SEARCH

SEARCH BY CITATION

Keywords:

  • mandibular alveolar bone;
  • ovariectomy;
  • protein undernutrition;
  • bone histomorphometry;
  • bone densitometry;
  • mastication;
  • proximal tibia

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Mandibular and systemic bone loss are poorly associated. We compared the effect of isocaloric protein undernutrition and/or ovariectomy on BMD and microstructure of mandibular alveolar and proximal tibia sites in adult rats. Mandibular bone was significantly less affected.

Introduction: Whether mandibular bone and axial or peripheral skeleton respond similarly to systemic bone loss remains a subject of controversy. We have previously shown that mechanical loading during mastication influences bone mass and architecture of the mandibular alveolar bone. Isocaloric protein undernutrition and ovariectomy are known to cause bone loss and deterioration of bone microarchitecture at various axial and peripheral skeletal sites. We studied how the mandible, which is subjected to heavy, abrupt, and intermittent forces during mastication, responds to low-protein intake and/or ovariectomy and compared this response to that of the proximal tibia in adult rats.

Materials and Methods: Forty-four 6-month-old female Sprague-Dawley rats underwent transabdominal ovariectomy (OVX; n = 22) or sham operation (n = 22) and were pair-fed isocaloric diets containing either 15% or 2.5% casein (sham 15%, n = 11; sham 2.5%, n = 11; OVX 15%, n = 11; and OVX 2.5%, n = 11) for 16 weeks. BMD and bone microarchitecture parameters (e.g., bone volume fraction [BV/TV] and trabecular thickness and number) of the mandible and the proximal tibia were measured at the end of the experiment using DXA and μCT.

Results: Mandibular alveolar bone was negatively influenced by both protein undernutrition and OVX, but to a significantly lesser extent than the proximal tibia. In sham-operated animals, low-protein intake led to a 17.3% reduction of BV/TV in the mandible and 84.6% in the tibia (p < 0.001). In normal protein diet–fed animals, OVX led to a reduction of BV/TV of 4.9% in the mandible but 82% in the tibia (p < 0.001). In the mandible, protein undernutrition resulted in thinner trabeculae (p < 0.05), whereas OVX led to a reduction of trabecular number (p < 0.05).

Conclusions: Mandibular alveolar bone was found to be less sensitive to either protein undernutrition or OVX than the proximal tibia spongiosa. We hypothesize that the mechanical loading of the alveolar process during mastication may protect the alveolar bone from the detrimental effects observed in other skeletal sites, such as the proximal tibia. Morphological and embryological differences between the two skeletal sites might also play a role.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Alarge number of the elderly suffer from osteoporosis, periodontal bone loss, tooth loss, and malnutrition. Periodontal bone loss is an infection-mediated process characterized by resorption of the alveolar bone (the bone surrounding the teeth) and loss of the soft tissue attachment to the tooth and is a major cause of tooth loss in adults.(1) Dental status has been shown to have a negative impact on food choice and on the intake of key nutrients.(2) Malnutrition is highly prevalent among the elderly population for various physiological, pathological, and psychosocial reasons.(3) Among the various nutrients, protein undernutrition has been associated with decreased bone mass and strength in both humans(4) and animals.(5–7)

The precise relationship between osteoporosis, periodontal mandibular bone, and tooth loss is currently not well understood.(8–12) Although there is growing evidence of an association between systemic BMD and alveolar BMD or tooth loss, this association seems to be, at best, significantly inferior to that observed in other skeletal sites.(13–18) A number of animal studies on the effect of ovariectomy (OVX) on the mandibular bone have yielded conflicting results as well.(19–22)

The mandible is both morphologically and functionally different from the other bones of the axial or peripheral skeleton. It also arises from a different embryonic germ layer (neuroectoderm) compared with bone of the axial and appendicular skeleton, which arises from mesoderm. In a previous animal study, we showed that the mechanical loading of the mandible during mastication has an impact on the mass, density, and microarchitecture of the mandibular alveolar bone.(23) This may give a clue as to why the majority of studies failed to show a strong association between systemic and mandibular bone loss. Another hypothesis is that, although these studies focus on the effects of estrogen depletion on the mandible, nutritional aspects such as malnutrition may play an even more important role in mandible osteoporosis. To the best of our knowledge, this is the first study of the effect of protein undernutrition on mandibular bone.

The aim of this study was to investigate and compare the effect of two well-known causes of systemic bone loss, protein undernutrition and OVX and their interaction, on the mandible and the proximal tibia in adult rats. The outcome measures were based on microtomographic histomorphometry, densitometry, and biochemical determinations.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Animals and diet

Forty-four 6-month-old female Sprague-Dawley rats (Novartis, Basel, Switzerland) were housed individually with demineralized water available ad libitum. Their weight at the beginning of the experiment was 288 ± 7 g. Room temperature was maintained at 25°C with a 12:12-h light-dark cycle. The animals were strictly pair-fed with isocaloric synthetic diets (Novartis Nutrition, Bern, Switzerland) containing 15% or 2.5% casein, 0.8% phosphorus, 1.1% calcium, 0.2% magnesium, 79–80% carbohydrates, and 5% fat throughout the entire study period. Isocaloric intakes were ensured through the addition of corn carbohydrate to the low-protein diet. Animals were also given a daily dose of vitamin D dissolved in peanut oil (100 IU/kg body weight). The study was approved by the Animal Ethics Committee of the Faculty of Medicine of the University of Geneva.

Experimental design

After 2 weeks of adaptation to a 15% casein-containing diet, the animals underwent transabdominal OVX (n = 22) or a sham surgical operation (n = 22) under anesthesia with intraperitoneal ketamine hydrochloride (100 mg/kg body weight). Effectiveness of OVX was verified at the end of the experiment by visualizing the atrophy of the uterus and the absence of ovarian tissue. The animals were pair-fed isocaloric diets containing either 15% or 2.5% casein (sham-operated 15%, n = 11; sham-operated 2.5%, n = 11; OVX 15%, n = 11; and OVX 2.5%, n = 11) for 17 weeks.

In a previous study, it was shown that the minimal protein intake to maintain a normal bone turnover and mass in the rat was a 5% casein diet.(5) Thus, a 2.5% casein diet corresponds to a 50% reduction of normal protein intake, which is close to that observed in the malnourished elderly.(24) At the end of the study, blood was sampled to determine IGF-I and osteocalcin plasma levels, and urine was collected over 24 h for the determination of total deoxypyridinoline excretion. The animals were killed by an overdose of ketamine hydrochloride. The mandibles and left tibia were removed for microtomographic histomorphometry and densitometric evaluation. Mandibles were separated at the symphysis into their two halves, and the left hemimandibles were used in this study (Fig. 1A).

thumbnail image

Figure Figure 1. Rat mandible. (A) Photograph superimposed on a drawing of the skull (a, site under study that is called alveolar process, bears the molars, and is the equivalent of the alveolar process in humans; b, not to be confused with the incisor alveolar process that does not exist in the human mandible; and c, the condylar process which articulates with the skull), (B) DXA screen (dashed line represents ROI), and (C) μCT scan (dashed line represents VOI as it appears in one layer).

Download figure to PowerPoint

Biochemical determinations

Plasma osteocalcin and IGF-I were measured using radioimmunoassay with reagents from Biomedical Technologies (Stoughton, MA, USA) for the former and with a kit from Nichols Institute (San Juan Capistrano, CA, USA) after extraction by acid-ethanol and cryoprecipitation for the latter. Total urinary deoxypyridinoline was determined after acid hydrolysis using a kit from Metra-Biosystems (Mountain View, CA, USA).

Densitometry by DXA

Mandibles and tibias were scanned using DXA with a pencil-beam densitometer (QDR-1000; Hologic, Waltham, MA, USA). BMC (mg), scanned area (cm2), and BMD (mg/cm2) were evaluated. The instrument was set on “ultra-high resolution” mode (line spacing, 0.254 mm; resolution, 0.127 mm), and a smaller collimator (0.9 mm diameter) was placed over the original one.(23,25) Each hemimandible was placed in a standardized manner in a plastic box filled with 5 cm of normal saline. This has been shown to improve edge detection by acting as a soft-tissue equivalent.(26,27) An anatomically and geometrically defined region of interest (ROI) that included the alveolar process was chosen (Fig. 1B). Reproducibility was evaluated using the CV of repeated measurements with repositioning (CV was <1.4%).

Microtomographic histomorphometry by μCT

Microtomographic histomorphometry of the secondary spongiosa of the proximal tibia and the mandibular alveolar process was performed with a high-resolution μCT system (μCT 40; Scanco Medical AG, Bassersdorf, Switzerland) as previously described.(23,28) In summary, 3D images of each proximal tibia and left hemimandible were acquired with a voxel size (nominal resolution) of 16 μm in all spatial directions. Samples were secured in a cylindrical sample holder in air. Resolution was set to medium (500 projections with 1024 samples each). The trabecular and cortical parts of the tibia were separated with semiautomatically drawn contours.

For the mandible, the volume of interest (VOI) was drawn on a slice-based method starting from the first slice containing the crown of the first molar and moving dorsally 100 slices in the area of the alveolar process between the roots of the molars and the root of the incisor. Trabecular bone was carefully contoured on the first and the last slice, whereas the intermediate slices were first interpolated by morphing (Fig. 1C). Each slice was subsequently visually inspected, and the contour was modified where deemed necessary.

Microstructural indices were calculated directly from the binarized VOI. Total volume (TV) was the volume of the whole sample examined. Bone volume (BV) and surface (BS) were calculated using tetrahedrons corresponding to the enclosed volume of the triangulated surface. Mean trabecular thickness (Tb.Th) was determined from the local thickness at each voxel representing bone. Trabecular number (Tb.N) was calculated by taking the inverse of the mean distance between the middle axes of the structure. The complete acquiring procedure was repeated a further four times with four mandibles chosen at random to calculate the CV (range, 1.05–2.25% for Tb.Th and BV/TV).

Statistical analyses

All data are represented as mean ± SE. ANOVA was performed to detect any differences between groups. Multifactorial ANOVA was also used to study the main and interaction effects of the two experimental factors, OVX (OVX versus sham-operated) and low-casein diet (2.5% versus 15%), on the variables under study. The Bonferroni correction was used to perform posthoc comparisons between groups. All statistical analyses were performed using the SPSS statistical package (SPSS 13.0; SPSS, Chicago, IL, USA). A result was considered as statistically significant at p < 0.05.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

Body weight

Protein undernutrition (isocaloric diet containing 2.5% casein) led to a statistically significant (p < 0.001) weight reduction at the end of the experimental period for both sham and OVX animals (213 ± 15 g compared with 261 ± 12 g and 213 ± 13 g compared with 290 ± 10 g, respectively; Table 1). Among the 15% casein–fed animals, OVX was associated with significantly higher weight (290 ± 10 g compared with 261 ± 12 g, p = 0.001). There was no weight difference between sham and OVX animals in the protein undernourished group.

Table Table 1.. Body Weight and Biochemical Parameters at the End of the Experimental Period
Thumbnail image of

Biochemical parameters

Plasma IGF-I at death was found to be strongly correlated to the weight of the animals (r2 = 0.79; Table 1). IGF-I was significantly higher in the 15% casein–fed animals (p < 0.001). OVX was found to be associated with higher IGF-I values only in the 15% casein–fed group (p < 0.001). OVX animals that received normal food had higher plasma osteocalcin levels at death (p < 0.01). Total urinary deoxypyridinoline was significantly higher in the protein undernourished animals (p < 0.001).

BMD (DXA)

Proximal tibia BMD was lower in both OVX and protein undernourished animals (∼11% and 9%, respectively; p < 0.001; Table 2; Fig. 2). There was no interaction between OVX and protein undernutrition, although undernourished OVX animals suffered a greater reduction (∼17%). In the alveolar process, only protein undernutrition had a significant negative effect on BMD. It was ∼4% lower (p < 0.01) in comparison with control animals. In contrast to its negative effect on the BMD of proximal tibia, OVX did not seem to have influenced alveolar BMD. No interaction was observed.

Table Table 2.. Main Effects and Interaction of OVX and Protein Undernutrition on BMD and Trabecular Microarchitecture
Thumbnail image of
thumbnail image

Figure Figure 2. Effect of OVX and/or isocaloric protein undernutrition on BMD. ANOVA was followed by posthoc pairwise comparison between groups. Bars with the same superscript letter are significantly different (p < 0.05).

Download figure to PowerPoint

In all groups, BMD reduction was significantly higher (p < 0.01) for proximal tibia compared with mandibular alveolar bone. BMD of the two skeletal sites was significantly but weakly correlated (p < 0.001, r2 = 0.22).

Trabecular microarchitecture (microtomographic histomorphometry)

In cross-sectional slices from both proximal tibia and mandibular alveolar process (Fig. 3), OVX and protein undernutrition were observed to have detrimental effects on the proximal tibia spongiosa (Table 2; Fig 4). In the alveolar process, there seems also to be a negative effect, but the changes were of lower amplitude.

thumbnail image

Figure Figure 3. Transverse slices of the proximal tibia and the mandibular alveolar process of one animal per group, selected at random (OVX, ovariectomy instead of sham operation for the controls; protein undernutrition, isocaloric diet containing 2.5% casein, 15% for the controls).

Download figure to PowerPoint

thumbnail image

Figure Figure 4. Effect of OVX and/or isocaloric protein undernutrition on microtomographic histomorphometry parameters. ANOVA was followed by posthoc pairwise comparison between groups. Bars with the same superscript letter are significantly different (p < 0.05).

Download figure to PowerPoint

Protein undernutrition led to significant changes in both the proximal tibia and mandibular alveolar process. These changes were significantly greater in the case of the tibia. Bone volume fraction decreased by 76% in the proximal tibia but only 17% in the mandible (p < 0.001). Connectivity density decreased by 90% in the tibia, whereas it did not change in the mandible (p < 0.001). On the other hand, trabecular thickness decreased only in the mandible (12%, p < 0.001). Trabecular number decreased by 56% in the tibia but only 2% in the mandible (p < 0.001).

OVX also had a more significant impact on the proximal tibia than the mandible. BV/TV decreased by 82% in the proximal tibia but only 5.5% in the mandible (p < 0.001). Connectivity density decreased by 94.5% in the tibia but only 18.5% in the mandible (p < 0.001). On the other hand, trabecular thickness was not affected in both skeletal sites. Trabecular number decreased by 63% in the tibia but only 6% in the mandible (p < 0.001).

An additive effect of OVX and protein undernutrition was detected in the case of BV/TV and connectivity density (p < 0.001). Bone surface fraction (BS/BV) was not influenced by OVX or protein undernutrition, but it was found to be significantly higher for the proximal tibia than mandibular bone (p < 0.001). The correlation between the two skeletal sites was statistically significant only for BV/TV (p < 0.01), but this correlation was weak (r2 = 0.17).

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

The comparative study of the effect of OVX and/or protein undernutrition in adult rats revealed significant differences in the response of two different skeletal sites (i.e., the proximal tibia spongiosa and the mandibular alveolar bone). Although the negative effect of both experimental factors could be shown in the two sites, the mandibular alveolar bone was significantly less affected. The effects of these two factors were additive for the tibia but not for the mandible. Protein undernutrition seems to have had a greater impact on alveolar bone than estrogen depletion. To the best of our knowledge, this is the first time that the role of protein undernutrition on mandibular bone was clearly shown and compared with its effect on another skeletal site.

Low protein intake has been associated with decreased bone strength,(5,6) altered intrinsic bone tissue quality,(7) and detrimental effects on long bone fracture healing(29) in rats. This study revealed important differences of the effect of low protein intake on the trabecular bone of the mandible and proximal tibia, with a much lower impact on the former site. Trabecular number decreased in the tibia but not in the mandible. Thickness of the few trabeculae left in the proximal tibia spongiosa was not affected, in contrast to the observed reduction of trabecular thickness in the mandible.

It is well documented that OVX in rats decreases bone mass and alters the microarchitecture at various skeletal sites containing trabecular bone(27) and is the most commonly used animal model method of postmenopausal osteoporosis.(30) For mandibular bone, there are some studies that have shown a negative effect of OVX on the mandibular body,(31) condyle,(32,33) and alveolar process.(34,35) However, many studies have failed to detect any effect or concluded that the influence of OVX on the mandibular condyle(36–38) or the alveolar process(20,22) was significantly weaker in comparison with the long bones. Rat strain(39) and site of measurement might partially explain these discrepancies. Indeed, the alveolar bone seems to be less affected than the condyle or the mandibular body.

In an animal model studying the relationship between oral bone loss and estrogen deprivation, OVX did not aggravate induced periodontal bone loss, although it did have a clear negative effect on the femur.(40) Evidence from epidemiological studies shows that mandibular BMD is correlated to systemic BMD.(41–43) However, although there seems to be an association between periodontal bone or tooth loss and osteoporosis, this association is rather weak.(13–18) A possible explanation is that, although these studies focus on the effects of estrogen depletion on the mandible, nutritional aspects such as malnutrition, frequent among the elderly, may play an even more important role in the osteoporosis of the mandibular bone.

Another explanation for the different response of the mandibular alveolar bone to estrogen deprivation in comparison with other skeletal sites is the mechanical loading during mastication. The alveolar bone houses the teeth and is subjected to heavy loading during mastication. The teeth transmit this stress to the alveolar bone through the periodontal ligament, which acts as a shock absorber. This unique biomechanical configuration is entirely different from any other skeletal site of the body. Of note, rat molars erupt normally, exactly like human molars, in contrast to rat incisors, which erupt continuously.(44) In the rat, the forces developed during incision have been estimated to range from 2 to 25N (but mainly <8N), depending on the hardness of the food, but are of relatively short duration (40–160 ms). During chewing (and swallowing), however, forces are of lower magnitude (4N) but of a relatively long duration (160–680 ms).(45,46) During normal function, the mechanical loading of the tibia is inferior to that of the alveolar bone during mastication. For most vertebrates, long bone peak functional strains range from <1000 με during walking to between 2000 and 3200 με for more vigorous activities.(47) Finite element data suggest that alveolar bone strain during mastication can reach up to 4000–6000 με depending on food consistency.(48,49)

There is strong evidence that bone formation is influenced by strain rate, frequency, amplitude, duration, and interpolation of rest periods.(47) It is known, for example, that bone cells respond better to a mechanical environment dominated by high strains changing at fast rates and presented in unusual distributions.(50,51) During each chewing session that lasts a few minutes, alveolar bone strain goes from almost 0 to 4000 με and back again to almost 0 in 160–680 ms. When the food is available ad libitum, eating frequency has been estimated to be ∼30 times per day in the rat.(52)

The correlation between mandibular alveolar trabecular structure, BMD, and mechanical loading has been shown in a recent study where masticatory loading was changed through the alteration of food consistency.(23) It was shown that a reduction of the masticatory loading of the alveolar bone (soft food) results in a reduction of BMD, trabecular bone volume, and thickness, as well as width reduction of the alveolar process itself. It seems that alveolar bone macro- and microarchitecture is directly influenced by its mechanical loading during mastication. In this study, the mandibular alveolar bone seems to resist more than the tibia spongiosa to both protein and estrogen deprivation, possibly because of the anabolic effect of bone strain during mastication. The protective effect of exercise during estrogen deprivation has been shown in a number of previous studies.(53) This hypothesis is supported by the fact that the mandibular condyle or body, which undergoes less strain during mastication, are more affected than the alveolar bone during estrogen deficiency.

In this study, trabecular bone of the alveolar process was approximately three times thicker than the proximal tibia in controls. In the proximal tibia, low protein intake or OVX resulted in an almost total disappearance of trabeculae. In mandibular alveolar bone, low protein intake did not decrease trabecular number but only thickness, and OVX resulted in a small reduction of trabecular number. A plausible hypothesis is that, because of higher trabecular thickness, a higher degree of bone loss should be reached before thinning of the trabeculae eventually results in structural deterioration of the mandibular alveolar bone. Bone surface fraction (BS/BV) was significantly higher for proximal tibia than for mandibular alveolar bone. This means that the bone surface available for remodeling was much higher for proximal tibia, and this provides another possible explanation for the different responsiveness of the two skeletal sites to bone loss stimuli. This experiment had a duration of 17 weeks. It is possible that a longer exposure to OVX or/and low protein intake might have a bigger impact on the alveolar bone. In humans, trabecular bone of the alveolar process is approximately two times thicker than that of the proximal tibia.(54,55) Whether this is a physiological adaptation to high bone strain during mastication or a site-specific particularity is unknown. The mandibular bone arises from a different embryonic germ layer (neuroectoderm) compared with bone of the axial and appendicular skeleton, which arise from mesoderm.

In this study, we showed that estrogen deprivation and protein undernutrition had a negative effect on the mandibular alveolar bone, but this effect was significantly lower than on the proximal tibia of the same animals. In contrast to OVX, protein undernutrition led to a reduction of IGF-I. Decreased bone formation and increased bone resorption (increased deoxypyridinoline levels) resulted in bone loss, which explains the important reduction of BV/TV and trabecular thickness, as well as the small but significant reduction of alveolar BMD in the animals on a low-protein diet. A hypothesis could be that the anabolic effect of mastication cannot occur with low levels of IGF-I. On the other hand, OVX seems to have a greater impact on the trabecular structure (reduction of trabecular number and connectivity density).

In conclusion, mandibular alveolar bone was found to be less sensitive to protein undernutrition and/or estrogen deprivation than the proximal tibia spongiosa. We hypothesize that the mechanical loading of the alveolar process during mastication may protect the alveolar bone from the detrimental effects observed in other skeletal sites. Higher trabecular thickness and lower bone surface fraction may blunt the deterioration of the trabecular microarchitecture in the case of the mandibular bone. The different embryological origin of the two skeletal sites may also play a role in their response to bone loss stimuli. The results of this study confirm and help to explain the existing evidence of a weak association between oral bone loss and postmenopausal osteoporosis.

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES

This study was supported by a grant from the Swiss National Research Foundation (3200B0-100714). The authors express their gratitude to Isabelle Badoud and Sylvie Vouillamoz for technical assistance.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. Acknowledgements
  8. REFERENCES