1. Background
Celiac disease (CD) is an autoimmune disease characterized by inflammation in the intestine, causing atrophy of the mucosal villi of the small intestine (1). In the general population, the prevalence of CD is about 1% (2). Celiac disease can manifest with various signs and symptoms, including extra-intestinal symptoms such as osteopenia and osteoporosis (1).
Osteoporosis is a systemic skeletal disease characterized by low bone mass and subsequent increases in bone fragility and fracture susceptibility. Among bone metabolic diseases, osteoporosis is a common global public health problem with high morbidity. Many factors contribute to the development of osteoporosis, including genetic, environmental, and endocrine factors (3). Therefore, it is recommended that local healthcare programs be designed to reduce the prevalence of avoidable risk factors for osteoporosis and related fractures (4).
Osteopenia/osteoporosis are often associated with CD, and the nutritional, metabolic, and endocrine status of patients should also be considered. Awareness of the bone conditions associated with CD should be present. Therefore, there is an increasing trend for studies investigating the link between osteopenia/osteoporosis and CD (5). Dual-energy X-ray absorptiometry (DEXA) is a non-invasive method for assessing bone mineral density (BMD). Decreased BMD in CD may be due to compromised absorption of calcium and vitamin D (6). Several studies have shown that low BMD is a common complication of CD and recommend screening for osteoporosis at the time of diagnosis (7). On the other hand, it is unclear whether every new case of CD should undergo DEXA (8-10). Furthermore, evidence supporting the effect of independent variables on osteopenia/osteoporosis in CD is limited (11, 12).
2. Objectives
Our study aimed to determine the frequency of osteopenia/osteoporosis in newly diagnosed CD cases and to evaluate the effect of various independent variables on the development of osteopenia/osteoporosis in these patients.
3. Material and Methods
3.1. Ethical Approval/Statement
Approval of this study was obtained from the Ethics Committee of Shiraz University of Medical Sciences (IR.SUMS.MED.REC.1399.412), and the research adhered to the ethics declaration of Helsinki for medical research. Informed written consent was obtained from all patients for the review of their medical records.
3.2. Population
This cross-sectional analytical study took place from October 2017 to July 2022. The primary objective was to assess the frequency of osteopenia and osteoporosis and explore the odds of various factors on the likelihood of these bone disorders in patients with CD.
All adult participants undergoing CD evaluation at the Celiac Clinic (Fars Celiac Registry, Approval ID: IR.SUMS.REC.1399.525), a referral center in southern Iran, were considered for inclusion based on specific criteria. A checklist encompassing age, sex, height, weight, physical examination, personal and family medical history, histological reports, anti-tissue transglutaminase (anti-tTG) levels, vitamin D levels, and other laboratory data was completed. The level of 25-hydroxy vitamin D was measured in included CD patients using the Radio-Immuno-Assay method. Dual-energy X-ray absorptiometry was employed to determine BMD in patients meeting inclusion criteria based on specified characteristics. Subsequently, demographic, clinical, and para-clinical findings of patients with and without osteopenia/osteoporosis were compared.
Body Mass Index (BMI) was categorized according to the World Health Organization (WHO) classification (13), with four groups: (1) less than 18.5 kg/m2 as underweight, (2) 18.5 to < 25 kg/m2 as normal weight, (3) 25.0 to < 30 kg/m2 as overweight, and (4) 30.0 kg/m2 or higher as obese.
3.3. Celiac Disease Definition
Celiac disease diagnosis relied on a duodenal biopsy and a positive anti-tTG (14). Serum levels of anti-tTG (IgA) were measured in all enrolled patients by the ELISA method, and a titer of 18 IU/mL or higher was considered positive.
Upper gastrointestinal (GI) endoscopy and duodenal biopsies were conducted in participants with positive anti-tTG. Specimens were stained with hematoxylin/eosin and histologically classified according to the Oberhuber-modified Marsh classification (15). In this study, CD was defined as an anti-tTG ≥18 IU/mL and Marsh type 2 or higher in histology. All patients had an ordinary diet and consumed gluten, and the interval between duodenal biopsy and anti-tTG testing was less than one month.
3.4. Inclusion and Exclusion Criteria
All adult participants meeting the defined criteria for CD were included in the study. Exclusion criteria encompassed individuals under 18 years old, those who were lactating or pregnant, those with a history of alcohol consumption (more than 10 g/day), those using any medication that affects BMD or vitamin D levels within the past 6 months (e.g., calcium and vitamin D supplements), and those with co-morbid disorders (including malignancy, kidney, liver, lung, heart diseases, etc.). Additional exclusions included individuals with immunoglobulin A deficiency, Marsh type 0 or 1 in histology, other causes of villi atrophy, patients on a gluten-free diet, and non-cooperative patients. Celiac disease patients exhibiting abnormal BMD underwent evaluation by an endocrinologist to exclude secondary causes of decreased BMD (including thyroid, parathyroid, diabetes mellitus, etc.) as potential confounders.
3.5. Dual-Energy X-ray Absorptiometry
Under the supervision of an endocrinologist, BMD (g/cm2) was measured in three areas: The lumbar spine, hip, and femoral neck, using the Hologic Discovery DXA system (MA, Bedford, USA). Osteoporosis (T-score ≤ -2.5) and osteopenia (T-score between -1.0 and -2.5) were defined using the WHO classification (16). Coefficients of variation based on preliminary measurements in 10 participants using this system were 0.51% for the lumbar spine and 2.4% for the femoral neck.
3.6. Statistical Analysis
Data were collected using SPSS version 25 (SPSS Inc., Chicago, IL, USA). Continuous variables were presented as means ± standard deviations, while categorical variables were expressed as frequency/percentages. Group comparisons involved the independent t-test for continuous variables and the chi-square test for categorical variables. The Mann-Whitney U test was applied when the Kolmogorov-Smirnov test indicated significance for nonparametric continuous variables. A P-value < 0.05 was considered statistically significant. Logistic regression analysis, estimating odds ratios (ORs) and confidence intervals (CIs), was employed to assess the odds of various independent variables on osteopenia/osteoporosis. For regression analysis, a cut-off P-value of less than 0.3 in univariate analysis was used for inclusion in multiple analyses.
4. Results
A total of 302 patients satisfied the inclusion criteria. The participants had a mean ± SD age of 29.73 ± 12.39, ranging from 18 to 70 years. Among them, 194 (64.2%) were female, and 108 (35.8%) were male. The mean ± SD BMI of the participants was 21.19 ± 5.05, ranging from 14.68 to 38.77 kg/m2. The mean ± SD serum levels of anti-tTG and 25-hydroxy vitamin D were 220.29 ± 208.98 IU/mL and 28.15 ± 15.50 ng/mL, respectively. The demographic, clinical, and para-clinical features of the participants are summarized in Table 1.
| Variables | Values |
|---|---|
| Age (y) | |
| Mean ± SD | 29.73 ± 12.39 |
| Minimum-maximum | 18 - 70 |
| Gender | |
| Male | 108 (35.8) |
| Female | 194 (64.2) |
| BMI (kg/m2) | |
| Mean ± SD | 21.19 ± 5.05 |
| Minimum-maximum | 14.68 - 38.77 |
| Ethnicity | |
| Fars | 232 (76.8) |
| Others | 70 (23.2) |
| Education | |
| Diploma or less | 208 (68.9) |
| More than diploma | 94 (31.1) |
| Anti-tissue transglutaminase (IU/mL) | |
| Mean ± SD | 220.29 ± 208.98 |
| Median | 132.85 |
| GI manifestations | 206 (68.2) |
| Breast-feeding Milk type | 212 (70.2) |
| Marsh classification | |
| Marsh 2 | 14 (4.6) |
| Marsh 3a | 97 (32.1) |
| Marsh 3b | 115 (38.1) |
| Marsh 3c | 76 (25.2) |
| 25-hydroxy vitamin D (ng/mL) | |
| Mean ± SD | 28.15 ± 15.50 |
| Median | 25.25 |
| Minimum-maximum | 3.50 - 104.00 |
| Celiac disease in the family | 19 (6.3) |
| Familial marriage in the parents2 | 32 (10.6) |
Characteristics of the Patients with Celiac Disease (N = 302) a
The frequency of osteopenia and osteoporosis in the three regions of the lumbar spine, hip, and femoral neck is shown in Table 2. Overall, 215 (71.2%) patients had osteopenia or osteoporosis in at least one of these regions.
| Variables | No. (%) |
|---|---|
| Lumbar spine osteopenia | 127 (42.1) |
| Lumbar spine osteoporosis | 46 (15.2) |
| Femoral neck osteopenia | 130 (43.0) |
| Femoral neck osteoporosis | 48 (15.9) |
| Hip osteopenia | 120 (39.7) |
| Hip osteoporosis | 39 (12.9) |
Frequency of Osteopenia/Osteoporosis in Celiac Disease Patients (N = 302)
A comparison of demographic, clinical, and para-clinical findings between patients with and without osteopenia/osteoporosis is presented in Table 3.
The mean age of patients with osteopenia/osteoporosis (30.68 ± 13.12 years) was significantly higher than that of the normal group (27.37 ± 10.03 years). Although the frequency of osteopenia/osteoporosis was higher in women (74.2%) compared to men (65.7%), the gender difference did not reach statistical significance (P = 0.12). Despite a slightly elevated mean anti-tTG level in participants with osteopenia/osteoporosis (228.69 ± 217.32 IU/mL) compared to the normal group (199.53 ± 186.33 IU/mL), this difference was not statistically significant (P = 0.41). Notably, the mean 25-hydroxy vitamin D level in participants with osteopenia/osteoporosis was significantly lower than in the normal group (P < 0.001).
| Variables | Osteopenia/Osteoporosis | Normal | P-Value b |
|---|---|---|---|
| Age (y) c | 30.68 ± 13.12 | 27.37 ± 10.03 | 0.019 |
| Gender d | 0.119 | ||
| Male | 71 (65.7) | 37 (34.3) | |
| Female | 144 (74.2) | 50 (25.8) | |
| BMI (kg/m2) d | 0.054 | ||
| Under weight | 82 (79.6) | 21 (20.4) | |
| Normal weight | 101 (67.3) | 49 (32.7) | |
| Overweight | 18 (58.1) | 13 (41.9) | |
| Obese | 14 (77.8) | 4 (22.2) | |
| Ethnicity d | 0.340 | ||
| Fars | 162 (69.8) | 70 (30.2) | |
| Non-fars | 53 (75.7) | 17 (24.3) | |
| Education d | 0.598 | ||
| Diploma or less | 150 (72.1) | 58 (27.9) | |
| More than diploma | 65 (69.1) | 29 (30.9) | |
| GI manifestations d | 145 (70.4) | 61 (29.6) | 0.651 |
| Celiac disease in the family d | 10 (52.6) | 9 (47.4) | 0.065 |
| Familial marriage in the parents d | 21 (65.6) | 11 (34.4) | 0.462 |
| Marsh classification d | 0.234 | ||
| Marsh 2 | 10 (71.4) | 4 (28.6) | |
| Marsh 3a | 67 (69.1) | 30 (30.9) | |
| Marsh 3b | 77 (67.0) | 38 (33.0) | |
| Marsh 3c | 61 (80.3) | 15 (19.7) | |
| Anti-tissue transglutaminase (IU/mL); mean ± SD - Median e | 228.69 ± 217.32 - 142.40 | 199.53 ± 186.33 - 119.26 | 0.408 |
| 25-hydroxy vitamin D (ng/mL); mean ± SD - Median e | 25.81 ± 14.19 - 23.50 | 33.96 ± 17.08 - 32.00 | < 0.001 |
Demographic, Clinical, and Paraclinical Findings in Celiac Disease Patients with (N = 215) and Without (N = 87) Osteopenia/Osteoporosis a
A logistic regression analysis was conducted to assess the odds of different independent variables on osteopenia/osteoporosis, as detailed in Table 4. Participants aged over 30 exhibited a significantly higher likelihood of developing osteopenia/osteoporosis compared to those under 30 years old (OR: 2.19; 95% CI: 1.22 - 3.92; P = 0.008). The odds of developing osteopenia/osteoporosis were significantly elevated in underweight individuals (BMI <18.5 kg/m2, OR: 2.38; 95% CI: 1.30 - 4.34; P = 0.005) and participants with histologically severe atrophy (Marsh 3c, OR: 2.22; 95% CI: 1.14 - 4.32; P = 0.019) when compared to the normal group. Additionally, participants with normal vitamin D levels (25-hydroxy vitamin D level ≥ 30 ng/mL) had a significantly lower likelihood of developing osteopenia/osteoporosis than those with vitamin D deficiency (OR: 0.37; 95% CI: 0.21 - 0.62; P <0.001). Notably, other variables such as gender, anti-tTG levels, GI manifestation, and ethnicity did not significantly augment the OR of developing osteopenia/osteoporosis.
| Variables | Univariate Analysis | Multiple Analysis | ||
|---|---|---|---|---|
| OR (95% CI) | P-Value | OR (95% CI) | P-Value | |
| Age (y) | 0.015 | 0.008 | ||
| ≥ 30 | 1.963 (1.142 - 3.377) | 2.187 (1.221 - 3.916) | ||
| < 30 | 1.0 | 1.0 | ||
| BMI | 0.021 | 0.005 | ||
| < 18.5 (under weight) | 1.938 (1.104 - 3.402) | 2.376 (1.301 - 4.342) | ||
| ≥ 18.5 (others) | 1.0 | 1.0 | ||
| Gender | 0.120 | 0.075 | ||
| Male | 1.501 (0.900 - 2.503) | 1.656(0.951 - 2.884) | ||
| Female | 1.0 | 1.0 | ||
| Celiac disease in the family | 0.072 | 0.065 | ||
| Yes | 2.365 (0.926 - 6.040) | 2.571 (0.941 - 7.024)1.0 | ||
| No | 1.0 | |||
| Familial marriage in the parents | 0.463 | 0.415 | ||
| Yes | 1.337 (0.615 - 2.906) | 1.439 (0.600 - 3.452) | ||
| No | 1.0 | 1.0 | ||
| Ethnicity | 0.342 | 0.134 | ||
| Fars | 1.347 (0.729 - 2.489) | 1.670 (0.854 - 3.264) | ||
| Others | 1.0 | 1.0 | ||
| Education | 0.598 | 0.197 | ||
| Diploma or less | 0.867 (0.509 - 1.476) | 0.669 (0.363 - 1.233) | ||
| More than diploma | 1.0 | 1.0 | ||
| Milk type | 0.984 | 0.607 | ||
| Breastfeeding | 0.994 (0.577 - 1.714) | 1.173 (0.639 - 2.153) | ||
| Others | 1.0 | 1.0 | ||
| GImanifestations | 0.652 | 0.244 | ||
| Yes | 1.133 (0.660 - 1.944) | 1.426 (0.785 - 2.590) | ||
| No | 1.0 | 1.0 | ||
| Anti-tissue transglutaminase level | 0.999 (0.998 - 1.001) | 0.274 | 0.999 (0.998 - 1.000) | 0.179 |
| Histological findings | 0.046 | 0.019 | ||
| Marsh 3c | 1.901 (1.012 - 3.571) | 2.217 (1.139 - 4.316) | ||
| Others | 1.0 | 1.0 | ||
| 25-hydroxy vitamin D level | 0.001 | < 0.001 | ||
| Normal | 0.409 (0.246 - 0.680) | 0.365 (0.214 - 0.623) | ||
| Deficiency | 1.0 | 1.0 | ||
The Odds of Different Independent Variables on Developing Osteopenia/Osteoporosis in Patients with Celiac Disease Using Logistic Regression Analysis
5. Discussion
This study assessed bone mineral density in adult patients with CD. More than two-thirds of our patients had osteopenia or osteoporosis in at least one of their regions. In examining various variables within the subgroup analysis, older participants and underweight patients had a higher likelihood (OR) of developing osteopenia/osteoporosis. Additionally, we observed a high probability of osteopenia/osteoporosis in participants with severe histologic atrophy (Marsh 3C) compared to other histologic groups. Furthermore, patients with normal vitamin D levels were less likely (OR) to develop osteopenia/osteoporosis than those with vitamin D deficiency.
The frequency of metabolic bone disease in patients with CD is increased due to several mechanisms (17, 18). One of the possible mechanism of osteoporosis in CD is calcium malabsorption leading to secondary hyperparathyroidism, which contributes to cortical bone loss. Several studies have demonstrated low BMD in CD, attributed to high levels of serum cytokines that trigger osteoclasts (IL-2, IL-6, and TNF-alpha) and low levels of cytokines that play an inhibitory role (IL-18, IL-12) (19, 20).
In our study, despite the higher frequency of osteopenia/osteoporosis in women, there were no statistically significant differences between the genders. This finding supports previous data indicating that the frequency of osteopenia and osteoporosis in CD does not significantly differ between genders (21, 22).
In the present study, participants over 30 had higher odds of developing osteopenia/osteoporosis. This observation corroborates findings from other studies (23, 24) that suggest aging increases the incidence of low BMD in patients with CD. A study by Marcella D. Walker et al. found a high prevalence of osteoporosis among people over 50 years of age (7).
In our research, similar to some reports (25, 26), severe histological atrophy (Marsh 3c) increased the odds of osteoporosis in patients with CD. Walker et al. showed that the increase in the severity of intestinal villus atrophy was associated with the male gender and lower T-score in the 1/3 radius (7). These data support the hypothesis about the impact of intestinal damage on BMD. In CD, atrophy of the intestinal epithelium leads to a decreased surface area for calcium and vitamin D absorption. Hypocalcemia and low levels of vitamin D contribute to secondary hyperparathyroidism. Consequently, there is high bone turnover and bone loss due to increased levels of serum PTH (27, 28).
Galli et al. (29) showed that underweight patients had a higher risk of osteoporosis compared to normal-weight CD patients. This finding is consistent with the results of our study, which showed an increased OR of osteopenia at a BMI of less than 18.5 kg/m2. This may be due to malabsorption resulting in deficiencies in micronutrients such as calcium and vitamin D (1, 30).
Kemppainen et al. demonstrated that approximately two-thirds of patients with CD had low 25-hydroxy vitamin D concentrations (31). In their studies, vitamin D deficiency was one of the main variables associated with low BMD in CD, which is similar to our findings. 25-hydroxy vitamin D is the principal form of vitamin D in circulation, and its level in the blood is thought to reflect nutritional status. Reduced calcium intake secondary to vitamin D deficiency results in increased PTH levels, which induces cortical bone loss (32).
There is no significant association between anti-tTG antibody levels and the odds of developing osteopenia/osteoporosis in our study. This observation is inconsistent with the report by Potter et al., who described an association between high levels of anti-tTG antibodies and low BMD. On the other hand, they found that gender, GI manifestations, and ethnicity were not associated with osteopenia/osteoporosis in CD, which is consistent with our results (33).
The strengths of this study include a substantial sample size accompanied by strict diagnostic evaluations for all patients. Additionally, we had appropriate inclusion and exclusion criteria for participants, compared to similar studies.
We acknowledge several limitations to our study. This study was single-center and lacked a control group of the normal population. Except for vitamin D, other metabolic bone markers were not measured due to their high cost. We did not have enough data to analyze bone mineral content and Z-scores of our population, which is recommended for consideration in future studies, especially in individuals under 30 years old.
5.1. Conclusions
The present study showed that advancing age, low BMI, severe villous atrophy, and low levels of vitamin D can significantly increase the odds of developing osteopenia/osteoporosis in CD patients. Until further studies are conducted, BMD evaluation is especially recommended for these high-risk subgroups of CD patients.
