Reducing cardiovascular disease risk among families with familial hypercholesterolaemia by improving diet and physical activity: a randomised controlled feasibility trial

Objective Familial hypercholesterolaemia (FH) elevates low-density lipoprotein cholesterol (LDL-C) and increases cardiovascular disease (CVD) risk. This study aimed to provide evidence for the feasibility of conducting a randomised controlled trial to evaluate the efficacy of an intervention designed to improve diet and physical activity in families with FH. Design A parallel, randomised, waitlist-controlled, feasibility pilot trial. Setting Three outpatient lipid clinics in the UK. Participants Families that comprised children (aged 10–18 years) and their parent with genetically diagnosed FH. Intervention Families were randomised to either 12-week usual care or intervention. The behavioural change intervention aimed to improve dietary, physical activity and sedentary behaviours. It was delivered to families by dietitians initially via a single face-to-face session and then by four telephone or email follow-up sessions. Outcome measures Feasibility was assessed via measures related to recruitment, retention and intervention fidelity. Postintervention qualitative interviews were conducted to explore intervention acceptability. Behavioural (dietary intake, physical activity and sedentary time) and clinical (blood pressure, body composition and blood lipids) outcomes were collected at baseline and endpoint assessments to evaluate the intervention’s potential benefit. Results Twenty-one families (38% of those approached) were recruited which comprised 22 children and 17 adults with FH, and 97% of families completed the study. The intervention was implemented with high fidelity and the qualitative data revealed it was well accepted. Between-group differences at the endpoint assessment were indicative of the intervention’s potential for improving diet in children and adults. Evidence for potential benefits on physical activity and sedentary behaviours was less apparent. However, the intervention was associated with improvements in several CVD risk factors including LDL-C, with a within-group mean decrease of 8% (children) and 10% (adults). Conclusions The study’s recruitment, retention, acceptability and potential efficacy support the development of a definitive trial, subject to identified refinements. Trial registration number ISRCTN24880714.

1. Dietary intakes: participants recorded their dietary intake on four non-consecutive days (including one weekend day) using Intake24, a validated online 24-hour recall tool. 1,2 Data were downloaded from Intake24 as an Excel file before being cleaned and analysed, in line with Intake24 data processing guidelines. After data quality checks, data not meeting the validity criteria (completion rate <2 minutes; daily energy intake <2000KJ or >16000KJ) were excluded and only participants with two weekdays and one weekend day of valid intake were included in the analysis. Using pivot tables, daily intakes of macronutrients were calculated. Intakes of fruits and vegetables were presented as mean portions consumed per day by dividing the sum intake of fruit and vegetables by 80, per participant and per day. The value of 80 was chosen as a single portion of fruit and vegetables, in U.K. national guidelines, is considered to be 80g. Dried fruit, fruit juices and pulses were included but a portion was considered to be 30g, 150ml and 80g respectively, and fruit juices and pulses were capped at one portion per day. Fruits and vegetables in composite dishes were not included as there is not yet a validated method to calculate this using the Intake24 database. In line with the specific dietary goals of the intervention described in appendix 2, mean daily intakes of fat, saturated fat, monounsaturated fats, polyunsaturated fats, cholesterol, fiber, fruit and vegetable portions, and plant sterol or stanol fortified foods were chosen as outcome measures.
2. Physical activity levels: average daily minutes of free-living moderate and vigorous physical activity (MVPA) was measured by asking participants to wear an ActiGraph GT3X+ accelerometer monitor (ActiGraph, Pensacola, USA) during waking hours for seven consecutive days. Participants also completed logbooks to capture waking/sleeping times and any activities during which the monitor was taken off.
Raw data were downloaded and analysed, in 15 and 60-second epochs for children and adults respectively, using the manufacturer's software (ActiLife software v6.13.4; ActiGraph, Pansacola, FL, USA). Logbook data and an algorithm, which detected non-wear as periods of 60 minutes or more of consecutive zeros with no allowance

Classificatio n of nonwear time
Periods of time that participant is not wearing Actigraph should be detected and deleted from the data to avoid misclassifying non-wear time as sedentary time. There are a variety of available algorithms to detect non-wear time, which differ in the criteria they use.
• An algorithm which detects nonwear as periods of 60 minutes or more of consecutive zeros with no allowance for interruptions • Participants are asked to complete logbooks to capture non-wear occasions and reasons, to cross check results from algorithm • Self-reported PA during non-wear times (e.g. swimming or contact sports) were not included in analysis Valid day & minimum number of days criteria The number of hours of weartime required to define a valid day, and the number of these valid days required for data to be included in the analysis.
• A valid day is defined as ≥ 8 hours (480 minutes) • Four valid days (including one weekend day) were required for inclusion in analysis

Uni-or triaxial counts
The ActiGraph detects activity counts across three axes (vertical, mediolateral and anteroposterior). Researchers can choose to use the vector magnitude (VM) score, which includes activity counts detected across all axes, or use activity counts detected across vertical axis only.
• Activity counts detected across vertical axis were analysed to match the approach used in the studies that developed the MVPA cut-points BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s)

Clinical outcomes (Study objectives 5 and 6)
Clinical outcomes were collected from participants at research contact two (baseline) and three (endpoint) by members of the research team.
1. Anthropometry and body composition: Height was measured using a stadiometer and weight, body fat percentage and fat free mass using bioelectrical impedance scales (Tanita™ MC-780MA). At sites where this equipment was unavailable, weight was measured using medical scales. These outcomes were used to calculate the body mass index (BMI) for adult participants. The LMS method was used to express the BMIs of child participants as BMI Z-scores and percentiles. 7 This was carried out using the lmsGrowth excel add-in 8 and the British 1990 reference population. 9 The 91 st and 98 th BMI percentiles were applied to define overweight and obesity as recommended for clinical interventions. 10 2. Resting arterial blood pressure: The mean of two sphygmomanometer readings (three if 1 st and 2 nd differed by more than 10mmHg) was calculated to provide systolic and diastolic blood pressure outcomes. Participants were seated during the measurements, which were taken after a period of at least five minute of rest to allow values to return to resting levels.
3. HRQoL: Children completed an age appropriate Pediatric Quality of Life Inventory ™ (PedsQL™) Version 4.0 and adults completed an EuroQol Group EQ-5D-3L health questionnaire. 11,12 For children, the completed inventories were scored using the recommended scoring system 13 to produce a score for HRQoL and a breakdown of the component physical and psychosocial (emotional, social and school) functioning scores. For adults, the recommended scoring system was used 12 to produce a five-digit health state profile representative of five dimensions of health which was converted into a single index value using a validated value set created for use in the United Kingdom. 14 The visual analogue scale data was presented as a single value between 1 and 100.
4. Blood lipid profile: 25ml fasted blood samples were collected and a proportion was analysed immediately in the local NHS laboratory. Analysis was done by routine homogeneous enzymatic methods using Cobas reagents (Cholesterol Gen 2; BMJ Publishing Group Limited (BMJ) disclaims all liability and responsibility arising from any reliance Supplemental material placed on this supplemental material which has been supplied by the author(s) BMJ Open doi: 10.1136/bmjopen-2020-044200 :e044200.
Triglyceride Gen 1; HDL Gen 3 or 4) on Roche/Hitachi c701/2 analysers. LDL-C concentration was then calculated using the Friedewald equation. 15 The remaining sample was processed and stored at -80°C as plasma or serum. Subsequent batch metabolomics analysis to determine 201 metabolomics is planned, but not presented in this manuscript due to laboratory closure during the COVID-19 outbreak.