your own: older adults’ food choice and dietary habits. Nutrients. 2018;10(4):413.
60. Sidney  S, Jacobs  DR  Jr, Haskell  WL, et al.  Comparison of two methods of assessing physical activity in the Coronary Artery Risk Development in Young Adults (CARDIA) study. Am J Epidemiol. 1991;133(12):1231‐1245.
61. Patsch  JM, Deutschmann  J, Pietschmann  P. Gender aspects of osteoporosis and bone strength. Wien Med Wochenschr. 2011;161(5-6):117‐123.
62. Zhang  YY, Xie  N, Sun  XD, et al.  Insights and implications of sexual dimorphism in osteoporosis. Bone Res. 2024;12(1):8.
63. Christiansen  BA, Kopperdahl  DL, Kiel  DP, Keaveny  TM, Bouxsein  ML. Mechanical contributions of the cortical and trabecular compartments contribute to differences in age-related changes in vertebral body strength in men and women assessed by QCT-based finite element analysis. J Bone Miner Res. 2011;26(5):974‐983.
64. Guglielmi  G, van Kuijk  C, Li  J, Meta  MD, Scillitani  A, Lang  TF. Influence of anthropometric parameters and bone size on bone mineral density using volumetric quantitative computed tomography and dual X-ray absorptiometry at the hip. Acta Radiol. 2006;47(6):574‐580.
65. Silva  BC, Leslie  WD, Resch  H, et al.  Trabecular bone score: a noninvasive analytical method based upon the DXA image. J Bone Miner Res. 2014;29(3):518‐530.
66. Muschitz  C, Kocijan  R, Haschka  J, et al.  TBS reflects trabecular microarchitecture in premenopausal women and men with idiopathic osteoporosis and low-traumatic fractures. Bone. 2015;79:259‐266.
67. Pothuaud  L, Carceller  P, Hans  D. Correlations between grey-level variations in 2D projection images (TBS) and 3D microarchitecture: applications in the study of human trabecular bone microarchitecture. Bone. 2008;42(4):775‐787.
68. Ho-Pham  LT, Nguyen  TV. Association between trabecular bone score and type 2 diabetes: a quantitative update of evidence. Osteoporos Int. 2019;30(10):2079‐2085.
69. Bagherzadeh  M, Sajjadi-Jazi  SM, Sharifi  F, et al.  Effects of metabolic syndrome on bone health in older adults: the Bushehr Elderly Health (BEH) program. Osteoporos Int. 2020;31(10):1975‐1984.
70. Leslie  WD, Binkley  N, Schousboe  JT, Silva  BC, Hans  D. Effect of abdominal tissue thickness on trabecular bone score and fracture risk in adults with diabetes: the Manitoba BMD registry. J Bone Miner Res. 2024;39(7):877‐884.
71. Totaro  M, Barchetta  I, Sentinelli  F, et al.  Waist circumference, among metabolic syndrome components, predicts degraded trabecular bone score: a retrospective study of a female population from the 2005-2008 NHANES cohorts. Front Endocrinol (Lausanne). 2024;15:1476751.
72. Shanbhogue  VV, Hansen  S, Frost  M, et al.  Compromised cortical bone compartment in type 2 diabetes mellitus patients with microvascular disease. Eur J Endocrinol. 2016;174(2):115‐124.
73. Shanbhogue  VV, Mitchell  DM, Rosen  CJ, Bouxsein  ML. Type 2 diabetes and the skeleton: new insights into sweet bones. Lancet Diabetes Endocrinol. 2016;4(2):159‐173.
74. Ballato  E, Deepika  FNU, Russo  V, et al.  One-year mean A1c of >7% is associated with poor bone microarchitecture and strength in men with type 2 diabetes mellitus. Calcif Tissue Int. 2022;111(3):267‐278.
75. Tomiyama  H. Vascular function: a key player in hypertension. Hypertens Res. 2023;46(9):2145‐2158.
76. Usiskin  IM, Mitchell  GF, Bouxsein  ML, Liu  CT, Kiel  DP, Samelson  EJ. Vascular function and skeletal fragility: a study of tonometry, brachial hemodynamics, and bone microarchitecture. J Bone Miner Res. 2024;39(7):906‐917.
77. Hernandez  JL, Olmos  JM, Pariente  E, Ramos  C, Martinez  J, Nan  D. The atherogenic index of plasma is related to a degraded bone microarchitecture assessed by the trabecular bone score in postmenopausal women: the Camargo Cohort Study. Maturitas. 2021;148:1‐6.
78. Paccou  J, Edwards  MH, Ward  KA, et al.  Ischemic heart disease is associated with lower cortical volumetric bone mineral density of distal radius. Osteoporos Int. 2015;26(7):1893‐1901.
79. Zhou  B, Li  S, Song  J. The impact of dyslipidemia on skeletal health: from an immunoregulatory perspective. Front Immunol. 2026;17:1774535.
80. Napoli  N, Chandran  M, Pierroz  DD, et al.  Mechanisms of diabetes mellitus-induced bone fragility. Nat Rev Endocrinol. 2017;13(4):208‐219.
81. Yamamoto  M, Sugimoto  T. Advanced glycation End products, diabetes, and bone strength. Curr Osteoporos Rep. 2016;14(6):320‐326.
82. Biamonte  E, Bendotti  G, Nigro  G, Cavigiolo  B, Gallo  M. Bone health in metabolic syndrome-is it a neglected aspect of dysmetabolic-related diseases?  J Clin Med. 2025;14(16):5785.
83. Pinar-Gutierrez  A, Garcia-Fontana  C, Garcia-Fontana  B, Munoz-Torres  M. Obesity and bone health: a Complex relationship. Int J Mol Sci. 2022;23(15):8303.
84. Kurylowicz  A. Estrogens in adipose tissue physiology and obesity-related dysfunction. Biomedicines. 2023;11(3):690.
85. Yu  EW, Thomas  BJ, Brown  JK, Finkelstein  JS. Simulated increases in body fat and errors in bone mineral density measurements by DXA and QCT. J Bone Miner Res. 2012;27(1):119‐124.
86. American Diabetes Association Professional Practice Committee for Diabetes . Introduction and Methodology: Standards of Care in Diabetes-2026. Diabetes Care. 2026;49(Suppl_1):S1‐S5.
87. Diabetes Prevention Program Research Group; Knowler  WC, Fowler  SE, et al.  10-year follow-up of diabetes incidence and weight loss in the Diabetes Prevention Program Outcomes Study. Lancet. 2009;374(9702):1677‐1686.
88. Hooper  R. To adjust, or not to adjust, for multiple comparisons. J Clin Epidemiol. 2025;180:111688.

Associated Data

Data Availability Statement

CARDIA data are available upon reasonable request from the CARDIA Coordinating Center. CARDIA investigators are eager to collaborate with investigators interested in using CARDIA data. Please see the CARDIA website (https://sites.uab.edu/cardia/) for publication policies and for a list of CARDIA Representatives. CARDIA data are also publicly available on the NIH-supported BioLINCC and dbGaP platforms.

Publication and source details

Original title: The relationship between cardiometabolic disease and vertebral bone strength, mineral density, and mass in US adults

Nicole C Wright; Rafeka Hossain; Mary L Bouxsein; Tony M Keaveny; David C Lee; Gerhard Hellemann; John Jeffrey Carr; James G Terry; Amy H Warriner; Cora E Lewis

Journal of the Endocrine Society · 2026

✉

Correspondence: Nicole C. Wright, PhD, MPH, Tulane University School of Medicine, 1440 Canal St., Tidewater 1503D, New Orleans, LA 70112, USA. Email: nwright6@tulane.edu.

© The Author(s) 2026. Published by Oxford University Press on behalf of the Endocrine Society.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. See the journal About page for additional terms.

Repository JATS snapshot; version as supplied

Source retrieved: 2026-11-10T21:10:30.686974+00:00

DOI: 10.1210/jendso/bvag193 · PMC13696833