Describing habitual levels of physical activity (PA) in older people in terms of impact loads and how this relates to bone and other systems

Context
Accelerometers have been widely used to record physical activity (PA) in studies evalauting effects of exercise on human health. Traditionally, these devices provide a measure related to energy consumption which is ideally suited to studying obesity-related outcomes. More recently, accelerometers have been developed which provide information about exposure to forces of different magnitudes. This may represent a more accurate approach to evaluating PA in the case of physiological systems such as the skeleton, which responds to brief high impacts as opposed to sustained loading.

Aims
1. To characterise the extent to which older people are exposed to impact loading of varying magnitudes.
2. To determine whether higher levels of impact loads resulting from everyday PA in older people are related to properties of bone, such as its density and size, whereas little relationship is observed with lower levels of impact.
3. To identify lifestyle, social, psychological and health related factors which affect the exposure of older people to impact loads of different magnitudes.
4. To characterise how relationships between PA and other physiological systems differ according to level of impact load.
5. To explore barriers to participation of older people in PA associated with higher levels of impact.

Methods
Setting: Relationships between PA and bone will be studied in two well characterised population cohorts of older people, namely the MRC National Survey for Health and Development (NSHD; age 68) and the Hertfordshire Cohort Study (HCS; age 74-82) (n=2000 under active follow up in both cohorts). A cohort of 500 elite athletes over 65 will also be recruited.
Study design: Participants from the three cohorts will be asked to wear an accelerometer for seven days, and information about exposure to different levels of impact subsequently extracted using methods developed in our previous studies. A questionnaire asking about factors related to PA will be completed at the same time. Bone measures will include bone density at the hip as measured by a widely used low dose radiation method (DXA), and more detailed measurements of bone size and shape derived from cross sectional images of the lower leg using a different low radiation imaging device (pQCT).
Analysis: PA data will be analysed in relation to bone parameters (i) currently being measured in a subgroup of HCS, and (ii) due to be measured in a subgroup of NSHD and elite athletes as part of this project. We also aim to extrapolate information about exposure to impacts from a previous NSHD data collection involving the whole cohort, relating this to hip bone density obtained at the same time. NSHD and HCS provide detailed information for other systems (eg muscle, obesity, metabolic, cardiovascular, respiratory, cognitive), which will also be analysed in relation to different impacts.

Potential application
Osteoporosis is a common condition predominantly affecting older people, in which weakening of bone leading to an increased risk of fracture. Since decreased PA is thought to contribute to this process, there is a strong need to develop PA interventions to preserve bone strength and reduce fracture risk in older people, particularly in light of emerging safety concerns over conventional pharmacological approaches. However, if successful PA interventions are to be developed for preserving bone, the relationship between PA and bone strength needs to be clearly understood, particularly if only impacts of a certain magnitude are effective in maintaining bone. Likewise, the types of activities which produce desired levels of impact, and the barriers to carrying out these activities, need to be understood if interventions are to be developed which bring about sustained alterations in behaviour required to prevent bone loss in older people.

We will use a combination of data previously or currently being collected in NSHD and HCS, future data collections funded separately, and data funded via this project comprising (i) accelerometers offered to all participants including a further cohort of Master Athletes, and (ii) and a bone study performed in a subgroup of NSHD and Master Athletes.

Objective 1. Describe habitual PA according to level of impact load in older people
The GCDC X16-1c tri-axial accelerometer will be posted out to participants, worn for 7 days, and number of counts derived within 0.3-0.9g (low eg normal walking), 1-1.5g (medium eg brisk walking), 1.6-2g (high eg aerobics class activities), >2g (very high eg jogging).

Objective 2. Associations between impact loads and bone in older people
Linear regression (adjusted for confounders - see objective 3) will be used to analyse counts within different impact bands in relation to: DXA and tibial pQCT as measured in the bone study; DXA, tibial pQCT and HRpQCT currently being measured in a subgroup of HCS; DXA obtained in the same HCS subgroup 12 months after accelerometry. Additionally, impact data will be extrapolated from previous from Actiheart data, and analysed in relation to hip BMD.

Objective 3: Determinants of exposure to different impact loads
Exposure to different impacts will be related to information about social, economic and psychological influences on PA from a questionnaire sent out with accelerometers, and information on physical performance and other health related factors obtained from face to face assessments.

Objective 4: Associations between impacts and other systems in older people
Exposure to different impacts will be related to variables relating to fat, muscle, cardiovascular, respiratory, and cognition.

Objective 5: Acceptability, barriers and facilitators of activities of different impacts
30 individuals will be purposively selected from the bone study sample and interviewed face to face.

Public: This research is anticipated to improve the health of older people. For example, those at increased risk of osteoporotic fracture will benefit from the greater understanding of relationships between PA impacts and bone provided by this research, since this information is expected to lead to the development of more effective PA interventions for preserving bone and reducing fracture risk. These benefits are expected to be realised over the relatively short term (ie five years following completion of the grant). Over the longer term, hypotheses concerning relationships between PA and other physiological systems arising from this project, may translate into PA interventions intended to reduce adverse effects of ageing in other patient groups.

Public sector organisations: In the short term, knowledge gained about barriers to participation in higher impact PA will help to inform public policies in the UK eg by promoting the provision of supervised aerobics classes targeted to older people, and improving the availability of exercise friendly neighbourhood spaces. This project will also benefit the public sector by identifying cultural and socio-economic influences on PA which public services need to overcome if PA participation is to be taken up widely. Other beneficiaries include agencies such as NICE, whose recommendations for PA will be modified to ensure these are optimised for preventing osteoporosis. In the longer term, knowledge gained from the research may impact advice and recommendations on PA from international organisations, such as the WHO.

Third Sector: In the short term, this project is expected to benefit patient-based charities such as the National Osteoporosis Society, and those involved in older people such as AgeUK. These charities are expected to play an important role in communicating recommendations arising from this project to the wider public through publicity campaigns, and production of information leaflets aimed at patients found to be at increased risk of osteoporosis, such as those attending hospital after sustaining a low trauma fracture. In the longer term, this research is expected to benefit international charities such as the International Osteoporosis Foundation. Charities funding research in this area, such as The British Society for Research into Ageing, and Arthritis Research UK, may also benefit in the longer term by providing a basis for targeting future research of PA interventions. Depending on the wider clinical applicability of our findings, in the longer term, charities representing other patient groups may also benefit, such as those representing patients with heart disease and dementia.

Business/Industry: This research may also provide opportunities for improving wealth, since the novel approach to PA developed through this project may attract investment from overseas charitable and commercial organisations, to support the development of novel PA interventions and accelerometer device technology. Over the short term, this research will benefit manufacturers of accelerometer devices by identifying further outputs that can be derived from raw accelerometry data, which are likely to be clinically useful, such as number of counts within pre-specified impact bands, and acceleration slopes. Our findings will provide an incentive for developing software which automatically derives these novel variables, thereby increasing the applicability of these devices. It may also be possible to develop versions of accelerometers aimed at older people, as a means of improving adherence to PA regimens, based on attainment of a target number of impacts within a pre-specified acceleration band, as exemplified by the Newtest device aimed at premenopausal women .