Whitehall I Study


The Whitehall study is a ground-breaking longitudinal (prospective cohort) study that clearly demonstrated the association between social determinants of health (the social gradient) and morbidity and mortality (cardiovascular disease) in a population of British civil servants (Breeze et al., 2001; Chandola et al., 2008; Marmot et al., 1978).

In no more than 1,000 words, please address the following short-answer questions:

  • What is the sampling frame for each phase of the Whitehall study (Whitehall I and II)?
  • How was disease risk assessed (both in data collection and analysis) in each of the three studies, and why?
  • To what extent can the results of each of the three studies can be generalised to other populations (include reasons for your answer)?
  • Would it be feasible to conduct a similar study in Australia using an existing cohort such as the 45 and up study cohort, or the Australian Women’s longitudinal study cohort? Why or why not?

Complete and submit the assignment by the end of Module 2.

Assessment criteria

  • Knowledge and understanding of prospective cohort studies
  • Knowledge and understanding of social determinants of health
  • Knowledge an understanding of the concepts of sampling and bias
  • Use of mathematical concepts to describe sampling frame and disease risk
  • Interpretation of the findings of the Whitehall study and its generalisability to other populations
  • Academic presentation including accurate referencing using APA style

Sampling Frame of the Whitehall Study

Whitehall I study sought to unmask the interconnection between coronary heart disease mortality, employment grade, and factors exposing one at risk of coronary disease (Killoran & Kelly, 2010). The number of people enclosed by the longitudinal study was 17530 civil servants based in London. Initially, a screening examination was attended by 18403 men. Each individual was presented with a standard questionnaire. The questionnaire mainly focused on employment grades, and based on received feedback, men were categorized into grades. They were; executive, professional, administrative, clerical, and others. However, 873 men from the British Council and the diplomatic service were excluded from the study. This was because their employment status’ were incomparable. The remaining sample population, therefore, comprised 17530 men from other employment departments (Naidoo & Wills, 2010). Records of over 99% of the participants were identified and tagged in the national health central registry. Any participant who has therefore died henceforth has been accounted for as birth certificates are available.

In Whitehall study II, phase one conducted between 1985-1988 saw the recruitment of 10308 people to be assessed who emanated from twenty departments of civil service based in London (Kirch, 2012). Phase 2 (1989-1990), phase 3 (1991-1993), phase 4 (1995), phase 5 (1997-1999), phase 6 (2001) and phase 7 (2002-2004) entailed data collection. In phases 2,4 & 6, postal questionnaires were used. Phases 3,5 & 7, however, entailed complete clinical examination (Marmot & Brunner, 2005)

Disease Risk Assessment

In the study to bring into the limelight the mechanisms associated with work stress and coronary heart disease, various criteria were used to assess the risk of the disease. Job strain questionnaires were used to collect data relating to self-reported work stress. The addition of the number of times that iso-strains were reported by participants in phases 1 and 2 enabled the creation of a cumulative measure of work stress. Data analysis involved using cumulative measures of work-related stress to display the hazardous ratios of coronary heart disease occurrence. It was therefore established that there was a direct connection between higher risks of coronary heart disease and greater reports of work stress.

The second study involved establishing a relationship between the employment grades of British civil servants and the prevalence of coronary heart disease amongst them. Data was collected by issuing questionnaires where the participants (men) were required to indicate their employment grades. It is from collected data that grades were classified into professional, administrative, clerical, and executive, and the lowest work profile grade was termed as ‘others.’ Data analysis sought to explain the coronary heart disease mortality percentage amongst workers in various employment grades but within the same age limits.

The third study was based on 29 years of following up on the Whitehall study. It sought to explain whether social-economic drawbacks persisted in old age. Given prior registration during Whitehall 1 with the National Health Service Central Register, the body isolated the various health authorities through which individual cohort members were registered to various family doctors. Permission was granted by the chief executives of the individual health authorities to the register to avail all the survivors’ addresses. Sent to the survivors identified were invitation letters, questionnaires, two reminders, and consent forms. A shortened version of the questionnaire entailing the crucial information was attached alongside the second reminder. Statistical analysis was conducted using heterogeneity Chi-square tests to establish univariate relationships. Odds ratios were estimated using logistics regression.

Generalization of the Results to Other Populations

Based on the three studies, the overall results apply to other populations. This is because the prevalence of baseline coronary heart disease is directly attributable to the income generated by the individuals in the population of choice (WHO, 2007). Pay grades determine the amount of income an individual earns, aligning the individual into respective socioeconomic measures. Work stress is also prevalent in other populations; thus, results from these studies could be generalized to several populations.

Feasibility of Conducting Similar Studies

The feasibility of conducting similar studies in Australia is high. The 45 and up study cohort could even prove to be more effective as it covers very large population of more than 26700 individuals at any single recruitment (Webb & Bain, 2010). Data collected would be, therefore, well distributed and emanating from a larger sample. The Australian Women’s longitudinal study cohort, on the other hand, could have a relatively low feasibility as all its participants are women, and thus, the data collected would not reflect the whole society but instead of one gender (Chang & Daly, 2012).


Chang, E., & Daly, J. (2012). Transitions in Nursing – E-Book: Preparing for Professional Practice (3 ed.). Elsevier Health Sciences.

Killoran, A., & Kelly, M. P. (2010). Evidence-based Public Health: Effectiveness and Efficiency (illustrated ed.). Patrick.

Kirch, W. (2012). Public Health in Europe: — 10 Years European Public Health Association — (illustrated ed.). Springer Science & Business Media.

Marmot, M., & Brunner, E. (2005). Cohort Profile: The Whitehall II study. International Journal of Epidemiology, 34(2). Retrieved 7 1, 2018, from https://academic.oup.com/ije/article/34/2/251/746997

Naidoo, J., & Wills, J. (2010). Developing Practice for Public Health and Health Promotion E-Book. Elsevier Health Sciences.

Nickitas, D. M., Middaugh, D. J., & Aries, N. (2010). Policy and Politics for Nurses and Other Health Professionals. Jones & Bartlett Publishers.

Orth-Gomer, K., & Schneiderman, N. (2013). Behavioral Medicine Approaches to Cardiovascular Disease Prevention. Psychology Press.

Webb, P., & Bain, C. (2010). Essential Epidemiology: An Introduction for Students and Health Professionals (2, revised ed.). Cambridge University Press.

WHO. (2007). Prevention of Cardiovascular Disease: Guidelines for Assessment and Management of Cardiovascular Risk (illustrated ed.). World Health Organization.