MR278 - Study of Birth Cohort from Hertfordshire

Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.

Objective for processing

The University of Southampton is the Data Controller for The Hertfordshire Cohort [58 words unchanged] the hypothesis that chronic, non-communicable diseases of ageing had their roots in fetal foetal and infant life. 37,000 men and women born in Hertfordshire between 1911 [39 words unchanged] mortality is due, hence the request for further Demographic and mortality data. [1 paragraph unchanged] processing is necessary for archiving purposes in the public interest, scientific or historical research purposes or statistical purposes in accordance with Article 89(1) based on Union or Member State law which shall be proportionate to the aim pursued, respect the essence of the right to data protection and provide for suitable and specific measures to safeguard the fundamental rights and the interests of the data subject. [11 paragraphs unchanged] Research findings based on NHS Digital data will be disseminated through journal articles, online newsletters and conference presentations. To our knowledge, there are no moral or ethical issues raised by this dissemination and there is no risk of harm to the public from this dissemination. This is because the information presented will only include general statements of relationships between various risk factors and health outcomes such as death and hospital admission and aggregated data such as means and medians; any low-frequency counts will be suppressed. The justification for requesting individual data from cohort members is that this is required for the following objectives stated above: to use mortality data to avoid contacting recently deceased cohort members; and to examine relationships between person-level sociodemographic, anthropometric, lifestyle, clinical and genetic characteristics from the cohort in relation to risk of mortality and hospital admission. These objectives cannot be met through the use of aggregated data or data that is not mapped to specific members of the Hertfordshire Cohort Study. These objectives also require the 10 years of NHS Digital data requested for all cohort members, regardless of their geographical location, to enable detailed histories of deaths and hospital admissions until 2020 to be ascertained. There are no alternative, less intrusive ways of achieving these research objectives. To minimise the NHS Digital data required, data already collected from participants of the Hertfordshire Cohort Study will be utilised as much as possible.

Processing activities

[7 paragraphs unchanged] The following steps mitigate risks of reidentification during linkage; only linkage between NHS Digital data and Hertfordshire Cohort Study databases is permitted; only a unique serial number acts as a common field between databases; no personal details, such as names or addresses, are included in any HCS databases used by researchers; researcher training emphasises the importance of participant confidentiality and data security. NHS Digital data can only be accessed at the MRC LEU by a select group of researchers that have been granted access to this data; remote access is not possible. The MRC LEU is a secure environment where data is only visible to researchers who have permission to use it.

Expected output

[9 paragraphs unchanged] • Occasional public meetings, most recently on 23/06/18 in Harpenden Harpenden. Going forward, these are likely to be online. Along with this material presented for cohort members, the MRC unit anticipate that findings will be published in journals specialising in ageing and musculoskeletal health. Examples of these journals include: https://www.springer.com/journal/40520, https://www.springer.com/journal/223, and https://www.journals.elsevier.com/bone. Relevant conferences include the British Society of Rheumatology Annual Conference in Manchester, UK. The Bone Research Society Annual Meeting, Manchester, UK and the World Congress on Osteoporosis, Osteoarthritis and Musculoskeletal Diseases in London. The nature and form of research outputs from the MRC LEU has remained the same for the last 5 years. As the MRC Unit only received NHS Digital data from this DSA in December 2021, no research outputs have been produced. Therefore, the MRC unit are not at the stage where the MRC unit can describe how this research has benefited the health and social care system. In all outputs concerning NHS Digital data, only general statements of relationships between various risk factors and health outcomes such as death and hospital admission and aggregated data such as means and medians will be presented; any low-frequency counts will be suppressed. This will ensure that no individuals can be identified from any research outputs.

Expected measurable benefits

[6 paragraphs unchanged] These benefits stated above remain the same and have not altered since the start of this agreement. As the MRC Unit only received NHS Digital data from this DSA in December 2021, no research outputs have been produced. Therefore, the MRC unit are not at the stage where the MRC unit can describe how this research has benefited the health and social care system.

Changed only in punctuation, spacing or capitalisation: Benefits reported.

Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.

Objective for processing

The University of Southampton are is the Data Controller for The Hertfordshire Cohort Study (HCS), which has been [23 words unchanged] was during the 1980's that the wider cohort was established by the MRC Medical Research Council (MRC) Environmental Epidemiology Unit (a forerunner of the MRC Lifecourse Epidemiology Unit (LEU)) [62 words unchanged] early life influences on mortality is due, hence the request for further MRIS Demographic and mortality data. The legally stated purpose for data processing is to investigate influences on chronic non-communicable diseases of ageing that act across the lifecourse, some of which begin in utero. The legally stated basis for processing is GDPR Article 6(1)(e) ‘the performance of a task carried out in the public interest’, and Article 9(2)(j). These activities can be deemed to be in the public interest as the processing has the potential to benefit the provision of health and social care in England. [2 paragraphs unchanged] The addition of another 10 years’ APC data will add considerable value to the existing database. It will: a) enable description of, and identification of risk factors for, longitudinal patterns of hospital admission between the ages of 60 and 90 years in a contemporary ageing cohort; b) provide sufficient cases of individual diagnoses and procedures to allow the exploration of causal pathways. In addition, since 1/3 of the cohort have died and many more are too frail to be followed-up in person, HES data now provide the only viable source of cohort-wide follow-up prior to death. The extended HCS HES database will be an important publicly funded resource for research. [1 paragraph unchanged] The work is being undertaken to extend a publicly funded resource for research. By comparing observed health outcomes from the NHS Digital data with prospectively measured risk factors from the cohort data, the research aims to elucidate pathways to ill-health and mortality in old age, and the concomitant need for hospital care. Such research is an iterative process, and whilst specific outputs planned are outlined below, it is important that the Hertfordshire Cohort Study (HCS) remains reactive to changing research priorities including those suggested by our Public and Participant Involvement Panel. Any changes to the purpose would however require an amendment to this agreement. Mortality data are used for three purposes: 1) avoid attempting contact with recently deceased cohort members 2) consider later life characteristics alongside those recorded at birth and during infancy as risk factors for mortality 3) account for mortality as a competing risk outcome in the analyses of risk factors for hospital admission in later life. The work is being undertaken to extend a publicly funded resource for research. By comparing observed health outcomes from the NHS Digital data with prospectively measured risk factors from the cohort data, the research aims to elucidate pathways to ill-health and mortality in old age, and the concomitant need for hospital care. Such research is an iterative process, and whilst specific outputs planned are outlined below, it is important that the Hertfordshire Cohort Study (HCS) remains reactive to changing research priorities including those suggested by The University of Southampton's Public and Participant Involvement Panel. Any changes to the purpose would however require an amendment to this agreement. To address the GDPR principle of the University of Southampton have reviewed the data received from NHS Digital and can confirm that it is: adequate, to confirm tracing of the cohort; relevant, to fulfil the stated objectives; and limited, to data storage which is necessary for audit purposes. The University of Southampton is the sole data controller who also processes the data for the purposes described in this agreement. This application is entirely funded from the transitional funding for the MRC Lifecourse Epidemiology Unit at the University of Southampton, documentation of which has been previously supplied.

Processing activities

Tracing and flagging of the wider cohort (n=C37,000) was carried out by [20 words unchanged] records (name, sex, DoB and address at birth). In return, current names, HA Health Authority (HA) ciphers of residence and NHS numbers of flagged individuals were notified to [14 words unchanged] subsequently been reported regularly by NHS Digital under this data sharing agreement. Mortality data are used for three purposes: In 2011, the NHS numbers of 2997 clinic participants originally provided to the Lifecourse Epidemiology Unit (LEU) by the NHS Central Register (NHSCR) were supplied to NHS Digital and an extract of HES data covering the period between the earliest baseline clinic and 31/03/2010 was received in return. HES data were cleaned and episode records collapsed to produce an admissions history for each HCS member from the date of their baseline assessment (which varied from person to person) to the universal end date in 2010. Admission histories (which could potentially be identified by date of birth in association with sex) were created through linkage to the HCS database and have been subjected to ongoing analyses using survival models in STATA (data analysis and statistical software). The HES data can be linked by the LEU with fact and date of death derived from civil registration data as well as with other study data, as explained in the following paragraph 1) avoid attempting contact with recently deceased cohort members Data derived from multiple sources (birth records, home interviews, clinics, postal questionnaires, health authorities, GPs and NHS Digital) are stored in a series of databases in a dedicated secure area of the MRC LEU computer server. A unique serial number acts as the primary key and is the only common field between databases. This allows the data to be linked as necessary to investigate emerging research questions. Any emerging research questions are not permitted to go beyond any significant extension of the purpose without an amendment being submitted and agreed by NHS Digital. 2) consider later life characteristics alongside those recorded at birth and during infancy as risk factors for mortality 3) account for mortality as a competing risk outcome in our analyses of risk factors for hospital admission in later life. In 2011, the NHS numbers of 2997 clinic participants originally provided to the LEU by the NHS Central Register (NHSCR) were supplied to NHS Digital and an extract of HES data covering the period between the earliest baseline clinic and 31/03/2010 was received in return. HES data were cleaned and episode records collapsed to produce an admissions history for each HCS member from the date of their baseline assessment (which varied from person to person) to the universal end date in 2010. Admission histories (which could potentially be identified by date of birth in association with sex) were created through linkage to the HCS database and have been subjected to ongoing analyses using survival models in STATA (data analysis and statistical software). The HES data can be linked by the LEU with fact and date of death derived from civil registration data as well as with other study data, as explained in the following paragraph Data derived from multiple sources (birth records, home interviews, clinics, postal questionnaires, health authorities, GPs, NHS Digital and ONS) are stored in a series of databases in a dedicated secure area of the MRC LEU computer server. A unique serial number acts as the primary key and is the only common field between databases. This allows the data to be linked as necessary to investigate emerging research questions. Any emerging research questions are not permitted to go beyond any significant extension of the purpose without an amendment being submitted and agreed by NHS Digital. [3 paragraphs unchanged] The LEU is located on the campus of the University Hospitals Southampton NHS Trust but is otherwise unrelated to the NHS. [1 paragraph unchanged]

Expected measurable benefits

[2 paragraphs unchanged] Mortality data for the 3000 people followed up in clinic and by questionnaires since 1998 allow us The University of Southampton to 1) avoid attempting contact with recently deceased cohort members, 2) consider [14 words unchanged] for mortality, 3) account for mortality as a competing risk outcome in our the analyses of risk factors for hospital admission in later life. [3 paragraphs unchanged]

Benefits reported

[1 paragraph unchanged] The work has enabled the creation of a publicly funded resource for [65 words unchanged] is important that the Hertfordshire Cohort Study remains reactive to changing research priorities. priorities, including COVID19. Any changes would of course require an amendment to this agreement. HCS has pioneered linkage between cohort and HES data and we The University of Southampton have published methods and technical statistical papers to assist other researchers to do likewise. In addition, timely notification of death has allowed us The University of Southampton to avoid approaching cohort members who have died, an occurrence which is distressing to the bereaved relatives and to the team. [19 paragraphs unchanged] Some of the key findings of the HCS with yielded benefits to health and social care have been expanded below: A lifecourse approach to ageing HCS was among the first cohort studies to show that small size at birth and during infancy are risk factors for chronic non-communicable disease in later life. Further evidence on the importance of lifecourse risk factors for healthy ageing has accumulated over time from HCS, and many other national and international cohort studies. This has led to the embedding of ‘a lifecourse approach to health’ in national and international public health policy and clinical frameworks. Public Health England advocates a lifecourse approach to the prevention of ill health. In their Minsk Declaration, the countries of the European Union state that a lifecourse approach is an investment in health and well-being for all countries and an essential step towards the implementation of the goals and targets in the United Nations 2030 Agenda for Sustainable Development. Globally, a lifecourse approach is central to the World Health Organisation’s Ageing and Health Programme. The WHO states that “factors start to influence the ageing process at an early stage. The environments that people live in as children – or even as developing foetuses – combined with their personal characteristics, have long-term effects on how they age”; HCS has contributed importantly to this position not only through its research findings, but also by the advocacy of its Principal Investigator to WHO strategy development (WHO strategy white paper 2017, Jacob, Baird, Barker, Cooper and Hanson, The Importance of a Life Course Approach to Health: Chronic Disease Risk from Preconception through Adolescence and Adulthood, https://www.who.int/life-course/publications/life-course-approach-to-health.pdf?ua=1). Sarcopenia (the loss of muscle mass, strength and function with ageing) Hertfordshire Cohort Study research in to sarcopenia is embedded within the Lifecourse Epidemiology Unit research program on Musculoskeletal Health in Later Life. HCS research has shown how smaller size at birth, shorter stature in adulthood, socioeconomic disadvantage, low physical activity and poor diet, combine throughout the lifecourse to affect muscle strength (grip strength) and physical function (gait speed) in later life. This research has contributed to international policy on the assessment, definition and population burden of sarcopenia, including the European Working Group definition (2019) and the FNIH/SDOC consensus definition (2019). This is a crucial contribution to policy because only by establishing a consensus definition can the assessment and monitoring of sarcopenia eventually be incorporated in to clinical care of older people. Physical frailty Research in HCS has contributed to the recognition that, set against the backdrop of an ageing population, frailty is a major source of concern for the healthcare systems of developed countries. Research evidence from studies such as HCS has contributed to the development of the following policies. First, the NHS Long Term Plan outlines several important changes to the way the NHS should work to support patients and their carers; improving care for older people living with frailty or multiple long-term conditions and supporting them to age well and stay independent at home for longer is one of its priorities. Second, changes to the GP contract in 2017/18 introduced routine frailty identification for patients who are 65 and over and made the NHS in England the first health system in the world to systematically identify older people who are living with moderate and severe frailty using a population-based stratification approach. It is hoped that early identification coupled with targeted support can help older people living with frailty to stay well and live independently for as long as possible. Lastly, efforts are being made to embed Comprehensive Geriatric Assessment (CGA) in primary care. CGA is a process of care comprising a number of steps. Initially, a multidimensional holistic assessment of an older person considers health and wellbeing and leads to the formulation of a plan to address issues which are of concern to the older person. Interventions are then arranged in support of the plan. Progress is reviewed and the original plan reassessed at appropriate intervals with the interventions reconsidered accordingly. In community settings, the evidence shows that complex interventions in people with frailty can reduce hospital admission and the risk of readmission in those recently discharged, and that individualised care planning can reverse the progression of frailty. A previous Principal Investigator in HCS has been closely involved in advocating for the importance of CGA in primary care. Osteoporosis and fracture risk As described above for sarcopenia, Hertfordshire Cohort Study research in to osteoporosis and fracture risk is embedded within Lifecourse Epidemiology Unit research program on Musculoskeletal Health in Later Life. HCS research has yielded benefit for the health and social care system by contributing to published literature, but also importantly through a member of the Lifecourse Epidemiology Unit’s role as an internationally recognised expert in rheumatology and osteoporosis, and his policy advocacy work in these areas. The HCS member has influenced national and international policy on the epidemiology, clinical risk assessment, prevention and treatment of osteoporosis. Following his role on the Department of Health Advisory Group on Osteoporosis, he produced widely adopted Guidelines for the Management of Osteoporosis for the Royal College of Physicians; contributed to NICE Technology Appraisals and Guidelines for the disorder; and translated these to local use through the UK National Osteoporosis Guideline Group (2015-2019). He has also co-authored European Guidelines for the Management of Osteoporosis (2015 and 2019), which provide a basis for risk assessment and treatment. Under his leadership, several international working groups have been formed, which have produced highly cited international consensus documents on the epidemiology of osteoporosis and secondary fracture prevention. His work with WHO has included the development of guidelines on osteoporosis prevention and treatment.

Unchanged: Expected output.

Text removed is struck through; text added is underlined. Unchanged paragraphs are summarised rather than repeated.

Datasets: + Civil Registrations of Death; + Demographics

Objective for processing

[3 paragraphs unchanged] In addition, the civil registration data that are available for the wider [33 words unchanged] from blood samples), combined with increasing numbers of deaths among clinic participants (700 (1000 approx. to date), provides the potential to examine the combined impact of genotype as well as health sociodemographic, anthropometric, lifestyle, clinical and lifestyle genetic characteristics in early old age on subsequent longevity and cause of death. This [15 words unchanged] because they have 'good genes' or because they have lived 'healthy lives'. The work is being undertaken to provide extend a publicly funded resource for research. By comparing observed health outcomes from [39 words unchanged] and whilst specific outputs planned are outlined below, it is important that The the Hertfordshire Cohort Study (HCS) remains reactive to changing research priorities. priorities including those suggested by our Public and Participant Involvement Panel. Any changes to the purpose would however require an amendment to this agreement.

Processing activities

Tracing and flagging of the wider cohort (n=C37,000) was carried out by [50 words unchanged] of deaths that had already occurred. Incident mortality has subsequently been reported monthly regularly by NHS Digital under this data sharing agreement. Mortality data are used for two three purposes: i) To investigate the associations between conditions in early life, ageing, longevity and cause of death. 1) avoid attempting contact with recently deceased cohort members ii) To avoid unnecessary contact with bereaved relatives of those who are deceased 2) consider later life characteristics alongside those recorded at birth and during infancy as risk factors for mortality 3) account for mortality as a competing risk outcome in our analyses of risk factors for hospital admission in later life. [4 paragraphs unchanged] Processing of NHS Digital data covered by this agreement occurs only within the MRC LEU. The LEU is a part of [21 words unchanged] of Southampton but are limited to those staff based in the LEU. [2 paragraphs unchanged]

Expected output

Over 250 300 publications in peer-reviewed scientific journals have resulted from the Hertfordshire Cohort Study [10 words unchanged] in 3 of these and HES data in 4. Over the next 1-2 5 years the analyses work proposed below will consider mortality or HES data as a specific outcome and, as explained lifecourse predictors of mortality, and morbidity recorded in the processing activities, continuous notification HES data. Outputs will take the form of death will underlie many more outputs. scientific papers published in peer-reviewed journals, with accompanying presentation at national or international conferences, press releases and dissemination via social media. HES data: An exploration of the risk factors for hospital admission in later life has produced 3 papers, all of which are drafted and ready for immediate submission: • Sociodemographic, anthropometric, lifestyle and clinical risk factors for mortality during a 20 year follow-up period among 3,000 community dwelling older men and women in England. • Predictive factors for 30-day readmission among older people • Lifecourse predictors of musculoskeletal admissions to hospital (including clinical fracture, joint arthroplasty, and injurious falls). • Predictive factors for emergency admission among older people • The role of sarcopenia (the loss of muscle mass, strength and function in later life) as a risk factor for hospital admission in later life. • Predictive factors for elective admission among older people • Lifestyle risk factors (smoking, alcohol, diet and physical activity) in later life as predictors of hospital admission during 20 year follow-up. Three further papers using HES data will be written: • Description of longitudinal patterns of hospital admission over a 20 year period and investigation of their predictors in a contemporary ageing cohort. • The contribution of co-morbidity to risk of hospital admission In addition to the outputs listed, material is produced for cohort members. This includes • The contribution of obesity to risk of hospital admission • An annual newsletter, the latest edition of which was mailed in February 2020 • The contribution of socio-economic status to risk of hospital admission • A website https://www.mrc.soton.ac.uk/herts/ and Twitter feed @MRC_LEU #HertsCohort In addition to the outputs listed, which are aimed at the scientific community, material is produced for cohort members. This includes • An annual newsletter, the latest edition of which will be mailed in February 2019 • A website https://www.mrc.soton.ac.uk/herts/ [1 paragraph unchanged]

Expected measurable benefits

The research described is focused on accruing evidence on risk factors for [37 words unchanged] intervention. For example, if the work on risk factors for hospital admission (described above) identifies a characteristic pattern of morbidity and behaviour that is associated with [69 words unchanged] intervention within clinical practice with the involvement of all necessary care-giving stakeholders. Mortality: The wider study birth cohort of 37,000 men and women born in Hertfordshire between 1911-39 aims to inform a body of evidence about how intrauterine (occurring within [91 words unchanged] evidence, an intervention in routine obstetric care could follow in the future. HES: Research using HES data has thus far addressed risk factors for hospitalisation generally, although the research is not principally service based. The aim is not to identify specific procedures, interventions and policies that will have immediate benefit for the health of older people and the care that they receive from the NHS but rather, to inform a body of evidence about likely risk factors for hospitalisation. In time, this could lead to interventions as described above. Mortality data for the 3000 people followed up in clinic and by questionnaires since 1998 allow us to 1) avoid attempting contact with recently deceased cohort members, 2) consider later life characteristics alongside those recorded at birth and during infancy as risk factors for mortality, 3) account for mortality as a competing risk outcome in our analyses of risk factors for hospital admission in later life. Nutrition programme: There is a very compelling case for knowing more about nutrition and health in older age. One third of older patients are categorised as being at risk of malnutrition on admission to hospital, and malnutrition is associated with poorer health outcomes and longer hospital stay. HCS is the only UK cohort that has collected dietary data in later life and offers the opportunity to define the role of nutrition as a determinant of health in later life. The MRC LEU are developing a programme that will run till 2020, using both HES and mortality data to chart the development of disease in relation to diet. HES: Research using HES data from HCS baseline to March 2010 addressed risk factors for hospitalisation, although the research is not principally service based. The aim is not to identify specific procedures, interventions and policies that will have immediate benefit for the health of older people and the care that they receive from the NHS but rather, to inform a body of evidence about likely risk factors for hospitalisation. In time, this could lead to interventions as described above. The addition of another 10 years’ APC data will add considerable value to the existing database. It will: a) enable description of, and identification of risk factors for, longitudinal patterns of hospital admission between the ages of 60 and 90 years in a contemporary ageing cohort; b) provide sufficient cases of individual diagnoses and procedures to allow the exploration of causal pathways. In addition, since 1/3 of the cohort have died and many more are too frail to be followed-up in person, HES data now provide the only viable source of cohort-wide follow-up prior to death. The extended HCS HES database will be an important publicly funded resource for research. The data remain as relevant today as when the work began, and the ongoing accumulation of hospital admissions and deaths among the wider cohort will only add to their value.

Benefits reported

Early findings from Hertfordshire provided the catalyst for a new field of [35 words unchanged] Weight in infancy and death from ischaemic heart disease. Lancet. 1989;2:577-80) has nearly more than 2000 citations to date. Over 260 papers have been published in peer-reviewed journals. Since 2004 HCS has focused on musculoskeletal disease. Key findings are summarised below: The work has enabled the creation of a publicly funded resource for research which details hospital admissions in a contemporary ageing cohort to 2010. By comparing observed health outcomes from NHS Digital data with prospectively measured risk factors from the cohort data, the research has elucidated pathways to ill health and mortality in old age, and the concomitant need for hospital care. Such research is an iterative process and whilst specific outputs have been proposed above, it is important that the Hertfordshire Cohort Study remains reactive to changing research priorities. Any changes would of course require an amendment to this agreement. Growth in utero and in infancy are determinants of adult bone mass, bone geometry, microarchitecture, strength and fracture risk HCS has pioneered linkage between cohort and HES data and we have published methods and technical statistical papers to assist other researchers to do likewise. In addition, timely notification of death has allowed us to avoid approaching cohort members who have died, an occurrence which is distressing to the bereaved relatives and to the team. Adult bone mass is related to circulating GH and cortisol concentration; SNPs within candidate genes; ischaemic heart disease and risk of diabetes Over 300 papers have been published in peer-reviewed journals. Since 2004 HCS has focused on musculoskeletal disease. Key findings are summarised below: Fracture risk is predicted by measures of bone density and architecture other than dual-energy X-ray absorptiometry • Muscle size and strength are related to growth in early life and are risk factors for hospital admission in early old age Markers of bone turnover, but not volumetric bone density, are associated with knee osteoarthritis • HCS data have contributed to the development of normative guidelines on grip strength Genetic factors and vitamin D status influence the incidence and progression of knee osteoarthritis • Sarcopenia and frailty are prevalent conditions in older men and women, newly recognised in clinical practice Functional limitation associated with hand OA is driven by pain rather than by comorbidity • Sarcopenia in later life is associated with altered muscle morphology Adult bone mass and grip strength are related to quality of life • Unhealthy lifestyles (obesity, smoking, poor diet, physical inactivity) are strongly linked to poor physical function and increased risk of hospital admission in early old age. Bone health is directly influenced by muscle function • Growth in utero and in infancy are determinants of adult bone mass, bone geometry, microarchitecture, strength and fracture risk Early life factors are related to body composition in late adulthood • Bone health is directly influenced by muscle function Muscle size • Adult bone mass and grip strength are related to growth in early quality of life Sarcopenia and frailty are prevalent conditions in older men and women, newly recognised in clinical practice • Adult bone mass is related to circulating GH and cortisol concentration; SNPs within candidate genes; ischaemic heart disease and risk of diabetes Sarcopenia in later life is associated with altered muscle morphology • Fracture risk is predicted by measures of bone density and architecture other than dual-energy X-ray absorptiometry HCS data have contributed to the development of normative guidelines on grip strength • Markers of bone turnover, but not volumetric bone density, are associated with knee osteoarthritis Early environment influences development of autoantibodies • Genetic factors and vitamin D status influence the incidence and progression of knee osteoarthritis Inflammaging is a powerful predictor of future frailty • Functional limitation associated with hand OA is driven by pain rather than by comorbidity Adult anthropometry is associated with several candidate genes • Inflammaging is a powerful predictor of future frailty Birth weight and adult fat consumption interact to determine serum cholesterol levels • Early environment influences development of autoantibodies Unhealthy lifestyles (obesity, smoking, poor diet, physical inactivity) are strongly linked to poor physical function and increased risk of hospital admission in older men and women. • Early life factors are related to body composition in late adulthood The data remain as relevant today as when the work begun, and the growing percentage of deaths among the wider cohort only add to their value. • Adult body composition is associated with several candidate genes In addition, notification of death is crucial for individuals with whom the study is in contact: given their age, death rates are high, but having received no mortality data for over two years sometimes unknowingly the team approach a cohort member who has died. This is distressing to the bereaved relatives and to the team and could be avoided by reinstating the flow of mortality data. • Birth weight and adult fat consumption interact to determine serum cholesterol levels