C ONCLUSIONES

Bone health can be assessed in several ways, depending upon the particular aspect of interest. As described above, the diagnosis of osteoporosis using the WHO definition requires assessment of bone mineral status using DXA, although there are other bone densitometry techniques also used to assess status. These include (with earliest first):

 Single Energy X-Ray absorptiometry (SXA)

 Single photon absorptiometry (SPA)

 Dual photon absorptiometry (DPA)

 Quantitative ultrasound (QUS)

 Dual energy x-ray absorptiometry (DXA)

 Quantitative computed tomography (QCT)

 Peripheral quantitative computed tomography (pQCT)

 High-resolution pQCT (HR-pQCT)

The earliest instruments used to assess bone density were SXA, S/DPA; however, since the development of DXA over 30 years ago, they are no longer used. Quantitative ultrasound (QUS) is less expensive and portable instrument used at peripheral sites (e.g. Os calcis, radius, and tibia).

It does not directly measure bone density, and cannot be used to diagnose osteoporosis; however, it is thought that the technique captures other parameters of bone structure and predicts individuals who later fracture. The machine provides a T-score, which correlates to the risk of fracture, however the measurement most commonly of the heel bone (Os calcis) is not a site associated with osteoporotic fracture.

DXA is now the most common bone-imaging technique used to assess bone mineral status. It provides measurements at both axial and appendicular skeletal sites. There are many advantages to DXA, including good precision, low radiation dose and its ability to measure sites prone to osteoporotic fracture e.g. hip and spine.

pQCT and HR-pQCT are more recent methods developed to assess the appendicular skeleton.

They can be used to quantify bone mineral status as well as provide additional parameters related to bone strength (shape and dimensions). Importantly, they provide true volumetric density of bone mineral; a limitation of DXA derived areal density. HR-pQCT also provides information on trabecular and cortical microarchitecture. Advantages and limitations of DXA and pQCT techniques are described in Table 3:1, further details of DXA used in the work described in this thesis are included in the Methods chapter, Sections 6.1.2. Details of pQCT are included in Appendix F, as the data were collected in the overall study protocol, but have not been presented in this thesis as cross-calibration data were required to ensure validity.

75 | P a g e Table 3:1 Advantages and limitations of DXA and pQCT

Densitometry

technique Advantages Limitations

DXA Non-invasive Size dependent, provides areal density (g/cm2) not true volumetric density, cannot measure tissue thickness

Quick scan time (approx. 10mins

for total body) Cannot differentiate a thin, high density bone from thick, low density bone

Fast scan time Affected by changes in body composition Good precision Not mobile and must be kept in a stable

environment

‘Gold standard’ clinically Unable to distinguish trabecular and cortical bone compartments

Low radiation dose Degenerative changes, aortic calcification and vertebral fractures can increase aBMD and bias results, particularly in older people

Assesses clinically relevant skeletal sites e.g. hip and lumbar

spine Limited when comparing different populations

Estimates body composition: lean and fat mass

Fan beam magnification can affect growing skeleton of children and young adults and effect estimated geometrical parameters e.g. hip axis length

Reference data available for Caucasian, Asian and

African-American No reference data for black African populations

pQCT Non invasive Sensitive to movement

Low radiation dose Only measures peripheral skeleton Able to distinguish trabecular

and cortical bone Longer scan time: positioning and adjustments increase scan time

Measures actual density (g/cm3) Software less user friendly Size-independent – able to

consider size and density

Significant variation in trabecular bone in the scan area, means slight shifts in positioning of ROI can alter BMD

Measures body composition: fat and muscle

Acquisition and analysis protocols not standardised, differ in % sites, difficult to compare across studies

Measures geometry e.g. CSA Unable to assess microstructure of bone Measures bone strength,

(IAEA, 2010) and (Langton and Njeh, 2003)

3.4.1 Assessment of fracture risk: FRAX

More recent consensus among bone health researchers has resulted in a definition of bone health that centres on the importance of considering ‘optimal bone strength’, of which bone mineral content (BMC) is but one component (Ward, 2011, Bouxsein, 2005). This shift resulted in the development of the Fracture Risk Assessment Tool (FRAX^®), by the University of Sheffield and WHO, and it is now under the auspices of International Osteoporosis Foundation (IOF). FRAX^® estimates fracture risk of patients and combines DXA derived BMD at the femoral neck with other risk factors. Data from population-based cohort studies in Europe, North America, Asia, and Australia informed the development of the FRAX algorithm. Users are given a 10-year probability of hip fracture specifically, or a major osteoporotic fracture at any site. FRAX has not yet been validated for use in sub Saharan Africa; see Section 2.5.3 for more details of the ‘SAMSON’

network and developments in this area. Figure 3.4 highlights the FRAX risk factors and other lifestyle characteristics that are associated with an increased risk of osteoporotic fracture. Several other algorithms have been developed in addition to FRAX, these include the Garvan and QFracture (ISCD, 2015).

Figure 3:4 FRAX tool web interface for UK

Screenshot of FRAX webinterface removed for copyright reasons. Copyright holder is Centre for Metabolic Bone Diseases, University of Sheffield, UK.

77 | P a g e

3.4.2 Limitations of current osteoporosis definition

Based on the current WHO definition, available data indicate that the incidence and prevalence of osteoporosis varies worldwide, with apparent geographic and ethnic disparity (Cauley et al., 2014). In 2004, Prentice (2004) published a comprehensive review outlining the evidence for diet and nutrition relating to osteoporosis. The review highlights the challenge of accurately determining the extent of osteoporosis worldwide, due to problems with definition and diagnosis.

The requirement of specialist bone imaging equipment to diagnose osteoporosis, along with lack of expertise and trained operators, has resulted in very little data in Africa and many parts of Asia.

As previously mentioned, DXA is the most common bone-imaging technique used to assess bone mineral status; however, while other techniques have been developed, for example Qualitative Ultrasound (QUS) and Quantitative Computed Tomography (QCT), they cannot be used with the WHO definition of osteoporosis. Furthermore, the lack of population reference data for sub-Saharan Africa and specifically black African populations has made it impossible to determine the prevalence of osteoporosis in these regions based on this WHO definition. It is also important to clarify, that black African populations of sub-Saharan Africa differ significantly from African-Americans included in the USA reference data sets (Pettifor, 2015). Additionally, the WHO definition is not valid when applied to men or pre-menopausal women.

Unfortunately, focussing on a quantitative measurement of bone mineral can lead to an oversimplified understanding of the relationship between bone health, defined only in terms of bone mass, and fracture risk. This narrow definition disregards other qualitative components including bone shape and size, internal structure and metabolism, and loading conditions associated with bone strength (Prentice, 2004).

The use of BMC and BMD to compare populations has also been criticised, as both parameters are strongly influenced by body size, a factor that also differs between populations (Prentice, 2004).

Evidence in more recent years has highlighted that not all populations with low bone mass are at high risk of fragility fracture, for example populations with shorter stature such as in Africa or

Asia, will also have lower bone mineral content. As a working definition within a population, it has been suggested that the WHO definition of osteoporosis is of some use, however, it is inadequate to compare across populations.

For this reason, fracture rate in older people has been advocated as most useful when comparing between populations. Hip fracture incidence is preferred, as many countries have hip fracture registers, which enable estimation of incidence. Other osteoporotic fractures, especially of the spine or wrist, may be asymptomatic and medical attention may not be sought. However, other challenges remain, including defining a low trauma, osteoporotic fracture and ensuring exclusion of traumatic fracture. Most challenging is the paucity of quantitative data from LMICs, and its potential unreliability, given the limited access to medical facilities, difficulty in ascertaining exact age and cause of fracture.

3.4.3 Assessment of bone turnover: biochemical markers of bone turnover and calcium homeostasis

To understand the rate of bone remodelling, biochemical markers can be measured in either blood or urine. Markers of bone turnover reflect bone resorption (osteoclast activity) and bone formation (osteoblast activity), but are not currently used to assess or diagnose osteoporosis.

There are various markers available and they tend to be enzymes or proteins synthesised during bone remodelling, or breakdown products (e.g. N- and C- terminal telopeptides of collagen).

There are several markers of bone resorption available these include CTX (C-terminal telopeptide) and NTX (N-terminal telopeptide) of type I collagen, measured in urine or serum/plasma. A further example is tartrate-resistant acid phosphatase (TRACP 5b), an enzyme highly expressed by osteoclasts.

Markers of bone formation measured in plasma or serum include osteocalcin, bone alkaline phosphatase (BALP), and C and N-propeptide of type I collagen (PICP and PINP). Osteocalcin is a protein derived from osteoblasts (the cells responsible for bone formation), secreted in significant quantities in bone tissue (Bonjour et al., 2014).

79 | P a g e In addition to bone turnover markers, the measurement of parathyroid hormone (PTH), metabolites of vitamin D, and growth hormones IGF-1 and -2 can also provide further insight into bone metabolism, calcium homeostasis, and factors which modulate growth. However, there are several factors, which can influence the concentration of these markers including, age, skeletal and sexual maturity, growth velocity, and mineral accrual, as well as sex and ethnicity. Additional sources of variability include season, diet, exercise, kidney function, and in women, the phase of menstrual cycle and the use of oral/injectable hormonal contraceptives. Therefore, it is important that these factors are considered in the study design and adjusted for in statistical models. These make the interpretation of results more difficult, and the implication is that such biochemical markers do not translate directly to amounts of bone formed or loss, and therefore net bone balance. Markers of bone turnover are therefore best measured along with bone mineral status in order to provide a better understanding of mechanisms and overall bone health (Prentice et al., 2006).