The influence of non-response in a population-based cohort study on type 2 diabetes evaluated by the Swedish Prescribed Drug Register

Undiagnosed type 2 diabetes is common - intensified screening of established risk groups is imperative in Sweden: the SDPP cohort

BACKGROUND

Undiagnosed type 2 diabetes (T2D) is a global problem. Current strategies for diagnosis in Sweden include screening individuals within primary healthcare who are of high risk, such as those with hypertension, obesity, prediabetes, family history of diabetes, or those who smoke daily. In this study, we aimed to estimate the proportion of individuals with undiagnosed T2D in Stockholm County and factors associated with T2D being diagnosed by healthcare. This information could improve strategies for detection.

METHODS

We used data from the Stockholm Diabetes Prevention Programme (SDPP) cohort together with information from national and regional registers. Individuals without T2D aged 35-56 years at baseline were followed up after two ten-year periods. The proportion of diagnosed T2D was based on register information for 7664 individuals during period 1 and for 5148 during period 2. Undiagnosed T2D was assessed by oral glucose tolerance tests at the end of each period. With logistic regression, we analysed factors associated with being diagnosed among individuals with T2D.

RESULTS

At the end of the first period, the proportion of individuals with T2D who had been diagnosed with T2D or not was similar (54.0% undiagnosed). At the end of the second period, the proportion of individuals with T2D was generally higher, but they were less likely to be undiagnosed (43.5%). The likelihood of being diagnosed was in adjusted analyses associated with overweight (OR=1.85; 95% CI 1.22-2.80), obesity (OR=2.73; 95% CI 1.76-4.23), higher fasting blood glucose (OR=2.11; 95% CI 1.67-2.66), and self-estimated poor general health (OR=2.42; 95% CI 1.07-5.45). Socioeconomic factors were not associated with being diagnosed among individuals with T2D. Most individuals (>71%) who developed T2D belonged to risk groups defined by having at least two of the prominent risk factors obesity, hypertension, daily smoking, prediabetes, or family history of T2D, including individuals with T2D who had not been diagnosed by healthcare.

CONCLUSIONS

Nearly half of individuals who develop T2D during 10 years in Stockholm County are undiagnosed, emphasizing a need for intensified screening of T2D within primary healthcare. Screening can be targeted to individuals who have at least two prominent risk factors.

Cancer patients' participation in population-based health surveys: findings from the HUNT studies

Background

The magnitude of participation bias due to non-participation should be considered for cancer patients invited to population-based surveys. We studied participation rates among persons with and without cancer in a large population based study, the Nord-Trøndelag Health Study (HUNT).

Methods

Citizens 20 years or above living in the Nord-Trøndelag County of Norway have been invited three times to comprehensive health surveys. The invitation files with data on sex, invitation date and participation were linked to the Cancer Registry of Norway. In a first step unadjusted crude participation rates (participants/invited persons) were estimated for cancer patients (CaPts) and non-cancer persons (NonCaPers), followed by logistic regression analyses with adjustment for age and sex. To evaluate the “practical” significance of the estimated odds ratios in the cancer diagnosis group, relative risks were also estimated comparing the observed rates to the estimated rates under the counterfactual assumption of no earlier cancer diagnosis among CaPts.

Results

Overall 3 % of the participants in the three HUNT studies were CaPts and 59 % of them had been diagnosed with their first life-time cancer >5 years prior to each survey. In each of the three HUNT surveys crude participation rates were similar for CaPts and NonCaPers. Adjusted for sex and age, CaPts’ likelihood to participate in HUNT1 (1984–86) and HUNT2 (1995–97), but not in HUNT3 (2006–2008), was statistically significantly reduced compared to NonCaPers, equaling a relative risk of 0.98 and 0.96, respectively. The lowest odds ratio emerged for CaPts diagnosed during the last 2 years preceding a HUNT invitation. Only one-third of CaPts participating in a survey also participated in the subsequent survey compared to approximately two-thirds of NonCaPers, and 11 % of CaPts participated in all three HUNT surveys compared to 37 % of NonCaPers.

Conclusion

In the three HUNT surveys no or only minor participation bias exist as to CaPts’ participation rates. In longitudinal studies selection bias as to long-term cancer survivorship should be taken into account, the percentage of repeatedly participating CaPts diminishing more strongly than among NonCaPers.

Assessment of participation bias in cohort studies: systematic review and meta-regression analysis

Abstract The proportion of non-participation in cohort studies, if associated with both the exposure and the probability of occurrence of the event, can introduce bias in the estimates of interest. The aim of this study is to evaluate the impact of participation and its characteristics in longitudinal studies. A systematic review (MEDLINE, Scopus and Web of Science) for articles describing the proportion of participation in the baseline of cohort studies was performed. Among the 2,964 initially identified, 50 were selected. The average proportion of participation was 64.7%. Using a meta-regression model with mixed effects, only age, year of baseline contact and study region (borderline) were associated with participation. Considering the decrease in participation in recent years, and the cost of cohort studies, it is essential to gather information to assess the potential for non-participation, before committing resources. Finally, journals should require the presentation of this information in the papers.

Systematic Review and Meta-Analysis of Response Rates and Diagnostic Yield of Screening for Type 2 Diabetes and Those at High Risk of Diabetes

Background

Screening for type 2 diabetes (T2DM) and individuals at risk of diabetes has been advocated, yet information on the response rate and diagnostic yield of different screening strategies are lacking.

Methods

Studies (from 1998 to March/2015) were identified through Medline, Embase and the Cochrane library and included if they used oral glucose tolerance test (OGTT) and WHO-1998 diagnostic criteria for screening in a community setting. Studies were one-step strategy if participants were invited directly for OGTT and two, three/four step if participants were screened at one or more levels prior to invitation to OGTT. The response rate and diagnostic yield were pooled using Bayesian random-effect meta-analyses.

Findings

47 studies (422754 participants); 29 one-step, 11 two-step and seven three/four-step were identified. Pooled response rate (95% Credible Interval) for invitation to OGTT was 65.5% (53.7, 75.6), 63.1% (44.0, 76.8), and 85.4% (76.4, 93.3) in one, two and three/four-step studies respectively. T2DM yield was 6.6% (5.3, 7.8), 13.1% (4.3, 30.9) and 27.9% (8.6, 66.3) for one, two and three/four-step strategies respectively. The number needed to invite to the OGTT to detect one case of T2DM was 15, 7.6 and 3.6 in one, two, and three/four-step strategies. In two step strategies, there was no difference between the response or yield rates whether the first step was blood test or risk-score. There was evidence of substantial heterogeneity in rates across study populations but this was not explained by the method of invitation, study location (rural versus urban) and developmental index of the country in which the study was performed.

Conclusions

Irrespective of the invitation method, developmental status of the countries and or rural/urban location, using a multi-step strategy increases the initial response rate to the invitation to screening for diabetes and reduces the number needed to have the final diagnostic test (OGTT in this study) for a definite diagnosis.

No increase in new users of blood glucose-lowering drugs in Norway 2006-2011: a nationwide prescription database study

BACKGROUND

National estimates for the occurrence of diabetes are difficult to obtain, particularly time trends in incidence. The aim was to describe time trends in prevalent and incident use of blood glucose-lowering drugs by age group and gender in Norway during 2005-2011.

METHODS

Data were obtained from the nationwide Norwegian Prescription Database. We defined prevalent users of "insulins only" as individuals having no oral antidiabetic drugs (OAD) dispensed from a pharmacy during the previous 24 months or in the subsequent 12 months. Incident users had no blood glucose-lowering drugs dispensed in the previous 24 months; incident "insulins only" users also had no OAD in the subsequent 12 months.

RESULTS

In 2011, 3.2% of the population had blood glucose-lowering drugs dispensed, and the incidence rate was 313 per 100,000 person years. The prevalence of OAD use increased from 1.8% in 2005 to 2.4% in 2011; however a decreasing trend in incidence of OAD use was observed, particularly in those aged 70 years and older. In 2010, 0.64% of the population had insulins only dispensed, with an overall incidence rate in the total population of 33 per 100,000 person years which was stable over time.

CONCLUSIONS

In this nationwide study, we found that although the prevalent use of OAD had increased in recent years, incident use was stable or had decreased. This may indicate that the increase in diabetes occurrence in Norway is levelling off, at least temporarily.

Longterm follow-up in European respiratory health studies - patterns and implications

Background

Selection bias is a systematic error in epidemiologic studies that may seriously distort true measures of associations between exposure and disease. Observational studies are highly susceptible to selection bias, and researchers should therefore always examine to what extent selection bias may be present in their material and what characterizes the bias in their material. In the present study we examined long-term participation and consequences of loss to follow-up in the studies Respiratory Health in Northern Europe (RHINE), Italian centers of European Community Respiratory Health Survey (I-ECRHS), and the Italian Study on Asthma in Young Adults (ISAYA).

Methods

Logistic regression identified predictors for follow-up participation. Baseline prevalence of 9 respiratory symptoms (asthma attack, asthma medication, combined variable with asthma attack and/or asthma medication, wheeze, rhinitis, wheeze with dyspnea, wheeze without cold, waking with chest tightness, waking with dyspnea) and 9 exposure-outcome associations (predictors sex, age and smoking; outcomes wheeze, asthma and rhinitis) were compared between all baseline participants and long-term participants. Bias was measured as ratios of relative frequencies and ratios of odds ratios (ROR).

Results

Follow-up response rates after 10 years were 75% in RHINE, 64% in I-ECRHS and 53% in ISAYA. After 20 years of follow-up, response was 53% in RHINE and 49% in I-ECRHS. Female sex predicted long-term participation (in RHINE OR (95% CI) 1.30(1.22, 1.38); in I-ECRHS 1.29 (1.11, 1.50); and in ISAYA 1.42 (1.25, 1.61)), as did increasing age. Baseline prevalence of respiratory symptoms were lower among long-term participants (relative deviations compared to total baseline population 0-15% (RHINE), 0-48% (I-ECRHS), 3-20% (ISAYA)), except rhinitis which had a slightly higher prevalence. Most exposure-outcome associations did not differ between long-term participants and all baseline participants, except lower OR for rhinitis among ISAYA long-term participating smokers (relative deviation 17% (smokers) and 44% (10–20 pack years)).

Conclusions

We found comparable patterns of long-term participation and loss to follow-up in RHINE, I-ECRHS and ISAYA. Baseline prevalence estimates for long-term participants were slightly lower than for the total baseline population, while exposure-outcome associations were mainly unchanged by loss to follow-up.

The Rotterdam Study: 2014 objectives and design update

The Rotterdam Study is a prospective cohort study ongoing since 1990 in the city of Rotterdam in The Netherlands. The study targets cardiovascular, endocrine, hepatic, neurological, ophthalmic, psychiatric, dermatological, oncological, and respiratory diseases. As of 2008, 14,926 subjects aged 45 years or over comprise the Rotterdam Study cohort. The findings of the Rotterdam Study have been presented in over a 1,000 research articles and reports (see www.erasmus-epidemiology.nl/rotterdamstudy ). This article gives the rationale of the study and its design. It also presents a summary of the major findings and an update of the objectives and methods.

Work stress, sense of coherence, and risk of type 2 diabetes in a prospective study of middle-aged Swedish men and women

OBJECTIVE

To investigate the prospective influence of work stress on type 2 diabetes (T2D).

RESEARCH DESIGN AND METHODS

This population-based cohort included 3,205 women and 2,227 men, aged 35-56 years, with baseline normal glucose tolerance measured with oral glucose tolerance test. At follow-up 8-10 years later, T2D was diagnosed in 60 women and 111 men. Work stress factors evaluated by questionnaire (i.e., demands, decision latitude, job strain, shift work, overtime work, and also sense of coherence) were studied in association with T2D. Odds ratios (ORs) and 95% CIs adjusted for age, education, BMI, physical activity, smoking, family history of diabetes, and psychological distress were calculated.

RESULTS

In women, low decision latitude was associated with T2D on its own (OR 2.4 [95% CI 1.1-5.2]) and combined with high demands: job strain (OR 4.2 [2.0-8.7]), adjusted for all available potential confounders. Also, shift work increased the risk of T2D in women (OR 2.2 [1.0-4.7]) when adjusted for age, education, and psychological distress, although this risk was diluted after multifactor adjustment (OR 1.9 [0.8-4.4]). In men, high work demands and high strain decreased the risk of T2D (OR 0.5 [0.3-0.9]) for both measures, as did an active job (high demands and high decision latitude, OR 0.4 [0.2-0.9]).

CONCLUSIONS

Work stress and shift work may contribute to the development of T2D in women. In men, the risk was decreased by high work demands, high strain, and an active job.