Skip to main content
NCBI home page
Journal of Animal Science logoLink to Journal of Animal Science
. 2020 Nov 20;98(12):skaa377. doi: 10.1093/jas/skaa377

Low dietary silicon supplementation may not affect bone and cartilage in mature, sedentary horses

Abby Pritchard 1,✉, Brian D Nielsen 1, Cara Robison 1, Jane M Manfredi 2
PMCID: PMC7749713  PMID: 33216909

Abstract

As osteoarthritis is a major cause of lameness in horses in the United States, improving collagen health prior to onset and increasing collagen turnover within affected joints could improve health- and welfare-related outcomes. Through its positive effects on bone mineral content and density and its role in increasing collagen synthesis, silicon (Si) may slow the development and progression of osteoarthritis, thereby reducing lameness. This study evaluated the hypothesis that Si supplementation would increase cartilage turnover through increased collagen degradation and formation markers, as well as bone formation markers, resulting in reduced lameness severity when compared with controls. Ten mature Standardbred geldings were assigned to either a Si-treated (SIL) or control (CON) group and group-housed on pasture for 84 d. Horses were individually fed to ensure no cross-contamination of Si other than what was present in the environment. For the duration of the study, SIL horses received a Si–collagen supplement at the rate of 0.3 g supplement/(100 kg body weight day). Serum samples were taken weekly for osteocalcin, and plasma samples were taken on days 0, 42, and 84 for plasma minerals. On days 0, 42, and 84, subjective and objective lameness exams were performed, and radiographs and synovial fluid samples were taken from reference and osteoarthritic joints. Plasma minerals were similar in both groups and were lower on day 84 than on day 0 (P < 0.05). Si supplementation, fed at the manufacturer’s recommended rate, did not improve lameness or radiographs when compared with controls, and supplemented horses did not show greater collagen degradation and/or synthesis markers in synovial fluid than controls, indicating that cartilage turnover remained unaffected. However, a minimum beneficial threshold and range for Si supplementation standardized to body weight need to be established.

Keywords: arthritis, cartilage, horses, silicon

Introduction

In the United States, over 65% of stables had one or more resident equids with a lameness problem at any time, and less than half of these horses recovered and remained sound (United States Department of Agriculture, 2017). Unfortunately, these pervasive lamenesses can lead to interrupted training and economic losses as 16.3% of horses retire, become a companion animal, are given away, or euthanized (United States Department of Agriculture, 2017). As over 20% of horses were identified as having a chronic joint problem like arthritis (United States Department of Agriculture, 2017), prevention and management of equine osteoarthritis (OA) is key to maintaining a horse’s soundness, value, and welfare.

OA is characterized by a loss of articular cartilage, often with associated bony and soft tissue changes, leading to pain and decreased function (Schlueter and Orth, 2004; McIlwraith et al., 2013). Articular cartilage loss results from an imbalanced turnover of collagen demonstrated by increased markers of collagen degradation found in synovial fluid collected from OA joints (Frisbie et al., 2008; Nicholson et al., 2010; McIlwraith et al., 2018). In addition, aging results in a loss of polysulfated glycosaminoglycans in healthy articular cartilage (Plat et al., 1998), further reducing collagen synthesis within the joint. Improving collagen health prior to OA onset and increasing collagen repair through higher rates of degradation and synthesis within OA-affected joints, as well as a more balanced ratio of these processes, could improve health- and welfare-related outcomes in horses with OA.

Silicon (Si) plays a crucial role in long bone development (Carlisle, 1980, 1981) and mineralization (Kim et al., 2013, 2014; Arora and Arora, 2017) through decreased bone resorption and increased osteoblast differentiation (Maehira et al., 2009; Mladenović et al., 2014; Dong et al., 2016). Through these effects and its role in increasing collagen synthesis (Calomme and Vanden Berghe, 1997; Reffitt et al., 2003; Dong et al., 2016), Si may influence the development and progression of OA, thereby reducing lameness. In addition, Si may affect the absorption and use of other minerals, such as calcium (Ca) and boron (B), important to bone mineralization (Kim et al., 2014; Jugdaohsingh et al., 2015a). Changes in the availability of these minerals may affect bone health.

Previous research found reduced carboxy-terminal pyridinoline cross-linked telopeptide region of type I collagen, a bone resorption marker, in Si-supplemented young horses compared with controls (Lang et al., 2001). In addition, Si supplementation increased distance trained prior to injury in racehorses (Nielsen et al., 1993). Unfortunately, previous equine research used sodium zeolite A (Frey et al., 1992; Nielsen et al., 1993; Lang et al., 2001), which contained large amounts of aluminum, up to 130,000 mg/kg (O’Connor et al., 2008), posing potential toxicity concerns and absorption interference with other minerals. Since these studies, new Si supplements have been developed, including a purportedly bioavailable silica-collagen peptide source that could be fed in lower amounts than supplements in previous studies. The objective of the current study was to determine the effect of feeding a Si-collagen supplement on markers of collagen degradation and formation and osteoblast activity as well as objective and subjective lameness in mature horses. This study evaluated the hypothesis that Si supplementation would increase bone formation markers and cartilage turnover through increased collagen degradation and formation markers and reduce lameness severity when compared with controls.

Materials and Methods

Horses and diet

All methods were approved by the Michigan State University Institutional Animal Care and Use Committee (AUF no. 11/17-201-00). Ten mature Standardbred geldings were striated by age and body weight (BW), pair-matched based on lameness degree and limb, and randomly assigned to either a Si-treated (SIL) or control (CON) group. Horses were group-housed on summer pasture for 84 d (12 wk) but individually fed from assigned feeders to ensure no cross-contamination of Si other than what was present in the environment. For the duration of the study, SIL horses received a Si-collagen supplement at the rate of 0.3-g supplement/(100 kg BW d), based on manufacturer’s recommendations (Privi Life Sciences Pvt Ltd, Mumbai, Maharashtra, India). After every daily meal, feeders were checked for refusals, and any feed refusals were weighed and recorded. Horses were weighed on day 0 and every 7 d during the study to ensure appropriate dosage, and total amount supplemented was adjusted as needed due to weight gain or loss. All horses received 500 g of textured feed (Meco 15, Mason Elevator Co., Mason, MI; 15% CP, 3% fat, 5.5% fiber) once per day and had free-choice access to grass and water.

Lameness exams and Lameness Locator

A lameness examination of each horse was conducted on days 0, 42, and 84 using a traditional scoring system and a Lameness Locator system (Equinosis, LLC, St. Louis, MO). A baseline for each horse was conducted on a straight line on a flat, hard surface, and distal and proximal flexion tests of all four legs were completed and measured by the system on the same line. A veterinarian who is a Diplomate of both the American College of Veterinary Surgeons-Large Animal and the American College of Veterinary Sports Medicine and Rehabilitation (Equine) and blinded to treatment group scored overall lameness and lameness after flexions using the American Association of Equine Practitioners (AAEP, 1991) scale in 0.5 increments ranging from 0 (sound) to 5 (nonweight bearing). The Lameness Locator system served as an objective observer.

The Lameness Locator consisted of three inertial sensor devices, two accelerometers and a gyroscope, which were positioned between the ears on the horse’s poll, over the pelvis along the dorsal midline, and on the dorsal side of the front right pastern (Keegan et al., 2011). The devices were placed by the same person every time to ensure consistency and remained in place throughout each trial. Each horse was then trotted in-hand on a hard, flat surface for a minimum of 15 strides. The variables of interest included the front vector sum (FVS) and the differences in maximum and minimum pelvic height (HDmax and HDmin). The FVSs were defined as the difference between the maximum and minimum positions of the head, and the threshold for a significant front limb lameness was considered at FVS ≥ 8.5 mm (Keegan et al., 2013). The maximum head difference was generated by the maximum height of the head between left and right portions of the stride compared with that of the left forelimb with the minimum head differences generated the same way by the differences at the minimum height. Similarly, HDmax and HDmin were generated by the differences in the maximum and minimum heights of the pelvis, respectively, during the stance phase of the right hind compared with the left hind limb. Thresholds for hind limb lameness detection using HDmax and HDmin were set at >3 mm for both (Keegan et al., 2013). Once this was complete, the devices were removed from the horse, and feedback from the Lameness Locators was saved for later analysis.

Radiographs

Digital radiographs (Eklin Model EDR3 Mark III, Carlsbad, CA; HF8015+dlp, MinXray Inc., Northbrook, IL) of the left middle carpal (LMC), left radiocarpal (LRC), and left metacarpophalangeal (LFF) joints were considered references as well as any other joint suspected of OA, including right carpal, right metacarpophalangeal, and both tarsal and metatarsophalangeal joints, prior to the start of the study. The same joints were radiographed again on days 42 and 84 to look for evidence or progression of arthritic changes. Dorsopalmar, mediolateral, dorsolateral–palmaromedial oblique, and dorsomedial–palmarolateral oblique views (Smallwood et al., 1985) of all joints were taken and compiled into files per joint per horse per day. Any identifying information for date or horse was removed from the digital radiographic images prior to assignment to a blinded reviewer, the same veterinarian who assessed lameness, for scoring. The reviewer scored each set of radiographs using a scale of 0 to 3 in 0.5 increments using a previously established scoring system (Frisbie et al., 2002; Kawcak et al., 2008). Briefly, a score of 0 indicated no evidence of OA, whereas a score of 3 indicated evidence of severe OA through a combination of subchondral bone lysis, bone proliferation at the joint capsule attachment, and/or osteophyte formation.

Blood samples

Immediately after feeding, blood samples were collected via jugular venipuncture into vacutainer serum separation tubes (BD Vacutainer: SST, Becton, Dickinson and Company, Franklin Lakes, NJ) on day 0 and every 7 d until the end of the study on day 84 for serum osteocalcin (OC) analysis. On days 0, 42, and 84, additional blood samples were collected into acid-washed, K2EDTA treated tubes for plasma Si, B, and Ca analysis. Serum and plasma samples were centrifuged and placed into microcentrifuge tubes before freezing at −20 °C. Serum samples were analyzed using an OC enzyme-linked immunosorbent assay (Microvue Osteocalcin EIA, Quidel, San Diego, CA). Plasma Si, B, and Ca and Si concentration within the supplement were determined by inductively coupled plasma mass spectrometry (Agilent 7900 Inductively Coupled Plasma Mass Spectrometer, Agilent, Santa Clara, CA) at a certified laboratory (Michigan State University Veterinary Diagnostic Laboratory, Lansing, MI).

Synovial fluid samples

Synovial fluid samples were collected aseptically on days 0, 42, and 84 from the LFF, LMC, and LRC joints that served as reference joints; samples were also collected from any other joint suspected of OA based on radiographic and lameness evaluations. One SIL horse had no additional joints suspected of OA, and for safety reasons, one CON horse only had LMC sampled. These samples were divided into cryotubes and placed immediately on dry ice until transferred to the −80 °C storage location. Synovial fluid was digested using 50 units/mL of hyaluronidase at 37 °C for 40 min to reduce viscosity (O’Connor-Robison et al., 2014). Synovial fluid was analyzed for procollagen type II C-propeptides (CPII) and triple helix fragments of both types I and II collagen (C1,2C) as markers of collagen formation and degradation, respectively, using commercial assays (Ibex Pharmaceuticals, Montreal, QC, Canada). Only reference joints were included for CPII analysis, whereas all joints sampled were included for C1,2C analysis. Differences between groups were examined at each sampling point, as well as for changes from baseline.

Statistical analysis

Residuals of data were tested for normality and log transformed as necessary. Absolute values of the variables generated from the Lameness Locator were used for analysis. Continuous data from biomarker, mineral, and Lameness Locator analyses were analyzed using the mixed model procedure (Proc Mixed) in SAS 9.4 (SAS Institute, Cary, NC). Data from radiographic and lameness scores were treated as continuous and analyzed using the generalized linear mixed model procedure (Proc Glimmix). Fixed effects for each analysis included group, day or week, and their interactions. For OC and plasma minerals, horse was included as a random effect with week used as a repeated measure for OC only. Lameness scores and radiographic evaluations included the trial or leg within horse as a random effect. Lameness Locator data were divided into a straight line, and flexion trials with data were analyzed separately by limb flexion. Concentrations of CPII and C1,2C were analyzed by each separate joint. Because nonreference joints were not pair-matched, concentrations of C1,2C were pooled to examine overall effect of group and joint was set as a random effect. Tukey’s post hoc analysis was used for comparisons among days or between groups for significant effects. Significance was set at P < 0.05, and due to the small number of subjects per group (n = 5), trends were considered at 0.05 ≤ P < 0.10.

Results

Horses and diet

Mean age and BW for the groups were 13 ± 1 yr and 541 ± 8 kg for SIL and 10 ± 2 yr and 542 ± 4 kg for CON. BW remained similar between groups, but BW increased in both groups from day 0 to day 84 (P < 0.01). The supplement contained 32.5 mg Si/g, whereas the textured feed had 0.004 mg Si/g. Similar to BW, supplement amounts increased from day 0 to day 84 (1.57 ± 0.1 and 1.65 ± 0.1 g, respectively; P < 0.0001), and mean supplement amount per horse over the course of the study was 1.62 ± 0.02 g/d. Si intake in CON was 1.8 mg/d from feed compared with 54.5 ± 0.8 mg/d in SIL from feed and 1.62-g supplement. There were no feed refusals over the course of the study.

Subjective and objective Lameness evaluations

Subjective lameness scores remained unaffected by group and day. When examined as a change from baseline, there was no difference between the groups (Table 1). From the Lameness Locator straight-line trials, there was no difference between groups or among days for any of the variables (Table 2). During all flexion trials, FVS was unaffected by group and day. Both HDmin and HDmax values experienced a day × group interaction (P = 0.003). In HDmin, CON values remained similar over the course of the study, but SIL values decreased from day 0 to day 84 (P = 0.02). For HDmax, SIL values were similar over time, but CON values decreased from day 0 to day 84 (P = 0.003).

Table 1.

Mean (± SE) subjective lameness scores by a licensed veterinarian for each leg using the American Association for Equine Practitioners lameness scale in CON (n = 5) and SIL (n = 5)

Day
Leg Group1 0 42 84
Left front CON 1.0 ± 0.4 1.2 ± 0.4 1.4 ± 0.4
SIL 0.4 ± 0.4 0.2 ± 0.4 0.4 ± 0.4
Left hind CON 1.0 ± 0.3 1.0 ± 0.3 0.2 ± 0.3
SIL 0.6 ± 0.3 0.2 ± 0.3 0.6 ± 0.3
Right front CON 0.8 ± 0.4 0.7 ± 0.4 0.5 ± 0.4
SIL 1.2 ± 0.4 0.9 ± 0.4 0.8 ± 0.4
Right hind CON 0.6 ± 0.3 0.4 ± 0.3 1.0 ± 0.3
SIL 1.1 ± 0.3 1.0 ± 0.3 0.9 ± 0.3
Open in a new tab

1CON, control; SIL, silicon treated.

Table 2.

Mean (± SE) mm of front vector sum (FVS), difference in minimum pelvic height (HDmin), and difference in maximum pelvic height (HDmax) generated from a Lameness Locator system in straight-line trials without joint flexions in CON (n = 5) and SIL (n = 5) horses

Day
Variable Group1 0 42 84
FVS CON 20.7 ± 3.2 19.5 ± 3.4 21.1 ± 3.2
SIL 13.7 ± 3.2 21.0 ± 3.6 12.5 ± 3.2
HDmin CON 6.2 ± 0.9 5.4 ± 1.0 4.8 ± 0.9
SIL 5.2 ± 0.9a 3.5 ± 1.0b 3.3 ± 0.9b
HDmax CON 5.0 ± 1.1a 4.9 ± 1.2a 2.8 ± 1.1b
SIL 4.5 ± 1.1 6.1 ± 1.3 4.2 ± 1.1
Open in a new tab

1CON, control; SIL, silicon treated.

a,bValues within a row lacking a common superscript differ (P < 0.05).

Radiographs

Overall, radiographic scores were unaffected by group in reference and OA joints, and scores were greater (P = 0.002) in OA joints than reference joints, verifying the scoring system. Radiographic scores in the LMC and LRC tended to increase over time (P = 0.07), but comparisons were nonsignificant in post hoc analysis. In remaining reference and OA joints, scores remained similar throughout the study.

Blood markers

OC concentrations did not differ between groups but did increase in certain weeks throughout the study (P < 0.001, Figure 1). Plasma Ca and B concentrations were unaffected by group. Ca concentrations remained similar from day 0 to day 42 (116 ± 1 and 117 ± 1 µg/mL, respectively) but decreased at day 84 below both previous sampling days (112 ± 1 µg/mL, P < 0.05 for both comparisons). However, B concentrations increased from day 0 to day 42 (0.094 ± 0.003 and 0.129 ± 0.003 µg/mL, respectively; P < 0.001) and decreased from day 42 to day 84 (0.104 ± 0.003 µg/mL; P < 0.001). Interestingly, Si supplementation did not increase plasma Si concentrations when compared with controls, but plasma Si decreased from day 0 to day 84 in all horses (1.06 ± 0.07 and 0.59 ± 0.07 µg/mL for day 0 and 84, respectively; P < 0.05).

Figure 1.

Figure 1.

Open in a new tab

Pooled mean serum OC concentrations (± SE) collected every 7 d beginning on day 0 until day 84 from control (n = 5) and supplemented (n = 5) horses. a,b,cDays lacking a common letter differ (P < 0.05).

Synovial fluid markers

In two of the reference joints, LRC and LMC, C1,2C concentrations did not differ by group or day (Table 3). In LFF, overall C1,2C concentrations tended to be greater in SIL when compared with CON (0.35 ± 0.2 and 0.30 ± 0.02 µg/mL, respectively; P = 0.08) but were similar at all sampling days. When analyzed as change from day 0, there were no differences between groups or days. In addition, in nonreference joints with OA, C1,2C concentrations were similar between groups (CON: 0.33 ± 0.3 µg/mL; SIL: 0.34 ± 0.3 µg/mL). There was evidence for a day × group interaction (P < 0.05), but there was no change within groups over the course of the study or difference between groups at each sampling day in post hoc analysis. CPII concentrations did not show any differences between groups or among days (Table 4). When analyzed as a change from baseline, there were no differences between groups or days.

Table 3.

Mean (± SE) triple helix fragments of C1,2C concentrations (µg/mL) as a marker of collagen degradation in synovial fluid taken from the left medial carpal (LMC), left radiocarpal (LRC), left metacarpophalangeal (LFF), and osteoarthritis (OA) joints in CON (n = 5) and SIL (n = 5) horses

Day
Joint Group1 0 42 84
LMC CON 0.37 ± 0.04 0.30 ± 0.04 0.32 ± 0.04
SIL 0.42 ± 0.04 0.35 ± 0.04 0.36 ± 0.04
LRC CON 0.36 ± 0.05* 0.27 ± 0.07 0.24 ± 0.06
SIL 0.36 ± 0.05 0.37 ± 0.08 0.28 ± 0.07
LFF CON 0.31 ± 0.04* 0.30 ± 0.04* 0.29 ± 0.04*
SIL 0.36 ± 0.04 0.45 ± 0.04 0.30 ± 0.04
OA CON 0.37 ± 0.05* 0.23 ± 0.08* 0.20 ± 0.08*
SIL 0.26 ± 0.05* 0.31 ± 0.08* 0.24 ± 0.09*
Open in a new tab

1CON, control; SIL, silicon treated.

*n = 4.

Table 4.

Mean (± SE) CPII concentrations (ng/mL) as a marker of collagen formation in synovial fluid taken from the left medial carpal (LMC), left radiocarpal (LRC), and left metacarpophalangeal (LFF) joints in CON (n = 5) and SIL (n = 5) horses

Day
Joint Group1 0 42 84
LMC CON 2,306 ± 812 949 ± 638 1,333 ± 638
SIL 2,037 ± 803 1,512 ± 605 1,516 ± 605
LRC CON 1,624 ± 363* 1,400 ± 314 1,137 ± 353
SIL 2,069 ± 321 1,489 ± 295 1,148 ± 346
LFF CON 1,490 ± 601* 717 ± 464* 846 ± 464*
SIL 1,532 ± 595 1,051 ± 444 883 ± 449
Open in a new tab

1CON, control; SIL, silicon treated.

*n = 4.

Discussion

Lameness and OA present large problems for the equine industry. While promising in previous studies, Si supplementation in the current study did not affect lameness as subjective and objective evaluations were similar between groups. While none of the changes in the study could be attributed to group, OC peaked during the first, third, sixth, and final weeks, and plasma mineral concentrations declined over the course of the study. For the mature horses in the current study, Si supplementation did not change whole-body or joint-specific measures of lameness and collagen metabolism.

Overall, determining the effect of Si supplementation on radiographic progression of OA and lameness proved difficult as horses in the current study presented little radiographic evidence of OA in reference joints with 71% of scores at 0.5 or below. In addition, even after flexion, 77.5% of trials scored at 1 or below on the AAEP scale. While these horses were mature and previously raced, most were relatively sound, making differences between groups and over time difficult to observe with subjective scoring. Trained clinicians can agree on lameness scores above 1.5, that is, around 93% of the time, but this agreement decreases to 62% when lameness is mild (Keegan et al., 2010), indicating the difficulty with using a subjective evaluation to determine group differences. However, subjective evaluations are standard clinical practice, making them an important incorporation in industry-applied research. In addition, the inclusion of an objective observer in the Lameness Locator system strengthens these subjective evaluations as this system produces similar findings to subjective evaluations (Keegan et al., 2013; Donnell et al., 2015). Previous research indicates that an FVS ≥ 8.5 mm but ≤20 mm will typically be scored at a 1 or below on the AAEP scale, whereas HDmin and HDmax range from ≥3 to ≤9 mm and ≤4 mm, respectively, for the same score (Keegan et al., 2013). In the current study, all Lameness Locator readings stayed close to these ranges, demonstrating the accuracy and consistency of the objective evaluations as well as the relative soundness of the research horses. While having healthy and relatively comfortable research horses is important, it did make it difficult in determining whole-body differences between groups. To better examine the effects of Si supplementation on lameness, a minimum lameness threshold with average scores above 1.5 should be established as inclusion criteria for horses in future studies.

Activity and exercise can affect bone metabolism markers more than nutrition alone (Nielsen and Spooner, 2008), and increased bouts of free-choice exercise in response to external stimuli may have influenced OC in the current study. While OC was unaffected by Si supplementation, concentrations of this marker spiked on days 0, 14, 42, and 84, potentially related to increased activity secondary to external factors. In the few days prior to sampling on day 0, horses were moved to a new pasture that served as group-housing for the current study, subsequently increasing exploration and flight response to stimuli in a new environment. The weeks of day 14 and day 42 corresponded with farm construction near the pasture and American Fourth of July celebrations, respectively, which could have spooked horses and increased their activity. While activity was not monitored in the current study, external stimuli outside of the researchers’ control may have increased free-choice exercise, while horses were on pasture, affecting OC concentrations in the current study.

Interestingly, plasma Si was not greater in the supplemented group when compared with controls. Previous research using sodium zeolite A as the Si source indicated that plasma Si concentrations should be higher in supplemented horses (Lang et al., 2001) and remain higher even at 6 to 9 h postfeeding (Frey et al., 1992). However, serum Si concentrations depend on source and renal clearance (Sripanyakorn et al., 2009). In a previous study in humans, serum Si did not peak until 3 h postconsumption of a choline-stabilized orthosilicic acid source and remained relatively stable after consumption of magnesium trisilicate (Sripanyakorn et al., 2009). As plasma samples were collected within an hour after feeding, this may not have been enough time to allow Si from the supplement to become fully bioavailable. However, supplementation of orthosilicic acid in horses did not produce greater plasma Si than controls (O’Connor et al., 2008), indicating that changes in plasma Si in horses may also depend on the Si source. The source in the current study was a Si-collagen supplement, which should have prevented the spontaneous formation of insoluble oligomers and silicates (Belton et al., 2012), which would have reduced absorption (Sripanyakorn et al., 2009). However, as there was no difference in plasma Si between groups and no increase in SIL over time, the source in the current study may not be as absorbable or alter plasma Si as much as previous sources.

The decrease in plasma Ca and Si concentrations in both groups probably resulted from pasture growth and declining digestibility with pasture maturation. Grasses and grains (such as oats and rice) accumulate Si and sometimes Ca more than other species as a mechanism to protect against biotic and abiotic stresses (Currie and Perry, 2007; Bauer et al., 2011; Belton et al., 2012). Si is usually deposited in plant cell walls and greater concentrations can be found in wounded and older cells (Bauer et al., 2011). In addition, increased Si concentrations in grasses are associated with reduced digestibility, providing both chemical protection of cell wall components and physical protection from chewing (Massey et al., 2008; Bauer et al., 2011). Due to the predominant cool-season grasses in Michigan and their growth cycle, pasture was likely older as this study took place over the summer. With reduced growth, more Si was deposited into older cells, likely reducing the digestibility and availability of nutrients from this pasture. This dormancy of cool-season grasses during the heat of the summer may have caused plasma Ca and Si to decrease over the course of the study. Unfortunately, grass samples were not collected, but given the decrease in plasma Si in both groups and group housing on pasture, this may have been the cause.

As group did not affect plasma B concentrations, the changes in both groups over the course of the study could be attributed to B accumulation and grass digestibility, as noted above. Grasses have low requirement for B but can concentrate it up to 800 mg/kg before reaching toxic levels (Gupta, 2016). B concentrations increase with plant maturity, mirroring similar increasing concentration in older bone (Jugdaohsingh et al., 2015a), and its role is less structural than Ca and Si as deficiencies interrupt protein, nucleic acid, and sugar metabolism, cellular membrane function, and seed production (Brown et al., 2002; Gupta, 2016). The peak of concentrations on day 42 and decline of concentrations to day 84 could represent this B accumulation; as grasses accumulated more B and Si and matured, digestibility decreased, affecting B concentrations in the second half of this study. Because both groups follow the same pattern, the cause was likely the pasture, but this remains speculative. Due to the potential effects of pasture growth in the current study, future research should provide single-source hay from the same cutting.

Overall, collagen metabolism markers in synovial fluid were relatively unaffected by Si supplementation. Use of these markers in synovial fluid, rather than serum, allowed for analysis at individual joints as collagen biomarker concentrations can vary by joint (Nicholson et al., 2010), and serum values may be inflated and poorly correlated with synovial fluid concentrations (Frisbie et al., 2008; Catterall et al., 2010; Nicholson et al., 2010). While C1,2C concentrations in LFF tended to be higher in SIL, changes from day 0 were not different between the groups. Rather than as a result of supplementation, the higher concentrations in SIL probably resulted from other factors outside the control of the study, such as amount or viscosity of synovial fluid produced by the horse. In addition, C1,2C concentrations were similar among metacarpophalangeal and carpal joints with values similar to previous reports in both joints without OA (Frisbie et al., 2008; Nicholson et al., 2010). Values for CPII are similar to previous studies ranging from around 1,000 ng/mL (Donabédian et al., 2008; Nicholson et al., 2010) to over 3,000 ng/mL (Frisbie et al., 2008) in normal and OA-affected joints. Concentrations increase in OA-affected joints and in response to exercise (Frisbie et al., 2008; Nicholson et al., 2010). As horses can have lower concentrations of CPII in metacarpophalangeal joints when compared with carpal joints (Nicholson et al., 2010), LFF concentrations remained relatively unchanged and similar to concentrations in non-OA joints (Donabédian et al., 2008; Frisbie et al., 2008; Nicholson et al., 2010). With consistent exercise, CPII concentrations can increase (Frisbie et al., 2008), but a bout of high-intensity exercise changes degradation markers in synovial fluid very little in the 24-h postexercise (Macnicol et al., 2020). As noted above with OC, activity may have increased during certain weeks, but this activity may not have been consistent enough to influence CPII concentrations. Overall, activity was probably low, leading to the decrease in CPII seen in the current study. As concentrations in both collagen markers were similar to previous studies and no differences between groups were observed after day 0, Si supplementation did not increase collagen degradation or synthesis within joints. However, exercise can stimulate extracellular matrix production, including collagen, by chondrocytes (Huber et al., 2000) and improve nutrient exchange through synovial fluid flow (Bertone, 2008), future studies should consider including a light-to-moderate workload for horses. The increased nutrient exchange may deliver more Si to synovial fluid to stimulate greater collagen synthesis. Including exercise and its potential interactions with Si supplementation will better inform owners and trainers of riding horses with OA.

Although supplemented horses received nearly 54 mg Si/d based on the manufacturer’s recommendations, this amount falls short of proposed beneficial threshold in humans on a kg BW basis. In humans, research suggests 0.3 mg Si/(kg BW d) as the minimum level to have positive influences on bone (Nielsen, 2014). In rats, favorable effects on bone are seen at a much higher supplementation rate of 54 to 94 mg and 37 to 60 mg Si/(kg BW d) for females and males, respectively (Jugdaohsingh et al., 2015b). In broilers, differences in bone strength and ash among controls and varying Si concentrations are achieved beginning at 14 mg Si/(kg BW d) (Scholey et al., 2018). Mean BW in the current study was around 540 kg, meaning SIL received 0.1 mg Si/(kg BW d), so the amount of Si in the current study falls below the minimum level recommended for humans and well below the experimental amounts given to rats. To achieve the proposed human minimum level, horses in the current study would need to receive 162 mg Si/d or around 5 g of supplement, and to match minimum experimental amounts, nearly 7.5 g Si/d or 231 g of supplement would need to be fed. The supplement would then equal nearly 1% to 46% of the textured feed given. These amounts, especially the latter, may present palatability issues, and many of the experimental amounts are well above the threshold for humans previously suggested by Nielsen (2014). Another study in broilers from an early age using between 0.005 and 0.015 mg Si/(kg BW d) (Sgavioli et al., 2016) found changes in the mineral profile of bones with supplementation but no difference in bone density or strength. However, these birds started receiving supplementation at an earlier stage of development compared with the horses in the current study, indicating that Si supplementation may be more influential during growth. Age at start of supplementation may play a role in bone response.

In addition, most of the proposed Si thresholds for amount fed do not account for amount absorbed. Absorption of Si, and therefore bioavailability, depends on source, and Si supplement absorption can range from 1% to 64% of amount ingested with colloidal silica being least absorbable and monomethyl silanetriol being most absorbable (Sripanyakorn et al., 2009). Orthosilicic acid, considered the most bioavailable form of Si, is around 41% digestible in horses (O’Connor et al., 2008) and 43% absorbable in humans (Sripanyakorn et al., 2009), whereas sodium zeolite A is around 51% digestible in horses (O’Connor et al., 2008). These absorption percentages, coupled with the low supplementation rate suggested by the manufacturer, may explain the lack of differences in the current study. Assuming the highest demonstrated absorption rate of 64%, the supplement would have provided around 34 mg Si/d or 0.06 mg Si/kg BW. Previous studies in horses provided around 430 mg to 16.5 g Si/d from sodium zeolite A and 474 mg/d from orthosilicic acid (Lang et al., 2001; O’Connor et al., 2008). Even at 474 mg Si/d, O’Connor et al. (2008) did not find an increase in plasma Si. These previous rates indicate that the supplementation rate recommended by the manufacturer in the current study (0.3 g/100 kg BW) could have been too low to demonstrate changes in plasma concentrations, despite purported greater bioavailability. Determining the minimum amount of Si needed to increase plasma Si may be necessary to calculate future intakes, rather than simply relying on manufacturer recommendations. However, as noted in the previous paragraph, these amounts may present palatability and/or cost issues in large animals. Implementing high experimental amounts that generate results in adult small animals may be difficult to achieve and unfeasible in larger species, depending on the source. Amount of Si provided, form (powder or liquid), and total amount of the supplement needed to achieve Si thresholds should be considered when designing future studies.

Conclusions

These results indicate that Si supplementation in older horses may not be as beneficial as in younger animals and that it may be difficult to achieve beneficial experimental amounts in large animals due to volume of Si supplement needed to be fed. In previous studies, Si supplementation showed potential to improve bone and cartilage health in many species, including horses and humans, but positive effects from Si were only achieved at high concentrations in the diet, making them difficult to replicate as animals get larger. Si can positively influence bone density, structure, and strength, as well as collagen metabolism in connective tissues, indicating the possibility to affect degraded cartilage and arthritis in older animals. Since arthritis is a prominent cause of lameness in the equine industry, supplementing Si would be an easy way to address this issue. However, Si supplementation did not improve lameness when compared with controls. In addition, supplemented horses did not show increases in collagen degradation and synthesis markers in synovial fluid than controls, and concentrations of these markers did not change from baseline, indicating that cartilage turnover remained unaffected. Future studies should determine the minimum threshold at which Si supplementation increases plasma or serum Si and establish a beneficial range on a kg BW/d basis.

Acknowledgments

The authors thank Emma Stapley and Brady Stutzman for their help with radiographs, lameness evaluations, and synovial fluid collection as well as all of the undergraduate assistants for their hard work taking care of the horses.

Glossary

Abbreviations

BW

body weight

C1,2C

triple helix fragments of both types I and II collagen

CPII

procollagen type II C-propeptides

FVS

front vector sum

HDmax

difference in maximum pelvic height

HDmin

difference in minimum pelvic height

LFF

left metacarpophalangeal

LMC

left medial carpal

LRC

left radiocarpal

OA

osteoarthritis

OC

osteocalcin

Funding

This project was funded as part of a grant from the Michigan Animal Agriculture Alliance (M-AAA-18-010), and the supplement was provided by Equi-Force Equine Products, LLC.

Conflict of interest statement

The authors declare no real or perceived conflicts of interest.

Literature Cited

  1. American Association of Equine Practitioners 1991. Definition and classification of lameness. In: American Association of Equine Practitioners, editor. Guide for veterinary service and judging of equestrian events. American Association of Equine Practitioners. Lexington (KY): American Association of Equine Practitioners; p. 19. [Google Scholar]
  2. Arora M., and Arora E.. . 2017. The promise of silicon: bone regeneration and increased bone density. J. Arthrosc. Jt. Surg. 4:103–105. doi: 10.1016/j.jajs.2017.10.003 [DOI] [Google Scholar]
  3. Bauer P., Elbaum R., and Weiss I. M.. . 2011. Calcium and silicon mineralization in land plants: transport, structure and function. Plant Sci. 180:746–756. doi: 10.1016/j.plantsci.2011.01.019 [DOI] [PubMed] [Google Scholar]
  4. Bertone, A. L 2008. Joint physiology: responses to exercise and training. In: Hinchcliff, K. W., R. J. Geor, and A. J. Kaneps, editors. Equine Exercise Physiology: The Science of Exercise in the Athletic Horse. 1st ed. Edinburgh: Elsevier Ltd; p. 132–142. doi: 10.1016/B978-0-7020-2857-1.50006-8 [DOI] [Google Scholar]
  5. Belton D. J., Deschaume O., and Perry C. C.. . 2012. An overview of the fundamentals of the chemistry of silica with relevance to biosilicification and technological advances. FEBS J. 279: 1710–1720. doi: 10.1111/j.1742-4658.2012.08531.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Brown P. H., Bellaloui N., Wimmer M. A., Bassil E. S., Ruiz J., Hu H., Pfeffer H., and Dannel F.. . 2002. Boron in plant biology. Plant Biol. 4:205–223. [Google Scholar]
  7. Calomme M., and Vanden Berghe D.. . 1997. Supplementation of calves with stabilized orthosilicic acid: Effect on the Si, Ca, Mg, and P concentrations in serum and the collagen concentration in skin and cartilage. Biol. Trace Elem. Res. 56:153–165. doi: 10.1007/BF02785389 [DOI] [PubMed] [Google Scholar]
  8. Carlisle E. M. 1980. Biochemical and morphological changes associated with long bone abnormalities in silicon deficiency. J. Nutr. 110:1046–1056. doi: 10.1093/jn/110.5.1046 [DOI] [PubMed] [Google Scholar]
  9. Carlisle E. M. 1981. Silicon: A requirement in bone formation independent of vitamin D1. Calcif. Tissue Int. 33:27–34. doi: 10.1007/BF02409409 [DOI] [PubMed] [Google Scholar]
  10. Catterall J., Stabler T., Flannery C., and Kraus V.. . 2010. Changes in serum and synovial fluid biomarkers after acute injury. Arthritis Res. Ther. 12:R229. doi: 10.1186/ar3216 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Currie H. A., and Perry C. C.. . 2007. Silica in plants: biological, biochemical and chemical studies. Ann. Bot. 100:1383–1389. doi: 10.1093/aob/mcm247 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Donabédian M., van Weeren P. R., Perona G., Fleurance G., Robert C., Léger S., Bergero D., Lepage O., and Martin-Rosset W.. . 2008. Early changes in biomarkers of skeletal metabolism and their association to the occurrence of osteochondrosis (OC) in the horse. Equine Vet. J. 40:253–259. doi: 10.2746/042516408X273657 [DOI] [PubMed] [Google Scholar]
  13. Dong M., Jiao G., Liu H., Wu W., Li S., Wang Q., Xu D., Li X., Liu H., and Chen Y.. . 2016. Biological silicon stimulates collagen type 1 and osteocalcin synthesis in human osteoblast-like cells through the BMP-2/Smad/RUNX2 signaling pathway. Biol. Trace Elem. Res. 173:306–315. doi: 10.1007/s12011-016-0686-3 [DOI] [PubMed] [Google Scholar]
  14. Donnell J. R., Frisbie D. D., King M. R., Goodrich L. R., and Haussler K. K.. . 2015. Comparison of subjective lameness evaluation, force platforms and an inertial-sensor system to identify mild lameness in an equine osteoarthritis model. Vet. J. 206:136–142. doi: 10.1016/j.tvjl.2015.08.004 [DOI] [PubMed] [Google Scholar]
  15. Frey K. S., Potter G. D., Odom T. W., Senor D. M., Reagan V. D., Weir V. H., Elslander J., Webb S. P., Morris E. L., Smith W. B., . et al. 1992. Plasma silicon and radiographic bone density in weanling quarter horses fed sodium zeolite A1. J. Equine Vet. Sci. 12:292–296. 10.1016/S0737-0806(06)81335-3 [DOI] [Google Scholar]
  16. Frisbie D. D., Ghivizzani S. C., Robbins P. D., Evans C. H., and McIlwraith C. W.. . 2002. Treatment of experimental equine osteoarthritis by in vivo delivery of the equine interleukin-1 receptor antagonist gene. Gene Ther. 9:12–20. doi: 10.1038/sj.gt.3301608 [DOI] [PubMed] [Google Scholar]
  17. Frisbie D. D., Al-Sobayil F., Billinghurst R. C., Kawcak C. E., and McIlwraith C. W.. . 2008. Changes in synovial fluid and serum biomarkers with exercise and early osteoarthritis in horses. Osteoarthr. Cartilage 16:1196–1204. doi: 10.1016/j.joca.2008.03.008 [DOI] [PubMed] [Google Scholar]
  18. Gupta U. C. 2016. Boron. In: Barker A. V. and Pilbeam D. J., editors. Handbook of plant nutrition. 2nd ed. Boca Raton (FL): Taylor & Francis Group; p. 241–277. [Google Scholar]
  19. Huber M., Trattnig S., and Lintner F.. . 2000. Anatomy, biochemistry, and physiology of articular cartilage. Invest Radiol. 35:573–580. doi: 10.1097/00004424-200010000-00003 [DOI] [PubMed] [Google Scholar]
  20. Jugdaohsingh R., Pedro L. D., Watson A., and Powell J. J.. . 2015a. Silicon and boron differ in their localization and loading in bone. Bone Rep. 1:9–15. doi: 10.1016/j.bonr.2014.10.002 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Jugdaohsingh R., Watson A. I., Bhattacharya P., van Lenthe G. H., and Powell J. J.. . 2015b. Positive association between serum silicon levels and bone mineral density in female rats following oral silicon supplementation with monomethylsilanetriol. Osteoporos. Int. 26:1405–1415. doi: 10.1007/s00198-014-3016-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kawcak C. E., Frisbie D. D., Werpy N. M., Park R. D., and Mcllwraith C. W.. . 2008. Effects of exercise vs experimental osteoarthritis on imaging outcomes. Osteoarthr Cartil. 16:1519–1525. doi: 10.1016/j.joca.2008.04.015 [DOI] [PubMed] [Google Scholar]
  23. Keegan K. G., Dent E. V., Wilson D. A., Janicek J., Kramer J., Lacarrubba A., Walsh D. M., Cassells M. W., Esther T. M., Schiltz P., . et al. 2010. Repeatability of subjective evaluation of lameness in horses. Equine Vet. J. 42:92–97. doi: 10.2746/042516409X479568 [DOI] [PubMed] [Google Scholar]
  24. Keegan K. G., Kramer J., Yonezawa Y., Maki H., Pai P. F., Dent E. V., Kellerman T. E., Wilson D. A., and Reed S. K.. . 2011. Assessment of repeatability of a wireless, inertial sensor-based lameness evaluation system for horses. Am. J. Vet. Res. 72:1156–1163. doi: 10.2460/ajvr.72.9.1156 [DOI] [PubMed] [Google Scholar]
  25. Keegan K. G., Wilson D. A., Kramer J., Reed S. K., Yonezawa Y., Maki H., Pai P. F., and Lopes M. A.. . 2013. Comparison of a body-mounted inertial sensor system-based method with subjective evaluation for detection of lameness in horses. Am. J. Vet. Res. 74:17–24. doi: 10.2460/ajvr.74.1.17 [DOI] [PubMed] [Google Scholar]
  26. Kim E. J., Bu S. Y., Sung M. K., and Choi M. K.. . 2013. Effects of silicon on osteoblast activity and bone mineralization of MC3T3-E1 cells. Biol. Trace Elem. Res. 152:105–112. doi: 10.1007/s12011-012-9593-4 [DOI] [PubMed] [Google Scholar]
  27. Kim M. H., Kim E. J., Jung J. Y., and Choi M. K.. . 2014. Effect of water-soluble silicon supplementation on bone status and balance of calcium and magnesium in male mice. Biol. Trace Elem. Res. 158:238–242. doi: 10.1007/s12011-014-9936-4 [DOI] [PubMed] [Google Scholar]
  28. Lang K. J., Nielsen B. D., Waite K. L., Hill G. A., and Orth M. W.. . 2001. Increased plasma silicon concentrations and altered bone resorption in response to sodium zeolite a supplementation in yearling horses. J. Equine Vet. Sci. 21:550–555. 10.1016/S0737-0806(01)70161-X [DOI] [Google Scholar]
  29. Macnicol J. L., Lindinger M. I., and Pearson W.. . 2020. A time-course evaluation of inflammatory and oxidative markers following high-intensity exercise in horses: A pilot study. J. Appl. Physiol. 124:860–865. doi: 10.1152/japplphysiol.00461.2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Maehira F., Miyagi I., and Eguchi Y.. . 2009. Effects of calcium sources and soluble silicate on bone metabolism and the related gene expression in mice. Nutrition 25:581–589. doi: 10.1016/j.nut.2008.10.023 [DOI] [PubMed] [Google Scholar]
  31. Massey F. P., Smith M. J., Lambin X., and Hartley S. E.. . 2008. Are silica defences in grasses driving vole population cycles? Biol. Lett. 4:419–422. doi: 10.1098/rsbl.2008.0106 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. McIlwraith C. W., Frisbie D. D., and Kawcak C. E.. . 2013. The horse as a model of naturally occurring osteoarthritis. Bone Joint Res. 1:297–309. 10.1302/2046-3758.111.2000132 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. McIlwraith C. W., Kawcak C. E., Frisbie D. D., Little C. B., Clegg P. D., Peffers M. J., Karsdal M. A., Ekman S., Laverty S., Slayden R. A., . et al. 2018. Biomarkers for equine joint injury and osteoarthritis. J. Orthop. Res. 36:823–831. doi: 10.1002/jor.23738 [DOI] [PubMed] [Google Scholar]
  34. Mladenović Ž., Johansson A., Willman B., Shahabi K., Björn E., and Ransjö M.. . 2014. Soluble silica inhibits osteoclast formation and bone resorption in vitro. Acta Biomater. 10: 406–418. doi: 10.1016/j.actbio.2013.08.039 [DOI] [PubMed] [Google Scholar]
  35. Nicholson A. M., Trumble T. N., Merritt K. A., and Brown M. P.. . 2010. Associations of horse age, joint type, and osteochondral injury with serum and synovial fluid concentrations of type II collagen biomarkers in Thoroughbreds. Am. J. Vet. Res. 71: 741–749. doi: 10.2460/ajvr.71.7.741 [DOI] [PubMed] [Google Scholar]
  36. Nielsen F. H. 2014. Update on the possible nutritional importance of silicon. J. Trace Elem. Med. Biol. 28:379–382. doi: 10.1016/j.jtemb.2014.06.024 [DOI] [PubMed] [Google Scholar]
  37. Nielsen B. D., Potter G. D., Morris E. L., Odom T. W., Senor D. M., Reynolds J. A., Smith W. B., Martin M. T., and Bird E. H.. . 1993. Training distance to failure in young racing quarter horses fed sodium zeolite A. J. Equine Vet. Sci. 13:562–567. 10.1016/S0737-0806(06)81526-1 [DOI] [Google Scholar]
  38. Nielsen B. D., and Spooner H. S.. . 2008. Small changes in exercise, not nutrition, often result in measurable changes in bone. Comp. Exerc. Physiol. 5:15–20. 10.1017/S1478061508914493 [DOI] [Google Scholar]
  39. O’Connor C. I., Nielsen B. D., Woodward A. D., Spooner H. S., Ventura B. A., and Turner K. K.. . 2008. Mineral balance in horses fed two supplemental silicon sources. J. Anim. Physiol. Anim. Nutr. (Berl). 92:173–181. doi: 10.1111/j.1439-0396.2007.00724.x [DOI] [PubMed] [Google Scholar]
  40. O’Connor-Robison C. I., Spencer J. D., and Orth M. W.. . 2014. The impact of dietary long-chain polyunsaturated fatty acids on bone and cartilage in gilts and sows. J. Anim. Sci. 92:4607–4615. doi: 10.2527/jas.2013-7028 [DOI] [PubMed] [Google Scholar]
  41. Plat D., Bird J. L., Bayliss M. T., Platt D., Bird J. L., and Bayliss M. T.. . 1998. Ageing of equine articular cartilage: Structure and composition of aggrecan and decorin. Equine Vet. J. 30:43–52. 10.1111/j.2042-3306.1998.tb04087.x [DOI] [PubMed] [Google Scholar]
  42. Reffitt D. M., Ogston N., Jugdaohsingh R., Cheung H. F., Evans B. A., Thompson R. P., Powell J. J., and Hampson G. N.. . 2003. Orthosilicic acid stimulates collagen type 1 synthesis and osteoblastic differentiation in human osteoblast-like cells in vitro. Bone 32:127–135. doi: 10.1016/s8756-3282(02)00950-x [DOI] [PubMed] [Google Scholar]
  43. Schlueter A. E., and Orth M. W.. . 2004. Equine osteoarthritis: a brief review of the disease and its causes. Equine Comp. Exerc. Physiol. 1:221–231. 10.1079/ecp200428 [DOI] [Google Scholar]
  44. Scholey D. V., Belton D. J., Burton E. J., and Perry C. C.. . 2018. Bioavailability of a novel form of silicon supplement. Sci. Rep. 8:17022. doi: 10.1038/s41598-018-35292-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Sgavioli S., de Faria Domingues C. H., Castiblanco D. M., Praes M. F., Andrade-Garcia G. M., Santos E. T., Baraldi-Artoni S. M., Garcia R. G., and Junqueira O. M.. . 2016. Silicon in broiler drinking water promotes bone development in broiler chickens. Br. Poult. Sci. 57:693–698. doi: 10.1080/00071668.2016.1206190 [DOI] [PubMed] [Google Scholar]
  46. Smallwood J. E., Shively M. J., Rendano V. T., and Habel R. E.. . 1985. A standardized nomenclature for radiographic projections used in veterinary medicine. Vet. Radiol. 26:2–9. [Google Scholar]
  47. Sripanyakorn S., Jugdaohsingh R., Dissayabutr W., Anderson S. H., Thompson R. P., and Powell J. J.. . 2009. The comparative absorption of silicon from different foods and food supplements. Br. J. Nutr. 102:825–834. doi: 10.1017/S0007114509311757 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. United States Department of Agriculture 2017. Equine 2015 Report 3: Equine management and select equine health conditions, 2015. No. 722.1017. Fort Collins (CO): USDA–APHIS–VS–CEAH–NAHMS. [Google Scholar]

Articles from Journal of Animal Science are provided here courtesy of Oxford University Press

RESOURCES