ALIFE@Work is a randomised controlled trial with a two-year follow-up. Randomisation takes place to three groups: to a control group (control) and to an intervention programme either counselled by phone (phone), or counselled by email (internet). The recruitment and data collection for this study started in January 2004. Data collection will continue until September 2006.
The study design, procedures and informed consent form were approved by the Medical Ethics Committee of the VU University Medical Center (under number 03/193), and all participants provided written informed consent.
Participants are employees of seven different companies in the Netherlands, including two IT-companies, two hospitals, an insurance company, the head office of a bank and a police force. Of the approximately 21,000 employees working at these companies about 40%, i.e. 8,400, were estimated to be overweight. All 21,000 employees were approached per company, through an invitational letter and a screening questionnaire. This was done in six months time, with an interval of three weeks to one month between the companies. The screening questionnaire included demographic questions and questions on body weight and body height.
Inclusion & exclusion criteria
Inclusion criteria of the study were: 1) BMI ≥ 25 kg/m2, 2) paid employment of at least 8 hours a week, 3) adequate knowledge of the Dutch language, 4) access to internet and skilled in using it, 5) at least 18 years of age. Employees were excluded for the following reasons: pregnancy, diagnosis or treatment of cancer, any other disorder that makes physical activity impossible.
A power calculation was carried out for the main outcome variable, i.e. weight change. Calculations were made for a comparison between two equal size groups. Power calculation for weight change was carried out for three levels of power, namely 80%, 85% and 90%. The standard deviation (SD) for two-year weight loss was expected to be 6.8 kg, based on prior weight treatment studies [15
]. Calculations showed that differences in mean two-year weight loss of about 1.4 kg between conditions will be detectable with 90% power in two-tailed tests with a significance level of 0.05 for a sample of 500 employees in each group. The sample size for the study was therefore determined at 1500.
Randomisation took place at an individual level. After baseline measurements, the employee was randomised to one of the three study groups and either to a group receiving basic weight measurements (80% of each study group) or to a group receiving additional measurements (20% of each study group). This two-step randomisation meant that there were six groups an employee could be assigned to. Randomisation to these six groups was done by block randomisation, with each block containing 15 allocations. A computerized random number generator drew up an allocation schedule. An administrative assistant put the group allocation in opaque sealed envelopes, numbered 1 to 1,500. These envelopes were taken to the locations of the baseline measurements and opened in the given order. The researchers were blinded for the allocation schedule, but were not blinded for allocation after randomisation. The participants were, in consequence of the nature of the intervention, not blinded for allocation after randomisation. Employees were not allowed to change groups after randomisation.
The employees in the control group received three information brochures with general information on overweight, physical activity and nutrition, and a calorie chart. These materials were briefly explained to the employee. All materials were published by the Netherlands Heart Foundation, for general use.
Both intervention groups received a lifestyle intervention programme. This intervention was similar in content, but differed in the way the content was distributed to them and in the way the participant communicated with a personal counsellor that was appointed to them. The intervention conditions are described later on.
The study design and participant flow (achieved at recruitment closure and brought up to closure of the baseline measurement period, August 2004) are shown in Figure .
Out of a total of approximately 21,000 employees, 4,619 returned the screening questionnaire. Of these, 2,615 were eligible and had no objections to receiving further information. They received an information brochure in which the study protocol was clearly described. Employees were free to further inquire about the study without any engagement. 1,454 employees were willing to participate and applied in time. They were all individually invited for a first appointment in which the basic measurements body weight and body height were assessed by the researchers. From these measurements, BMI was calculated. Those with a BMI lower than 25 kg/m2, which was the case for nine employees, were excluded. One employee became pregnant between screening and baseline measurements and was also excluded. One employee withdrew before randomisation. 57 employees did not show up for the baseline measurements. Therefore, 1,386 employees were randomised to the two intervention conditions (phone group N = 462, internet group N = 464), or to the control group (N = 460). Within these groups a number of subjects received additional measurements as follows: phone group N = 91, internet group N = 93 and control group N = 92. A second appointment was arranged within one week after randomisation, to perform these measurements. In each group a few employees were unable to attend the additional measurements: five out of the phone group, two out of the internet group, and ten out of the control group.
The lifestyle intervention programme is an adapted version of previous work of HealthPartners in Minnesota, USA [13
]. This Dutch version of the lifestyle programme is called "Leef je Fit". The programme is based on social cognitive theory [17
] and emphasizes 1) identifying behaviours in need of change, 2) setting goals for change, 3) monitoring progress, 4) modifying environmental cues to facilitate change, and 5) modifying consequences to motivate change. An essential part of the programme is coaching on these activities by a personal counsellor.
The programme consists of ten lessons or 'modules'. These provide information on nutrition and physical activity, and teach techniques for changing behaviour (e.g., self-monitoring). Assignments in the modules assist in applying these techniques. The programme can be worked through either at work or at home.
Participants are asked to study a module and finish the assignments every two weeks, so they are able to finish the programme within six months. After finishing each module, participants are contacted by their counsellor. Counselling is done according to two comparable standardized counselling protocols (i.e., for the two communication methods).
Next to the programme materials, the participant receives a pedometer at the start of the programme.
Programme materials for the phone group are provided after randomisation. The employee receives a binder with the ten modules, and the same brochures that are given to the control group. The intervention starts when the personal counsellor first calls the participant, about two weeks after randomisation. In this first contact the counsellor explains the workings of the programme and sets a time and date for the next contact in which the first module will be addressed. In between these contacts, the employee studies the module and fills out the assignments. The counsellor calls the employee at the scheduled date and time, to go through the module and to talk about the assignments according to the standardized protocol. At the end of the call, a new date and time are set for the next call about two weeks later. This interactive process continues until the employee completes all modules, or until the participant declines contact. Participants also have the possibility to contact their counsellor by phone between modules, if they have additional questions.
A website is developed for the intervention. The internet-based lifestyle intervention starts when the employee receives a welcome e-mail, about two weeks after randomisation, with a unique username and password. The first time an employee logs on to the site, he is directed to an introduction which explains the workings of the programme. The employee is asked to fill out some personal details for the counsellor. The employee then starts with the first module. The assignments are filled out on the site and stored in a personalized area of the website. The personal counsellor receives an alert when the employee has finished a module. The counsellor then checks the assignments and comments on them through e-mail within 5 working days, according to the standardized protocol. One week after finishing a module, the participant is able to start with the assignments of the next module. Modules have to be worked through in the given order.
When an employee does not log on to the website according to schedule, he receives an e-mail reminder twice a week. These reminders are continued for the full duration (i.e., six months) of the programme. Besides being reminded by e-mail, employees can also choose to be reminded by text message on their mobile phone.
Participants have the possibility to e-mail their counsellor between modules, if they have additional questions.
Four counsellors with a degree in nutrition or physical activity provide coaching for the participating employees. Prior to the start of the intervention, they receive a four week training. The counsellors are made familiar with the principles behind the programme and go through the intervention materials. A psychologist trains them in counselling techniques. A pilot group is put together to give the counsellors the opportunity to practise with the counselling protocols, counselling skills and administrative procedures.
Counselling takes place according to a standardized protocol, which is developed for each communication method. Both protocols are similar to each other in approach.
The protocol for the phone-condition consists of a timed outline for each module and provides the counsellors with guidelines on how to counsel the assignments.
The protocol for the internet-condition consists of a semi-prepared e-mail for each module. The e-mail discusses the module by giving generic comments. To this the counsellor adds the answers the participant has filled out in the assignments and addresses those individually in the same manner as in the phone calls.
In order to keep track of appointments, purport of phone calls and e-mails, and the progress of the participants, a web-based participant management system is developed. Counsellors have access to the assignments and progress from the internet group through this participant management system.
Outcome measures, confounding- and mediating variables
Primary outcomes of the study are 1) weight change and change in BMI, 2) change in physical activity level, and 3) change in dietary intake of fat, fruit, vegetables, sugar and alcoholic beverages. Effects in both intervention groups combined against the effects in the control group are studied, as well as effects in the phone group against effects in the internet group.
Besides these primary outcomes, other outcomes are studied also. Among those are changes in waist circumference, sum of skin folds, blood pressure, total blood cholesterol and aerobic fitness. Other outcomes that are studied include change in work performance/productivity, perceived health, empowerment, stage of change and self efficacy concerning weight control, physical activity and eating habits. Changes in direct medical costs and changes in indirect costs for sickness absence and loss of work productivity are studied as well.
In conclusion, a process evaluation of the execution of the intervention and of participant satisfaction is carried out after completion or the intervention.
Assessment of the aforementioned outcomes, is done either by questionnaire or by physiological measurements. Physiological measurements take place at the work site or near the work site. Questionnaires are sent to the home address of the participant. The scheduling of all measurements is shown in Table .
Primary outcome measures
Body weight and BMI
Body weight and body height are assessed in all participants. Body weight is measured in kg, to the nearest 0.1 kg, with a digital scale (Seca 770; Seca GmbH & Co, Hamburg, Germany). Participants are wearing light clothing and no shoes. Body height is measured in m, to the nearest 0.001 m, with a portable stadiometer (Seca 214, Leicester Height Measure; Seca GmbH & Co, Hamburg, Germany). Positioning of the body is standardized by asking the subject to stand straight, without shoes and with the heels together. Both weight and height are measured twice, and for each mean value of the two measurements is computed. BMI is calculated by dividing the measured body weight (kg) by the squared measured body height (m).
In addition, in a questionnaire self-reported body weight is assessed. Participants are asked to weigh themselves wearing light clothing and no shoes.
Dietary intake and physical activity
Dietary intake and physical activity are assessed by means of a questionnaire. The focus of dietary intake is on fat, fruit and vegetable intake. Fat intake is assessed by the validated Dutch Fat List [18
]. Fruit and vegetable intake are asked with a short fruit and vegetable questionnaire, that has been validated as well [19
]. Intake of sugar and alcoholic beverages is assessed with a set of questions which were developed for this study. Physical activity level is assessed with the validated Short Questionnaire to Asses Health enhancing physical activity (SQUASH) [21
Physiological outcome measures
For a random sample of each group, several additional physiological measurements are assessed in the following order: waist circumference, sum of skin folds, blood pressure, total blood cholesterol, and aerobic fitness. They are all done according to standardized protocols and take approximately 45 minutes per session.
Waist circumference (in cm) is measured twice with a measuring tape (Gulick; Creative Health Products, Ann Arbor, MI, USA) with a range of 0–150 cm. Waist circumference is measured to the nearest 0.1 cm, at the midpoint between the lower border of the ribs and the upper border of the pelvis. A mean value of the two measurements is computed.
In addition to the objectively assessed waist circumference in the sub-sample, all 1386 employees are asked to report their self measured waist circumference in each questionnaire. For that aim, a measuring tape is sent to all participants along with the baseline questionnaire. This measuring tape is developed for the study and has a range of 0–135 cm. Participants get instructions on how to use the measuring tape and are asked to report their waist circumference to the nearest cm.
Sum of skin folds
According to the method of Weiner and Lourie [22
] the following four skin folds are measured twice with a Harpenden caliper (HSK-BI; Baty International, Burgess Hill, UK) up to the nearest 0.1 mm and on the right side of the body: sub scapular, suprailiac, triceps and biceps. In case the two measurements of a fold differ more than 1.0 mm, the skin fold is measured a third time. Next, for each skin fold, a mean value of the measurements is computed and the four skin folds are added up.
Blood pressure is measured with a fully automated blood pressure monitor (Omron HEM 757E [M5-I]; Omron Healthcare Europe BV, Hoofddorp, The Netherlands). This blood pressure monitor is validated and recommended for clinical use [23
]. A regular size cuff is used on the right upper arm. When the subject has an upper arm circumference of 33 cm or more, a large size cuff is used. The right arm is placed on a table so as the cuff is on a level with the heart. After the employee has rested for 5 minutes in a sitting position, blood pressure is measured twice. A mean value of the two measurements is computed.
Employees that have a blood pressure of 140/90 mmHg or higher are advised to visit their general practitioner.
Total blood cholesterol
Total blood cholesterol level is determined with the Reflotron®
Plus (Roche Diagnostics GmbH, Mannheim, Germany) portable blood analysis system. The Reflotron®
Plus provides a good risk classification [24
]. Total cholesterol is determined in non-fasting capillary blood, collected from a finger prick. If total cholesterol is lower than 3.0 mmol/l or exceeds 6.5 mmol/l, the measurement is repeated. If both measurements are over 6.5 mmol/l, the employee is advised to visit his general practitioner.
Aerobic fitness is assessed by means of the Chester Step Test (CST) [25
]. The CST is a sub-maximal test that gives a reliable prediction of VO2
]. The employee is asked not to drink coffee or smoke two hours prior to the test. During the CST, the employee steps on and off a 10, 15, 20 or 25 cm gym bench. The height of the bench depends on the participant's age and current fitness level, as described in the manual of the test [27
]. The test starts at the pace of 15 steps per minute. The pace increases every two minutes to 20, 25, 30 and 35 steps per minute, respectively. A metronome sets this stepping rate. The heart rate of the participant is monitored continuously by means of a heart rate monitor (Polar S610; Polar Electro Oy, Kempele, Finland). Also, the subject must report his subjective rate of exertion at the end of each stage, using a Borg scale ranging from 6 tot 20 [28
]. The test is terminated at the end of a stage at which the subject's heart rate reaches 80% of his age predicted maximal heart rate (220 minus age), or when the reported rate of perceived exertion exceeds 14 (hard) [27
]. The estimated VO2max is calculated with calculator software that comes with the Chester step test (ASSIST creative resources Limited, Redwither Business Park, UK).
Other outcome measures
Perceived health and empowerment are respectively assessed with the RAND-36[29
] and the Mastery Scale of Pearlin [30
]. Stage of change in relation to weight control, physical activity and eating habits is assessed using questions about intentions to start changing these behaviours [31
]. The behavioural determinant self efficacy, concerning weight control, physical activity and eating habits is assessed with a set of questions developed for this study.
The cost-effectiveness and the cost-utility of the lifestyle programme will be ascertained. The cost-effectiveness analysis will be performed from both a company perspective as well as from a societal perspective. The primary outcome measures weight change, change in physical activity level, and change in dietary intake will be included in this cost-effectiveness analysis.
For the company perspective, intervention costs will be compared with obtained benefits from reduced sickness absence and increased work productivity. Intervention costs include costs for the development of the intervention, such as the development of the website and the binders, and costs for the implementation of the intervention, such as salary costs of the counsellors and costs for hosting the website. Self-reported sickness absence is prospectively measured by keeping a six-monthly diary. On a monthly basis the employee records full days absence from work due to sick leave. Diaries are provided at the start of each six month measuring period. In addition, data on sickness absenteeism will be derived from employer records, from six months before baseline till two years after baseline. Work productivity is measured with the work related questions from the WHO Health and Work Performance Questionnaire (HPQ) [34
]. Work productivity assessed by the HPQ is based on self-report. The HPQ is validated and available in several languages, including Dutch.
The cost-effectiveness from a societal perspective will be addressed by assessing health care utilization and medical costs. Health care utilization is based on actual resource use, using prospective data collected by a diary in which the employee records use of medical services and medication on a monthly basis. This diary is combined with the diary recordings on sickness absence. Direct medical costs of health care utilization are calculated using cost prices if available and otherwise tariffs will be used.
Utilities for the cost-utility analysis are based on the EuroQol [35
Confounding and mediating variables
Possible confounders that are assessed include certain demographics, smoking behaviour, health conditions, weight outcome evaluation and weight control behaviours. Demographic variables that are measured are age, educational level, personal income, country of birth, gender, marital status and number of adults and children living in the employee's household. Smoking is defined as use (yes/no) of cigarettes, or other tobacco products (cigars, pipes). Having quit smoking (yes/no) in the past six months is also assessed. Health conditions related to overweight, physical inactivity and unhealthy eating habits are assessed using a series of questions about the use of medication for hypercholesterolemia, hypertension, angina, heart disease, myocardial infarct, depression and diabetes. Weight outcome evaluation is assessed by three questions concerning acceptable weight, desirable weight in six months and the wish for weight loss or weight management. Weight control behaviours are measured by reports of frequency of self weighing, frequency of weight loss attempts in the last two years, and employed methods for weight control.
The process evaluation among participants assesses the participants opinion of the allocated study group, use and appraisal of the components of the lifestyle intervention programme or (in the case of the control group) of the provided general brochures, evaluation of coaching by the personal counsellor (if applicable) and a generic grade for the programme. Additional questions are asked about contentment with achieved results, reasons for not finishing the programme, and how much the participant would be willing to pay for the offered programme.
Objective measures for use of both intervention methods will be obtained from the database that underlies the participant management system and the intervention website.
Baseline values will be analysed for differences in the three groups, by one-way-analysis of variance for numerical data and chi-square for categorical data.
Analyses for the primary outcomes, for the physiological outcomes and for the other outcomes (e.g., perceived health) will be at six months (short term) and after two years (long term). They will be performed according to the intention-to-treat principle. Analyses for short term results will be performed with multiple linear regression. Analyses for long term results will be performed with multiple longitudinal regression.
Bootstrapping will be used for comparison of mean direct, indirect and total costs between groups [36
]. Confidence intervals for the mean difference in costs will be obtained by bias corrected and accelerated bootstrapping. Cost-effectiveness ratios and cost-utility ratios will also be calculated with bootstrapping according to the bias corrected percentile method. Separate cost-effectiveness ratios will be calculated for the primary outcomes [37
]. Bootstrapped cost-effect pairs will be graphically represented on a cost-effectiveness plane [38
]. Acceptability curves will be calculated, which show the probability that a treatment is cost-effective at a specific ratio.