The Meaning of the Study of Epidemiology

Article Critique 

Ganwisch, J.E., Feskanich, D., Malaspina, D., Shen, S. & Forman, J.P. (2013). Sleep duration and risk for hypertension in women: Results from The Nurses’ Health Study. American Journal of Hypertension, 26 (7), 903-911. 

1. WHY WAS THE STUDY DONE? 

What was the purpose of the study? 

The study aimed at determining the correlation between acute sleep restriction and increase in blood pressure and sympathetic nervous system activity. The research further aimed at establishing other factors which accelerate hypertension regarding which risk factors are contributing to cardiovascular diseases in women. 

Before reading the article, based on the title alone, list all the factors that you think might be associated with hypertension risk in nurses? 

The factors, which may be associated with hypertension risk in nurses, may include long working hours, inconsistent sleep patterns, short sleep duration, age, menopause and sleep deprivation. 

Was the literature review current, credible, and sufficient to support the need for a study? 

(Rationale) 

The literature review is mainly credible and sufficient to support the need for study considering that this is an outcome of several decades of study. The study relies on in-depth studies carried out in 1986 and 2000. However, when it comes to the question of whether the review is current one would feel that this review is not as current considering it goes back to 1986. A lot may have changed between then and when the article was established which may affect the currency of the information. 

What was the funding source for this study? Does this cause you any concern (rationale)? 

The study was funded by National Institutes of Health Grants from the National Heart, Lung, and Blood Institute, National Cancer Institute and American Heart Association grant-in-aid. This source of funding does not cause any concerning since the issue of cardiovascular diseases is a matter related to these institutes. Any study with the aim of finding the root cause of these diseases related to the study is a welcome relief and a reason for funding by the national institutes 

2. WHO WAS IN THE STUDY? 

What type of sampling strategy was used? 

The study made use of random sampling and included two cohorts one established in 1976, and the other was established in 1989. 

What was the population for the study? 

The population for the study was female, married nurses divided into two cohorts. The first cohort established in 1976 comprising of 121,700 nurses whose ages ranged from 30 to 55 years drawn from 11 large states in the US. The second cohort established in 1989 included 116,686 nurses aged 25-42 years drawn from fourteen states in the US. 

How many people were in the accessible population? 

The accessible population includes those individuals who stated their sleep period on the in the questionnaires 1986, 2000 and 2001. The accessible population is as follows: 82,130, 71,658, and 84,674, respectively. 

How many people were in the sample? Is there a problem with sample size? 

The actual sample narrowed down to 60,009, 32,105, and 68,784 in 1986, 2000 and 2001 respectively after excluding the prevalent hypertension cases. This sample size may present several challenges because it is a large sample size and because it mainly differs from the figures in the original sample in 1976 and 1989. 

What were major characteristics of the sample (i.e., age, gender)? 

The major characteristics of the sample are the fact that it involved female, married and registered nurses. Moreover, the sample focused on age with the age limits ranging from 25-55 years where the risk for hypertension is measured. 

Was there a difference between the sample and the responding/analytic sample? If yes, what was the difference and why did it occur? 

There was a difference between the sample and the analytic sample. The initial sample was 121,700 in 1976 and 116686 in 1989. However, the sample was narrowed down to 60,009, 32,105, and 68,784 in 1986, 2000 and 2001 as it excluded individuals with prevalent hypertension. 

3. HOW WAS THE SAMPLE STUDIED? 

Design 

What type of study design(s) was used for this study? 

The study made use of longitudinal study design as the results have been obtained from long-term studies. The initial study began in 1976 with the first cohort and the second cohort in 1986. There have been comparative studies from time to time to give differentiated and concrete research findings. 

Exposure 

What is the exposure(s) of interest in this study? 

The exposure of interest in the study is the amount of sleep or sleep durations that the population got in the course of the study. 

How was the exposure(s) measured? 

The duration of sleep achieved by each individual was gotten through self-reporting. The study team ranged the sleep hours from <5 to ≥11 with 7 hours being the reference group. 

What level of measurement was used? 

The team made use of Spearman correlation of 0.79 (P < 0.0001) between the average sleep time reported by the study group. 

How reliable/valid was the measure? 

The Spearman correlation was reliable and valid to a great extent as it offered an insight into the relationship between sleep duration and hypertension. The validity of self-reported hypertension had been assessed in 1982 and had been used since. 

Outcome 

What is the outcome of interest in this study? 

The outcome of interest in this study is the fact that women who sleep for up to 7 hours had a decrease in hypertension prevalence as compared to those who slept for lesser or more hours. 

How was the outcome of interest measured? 

The study team used multivariable logistic regression to calculate odds ratios (ORs) to evaluate the cross-sectional association between sleep duration and hypertension incidence. More so, Cox proportional hazard was used to calculate hazard ratios (HRs) for hypertension incidence in longitudinal analyses. 

What level of measurement was used? 

The study team used an attenuation of ≥10% in the β coefficients on sleep duration after introducing the hypothesized mediating variables. 

How reliable/valid was the measure? 

The measure was reliable and valid as the results were consistent with results gotten from the previous longitudinal studies, which established a relationship between short sleep duration and hypertension rates in younger cohorts . 

Analysis 

What other factors that might affect the outcome were included in the multivariable analysis? 

Age, hypercholesterolemia, diabetes, body mass index, ethnicity, diet, race, ethnicity, smoking, pollution, menopause, physical activity, alcohol, genetics, and snoring. 

Were all of the factors you identified in question 1included? 

Yes. All the factors were included and their relationship with hypertension established largely 

Are there others you think should have been included? 

No. The study covered almost all of the possible factors in an exhaustive manner. 

What type of analyses was done to determine incidence and prevalence of hypertension? 

The study team conducted cross-sectional, longitudinal multivariable and meta-analyses to determine the incidence and prevalence of hypertension. 

Was there a difference in the outcome of interest due to confounders? Discuss confounders in terms of the level of measurement and give examples) 

There was a difference in the outcome of interest due to several confounders such as apnea, anxiety, and treatment for hypertension. In essence, it was not easy to establish the levels of anxiety, sleep durations, and apnea as they are not measurable variables. On the other hand, the respondents only answered questions based on their recent sleep durations. 

What type of analysis was done to determine the relationship between exposure of interest and outcome of interest? 

The type of analysis done to establish the relationship between sleep duration and hypertension prevalence was a meta-analysis. Here sleep durations ranging at ≤5 hours were linked with considerably augmented rates of hypertension occurrence. 

4. WHAT WERE THE RESULTS? 

What were the main results? (Describe cross-sectional, meta-analyses, and longitudinal analyses results). 

The cross-sectional analyses established that the participants with sleep duration of ≤5 hours and 6 hours had significantly increased age-adjusted odds of prevalent hypertension. The meta-analyses established that sleep duration of ≤5 hours was associated with significantly increased odds of hypertension prevalence. Results from longitudinal analyses indicated that participants with sleep duration of ≤5 hours had increased age-adjusted HRs of hypertension compared to those with sleep durations of 7 hours. 

Which confounder(s) were found to affect the relationship between exposure and outcome? 

The main confounders affecting the relationship between sleep duration and hypertension prevalence were BMI, diabetes, and hypercholesterolemia. 

Describe the amount and range of increased risk found for women less than 50 years of age based on the degree of exposure . 

Women who are younger than 50 years and who sleep for less than 5 hours are more likely to have higher hypertension rates as compared to the older women. The reason for this is the fact that, increased sympathetic activation from short sleep duration could cause diastolic and systolic hypertension. 

5. WHAT DO THE RESULTS MEAN FOR CLINICAL PRACTICE? 

Threats to Validity 

What type of threats to validity might occur given this sample? (Consider the four types of validity threats). 

The validity threats, which might occur in this sample, is mortality and attrition as this is a lengthy study. The participants may have become healthier and thus affect the final findings. Moreover, continuous testing posed by the longitudinal study is a validity threat, as the participants have to give feedback continually, which may be inconsistent. Additionally, instrumentation may be a validity threat as the researchers may be forced to change instruments from time to time. Finally, the statistical regression may pose a validity threat as results may drift or become extreme. 

How might the data presented here affect your clinical practice? 

The data presented in this study is an eye-opener as it puts into place all the variables at interplay when it comes to hypertension. This information will help me approach the topic of hypertension in a different perspective in an attempt to avert or reduce its negative impacts in the course of my practice.