1. Collection of height, weight and body fat.

As well as reflecting the substantive comments from earlier discussions this paper in setting out the main details of the enhancement, this paper outlines pre-and post- March 2015 costs, in order to enable a profiling of spend. In the original proposal all staffing costs were censored at March 2015, given that was the limit of the spend period. However, this was not in expectation that all the work would be completed by that point, but a reflection of the limits of the funding source. It was anticipated that the extension of staffing costs would be found by some other means (e.g. through the core CLS funding, or through a separate funding proposal).

This paper instead builds in staffing costs that are appropriate for the enhancement even when they extend beyond March 2015. However, for consistency with the earlier paper, the post-March 2015 totals are also shown excluding these post-March 2015 staffing costs.

Key elements of enhancement

The continued collection of height and weight will allow a robust understanding of contemporary children's physical and developmental trajectories over time across the UK and how this physical development relates to family context, child and family practices and behaviours, neighbourhood and environment, and other aspects of children's development, and wellbeing.

Height will be measured by trained and accredited interviewers using a height measure that is appropriate for teenagers and using best practice protocols. As usual we will check what the most up-to-date protocols are and ensure that the MCS team and the fieldwork agency are fully versed in them. This will also maximise comparability with measurement carried out for example on the National Child Measurement Programme (NCMP) using the same protocols.

Weight and body fat will be measured by trained interviewers using Tanita scales that capture both weight and body fat percentage.

Following full briefing, interviewers will practice the height, weight and body fat measurement on young people of the relevant age, as part of their training, and will be require to be accredited in height measurement before they can implement it in the field.

All measures will be captured in CAPI, alongside a record of any specific circumstances relating to the measurement that might have affected its accuracy. These data will be returned to CLS by the contracted fieldwork agency within four months of the conclusion of fieldwork.

The existing data manager team will check and clean the data, create appropriate derived variables and prepare it for release within six months of receipt from the fieldwork agency.

The data manager team will be able to provide value added to the physical measurements data by linking the height weight and body fat measurements to updated growth charts constructed according to standard methodology but based on contemporary growth patterns, using the National Child Measurement Programme (NCMP) data. We are currently in discussion with the Department of Health about access to measures collected in the NCMP but not currently deposited (specifically ethnicity). Growth charts based on national NCMP data and incorporating ethnic differences would be a more appropriate reference population than currently used growth charts. We are thus seeking to develop such contemporary growth charts employing the methodology used in the 1990 charts but based on a contemporary, nationally representative and multi-ethnic population.

2. Collection of saliva for DNA extraction.

This element represents a collaboration between Dr Sue Ring of Bristol University and the MCS team at CLS. The collection of saliva samples would enable a genetic record of both the child and their parents to be preserved. This would provide enormous potential for both genetic and epigenetic research on a young, nationally representative cohort and enhance the multidisciplinary, bio-social potential of the study both now and in the future.

There is already great interest in the potential of a DNA resource derived from MCS. Given the wealth of existing data collected in MCS from both survey questions and direct measures and assessments, there is huge potential to understand associations between genetic make-up and a whole series of outcomes relating to e.g. learning, dyslexia, obesity and overweight, behaviour and mental health.

The large sample size and national coverage of MCS are particularly beneficial: as well as offering detailed family-level, individual and contextual information, MCS enables analysis to explore the contexts of varying environmental exposures as well as genetic dispositions. In particular, the value of the oversampling of those from more disadvantaged socio-economic backgrounds is regarded as being particularly valuable in

There is clearly great potential from the collection of triads of samples (both biological parents and the cohort child) for the investigation of epigenetic processes. It is increasingly recognised that non-target tissue sources of DNA can be informative about the relationship between epigenetic variation and phenotype.

It is clear that, while attention would need to be given to ensuring the resource was both known about and used (utilising both the Access Committee and Closer in doing this), there would be a number of scientists ready to build proposals to exploit the data once they became available. For example, researchers who have already identified genetic data from MCS as offering enormous potential include, Silvia Parracchini who has expressed interest in utilising MCS to enhance her existing studies of genetic variants associated with dyslexia; and Caroline Relton, who is working on epigenetics using non-target tissues, and is developing an approach which places DNA methylation changes as in intermediate phenotype on a pathway between exposure and outcome, including such outcomes as obesity, and cognitive and behavioural outcomes.

With ensuing genome-wide genotyping, an MCS DNA back can be expected to provide an extremely valuable resource for analysis that is likely to be very well used. The success of the NCDS resource is testament to the volume of research that such a resource can invite, and we would expect certain of the key features of MCS, such as its demographic coverage, to enhance its utility.

In terms of the planned provision to result from this enhancement; the key element is the collection of saliva, from both the cohort member and from their co-resident natural parents, where applicable.

Such collection was successfully piloted at age 11, but we would expect that the change in age and the context of different instruments would require some additional pretesting. We would then implement in pilot, dress-rehearsal and main stage. We would also need to ensure proper briefing of interviewers and the available of appropriate subsidiary materials such as clear question and answer guidance.

The second key element is the preparation and storage of the samples, which would be undertaken in Bristol using a similar approach and protocols as were developed for the Age 11 pilot in relation to delivering the samples from the field, conduction quantification and extraction. Clearly in the main stage, this would take place on a much larger scale and would require adequate systems being fully in place for that

The third key element is the genotyping of the samples. This would also be undertaken by Bristol. While indicative support has been indicated for this element of the proposal, it would need to be aligned with the ESRC biosocial strategy which is expected later in 2013. Indicative costs for the genotyping would also need to cover staff to support accessing the samples and ongoing storage and maintenance.

Access to the sample would also need to be ensured through adequate resourcing of the Access Committee to support their existing work. The existing access arrangements for NCDS would clearly be suitable for accessing the genotyped data from MCS; and the learning that is being developed in relation to 'accidental' findings would also need to be built into procedures for MCS, as well as informing the consent process itself.

Substantive proposals for analysis would be subject to separate applications to relevant funding organisations but would be enabled and facilitated by this support.

3. Collection of physical activity data.

There is a great deal of interest in understand the patterns of physical activity, including sedentary behaviour, and how these relate to young people's growth patterns, including overweight, as well as to other aspects of their wellbeing and to their long-term outcomes and other.Self-report measures need to be extensive and detail to capture the range of patterns of physical activity and inactivity, and without a complex battery of questions can only provide limited insight into young people's activity patterns. Objective measures can be more convincingly obtained through direct measurement using accelerometers. Direct measures of physical activity can give a much better insight into patterns of moderate and intense activity, and into sedentary behaviour than report measures.

Accelerometers has been successfully collected not only in health-related surveys but also in MCS itself at sweep 4 (the age 7 sweep). Best practice in collection is also improving. At MCS4, 7-day wear was recommended. However, this was not only hard to achieve for many of the children, it is becoming accepted that two days of complete wear may be sufficient to understand activity patterns. Collection of data can contribute to ongoing debates about optimal forms and intensities of activity in relation to health and wellbeing. Combined with the planned age 14 diary, physical activity measures themselves can be enhanced by understanding the context of the activity as well as periods of non-wear.

There is substantial appetite within the scientific and policy community for utilising direct measures of physical activity and for repeating the measurement of physical activity at the critical age of age 14.

Thus the key element of this proposal is to implement collection of two full days of physical activity monitoring of MCS cohort members, using age appropriate and high quality accelerometers.

The proposed collection of physical activity using accelerometers would involve the young people being asked to consent to wear the accelerometer for 2 days, days that coincided with the days they completed a time diary. Parental consent would also be sought prior to the young person being approached.

There are two models that could be implemented. At the age 7 survey, those families that consented to wear the accelerometer were then posted from an external unit the accelerometer, and that unit, based at the Institute of Child Health (ICH) also dealt with follow-up, sending feedback and reminders to return. The ICH unit was responsible for charging the accelerometers, setting them in advance of posting them out, and downloading the data from them on return, before sending them out to the next batch of respondents. The first model would follow this approach but rather than an external unit, the fieldwork agency would themselves log the consenting cases, and send out the equipment, with a pre-paid envelope for return, download data on return and reset the devices for the next batch of consenting participants. On this model we would anticipate a 90 per cent consent rate and that up to 70 per cent of cohort families would return the accelerometers resulting in a gradual depletion of the stock.

A variant on this model (Model 1a) would be to integrate the enhancement fully into the MCS6 fieldwork. Interviewers would supply the accelerometers at the point of the survey having guided the cohort members on use, setting and administration. The interviewers would then be responsible for ensuring the devises were charged and set, and would need to be thoroughly briefed on this. They could then leave the accelerometers to be returned by post, as with the original version, and the fieldwork agency would need to ensure that reminders were sent in a timely fashion to maximise return. This would be more intuitive approach, in that distributing and implementing while in the household would accord with the general practice of the survey and enable questions to be dealt with on the spot, as well, potentially as encouraging wear. However, there would be an additional layer of complexity introduced as a result of the fact that the interviewers would need to be supplied on an ongoing basis with sufficient numbers to carry out the fieldwork, while the accelerometers were being returned, and data downloaded from the respondents. To ensure a smooth process would probably require a larger number of accelerometers than the total used at MCS4. The consent and return rates overall would be expected to be similar, though we could anticipate that a larger number would actually wear the devices prior to return increasing the overall data collection, and that thus the level of productive cases would be higher.

A second model would involve the interviewers not only administering but also collecting the accelerometers after the week containing the two 'wear days' had passed, and while they were in the area. This would increase fieldwork costs, possibly substantially and would require the interviewers themselves to be trained in downloading the data before resetting for subsequent participants or that they returned them in batches to the central office; but would increase the number of returned accelerometers, and thus reduce the total number needed / lost. It would obviate the need for requests for return, except in those cases where the interviewer was unable to follow up.

There are clearly advantages and costs associated with each of the approaches, and the tendering fieldwork agencies will propose and cost what they think is the optimal model. At present we judge this is likely to be closest to Model 1a, and we have estimated costs on that basis. For this we have assumed around 4000 accelerometers will be needed, rather than the 3000+ used in the age 7 survey, since at Age 11 the total stock was almost totally depleted by the end and left no room for flexibility in distribution.

As well as collecting the data, since these are complex data, to provide suitable outputs for users there would be the need to construct age-appropriate derived measures of activity and sedentary behaviour from the data, as was achieved for MCS4. This would be the key data deposit from this element of the data collection, though the 'raw' data would also be deposited for those who wished to work with it directly.

The derived variables would be constructed by an experienced data manager working with a statistician to produce the derived variables useable by the wider research community.