Danielle H. Morris

Addressing between-study heterogeneity and inconsistency in mixed treatment comparisons: Application to stroke prevention treatments in individuals with non-rheumatic atrial fibrillation

Mixed treatment comparison models extend meta-analysis methods to enable comparisons to be made between all relevant comparators in the clinical area of interest. In such modelling it is imperative that potential sources of variability are explored to explain both heterogeneity (variation in treatment effects between trials within pairwise contrasts) and inconsistency (variation in treatment effects between pairwise contrasts) to ensure the validity of the analysis.The objective of this paper is to extend the mixed treatment comparison framework to allow for the incorporation of study-level covariates in an attempt to explain between-study heterogeneity and reduce inconsistency. Three possible model specifications assuming different assumptions are described and applied to a 17-treatment network for stroke prevention treatments in individuals with non-rheumatic atrial fibrillation.The paper demonstrates the feasibility of incorporating covariates within a mixed treatment comparison framework and using model fit statistics to choose between alternative model specifications. Although such an approach may adjust for inconsistencies in networks, as for standard meta-regression, the analysis will suffer from low power if the number of trials is small compared with the number of treatment comparators.

Diabetes and cardiovascular events in women with polycystic ovary syndrome: a 20‐year retrospective cohort study

Women with polycystic ovary syndrome (PCOS) are potentially at increased risk of cardiovascular (CV) diseases due to well‐established risk factors, including insulin resistance, obesity and type 2 diabetes mellitus (T2DM). However, data showing excess CV events in this population are still lacking. We investigated the incidence and prevalence of CV events in a cohort of women with PCOS.

Determinants of age at menarche in the UK: analyses from the Breakthrough Generations Study

Background:Early menarche increases breast cancer risk but, aside from weight, information on its determinants is limited.Methods:Age at menarche data were collected retrospectively by questionnaire from 81 606 women aged 16–98, resident in the UK and participating in the Breakthrough Generations Study.Results:Menarche occurred earlier in women who had a low birthweight (Ptrend<0.001), were singletons (P<0.001), had prenatal exposure to pre-eclampsia (P<0.001) or maternal smoking (P=0.01), were not breastfed (Ptrend=0.03), were non-white (P<0.001), were heavy (Ptrend<0.001) or tall (Ptrend<0.001) compared with their peers at age 7 and exercised little as a child (Ptrend<0.001). Menarcheal age increased with number of siblings (P<0.001) independently of birth order, and had an inverse association with birth order after adjustment for sibship size (P<0.001). In a multivariate model, birthweight, ethnicity, weight, height, exercise, sibship size and birth order remained significant, and maternal age at birth became significant (positive association, P<0.001).Conclusion:Age at menarche was influenced by both pre- and post-natal factors, and these factors may affect breast cancer risk through this route.

Body Mass Index, Exercise, and Other Lifestyle Factors in Relation to Age at Natural Menopause: Analyses From the Breakthrough Generations Study

The authors examined the effect of womens lifestyles on the timing of natural menopause using data from a cross-sectional questionnaire used in the United Kingdom-based Breakthrough Generations Study in 2003-2011. The analyses included 50,678 women (21,511 who had experienced a natural menopause) who were 40-98 years of age at study entry and did not have a history of breast cancer. Cox competing risks proportional hazards models were fitted to examine the relation of age at natural menopause to lifestyle and anthropometric factors. Results were adjusted for age at reporting, smoking status at menopause, parity, and body mass index at age 40 years, as appropriate. All P values were 2-sided. High adult weight (P(trend) < 0.001), high body mass index (P(trend) < 0.001), weight gain between the ages of 20 and 40 years (P(trend) = 0.01), not smoking (P < 0.001), increased alcohol consumption (P(trend) < 0.001), regular strenuous exercise (P < 0.01), and not being a vegetarian (P < 0.001) were associated with older age at menopause. Neither height nor history of an eating disorder was associated with menopausal age. These findings show the importance of lifestyle factors in determining menopausal age.

Secular trends in age at menarche in women in the UK born 1908-93: results from the Breakthrough Generations Study

Menarcheal age decreased over time in Western countries until cohorts born in the mid-20th century. It then stabilised, but limited data are available for recent cohorts. Menarche data were collected retrospectively by questionnaire in 2003-10 from 94,170 women who were participating in the Breakthrough Generations Study, aged 16-98 years, born 1908-93 and resident in the UK. Average menarcheal age declined from women born in 1908-19 (mean=13.5 years) to those born in 1945-49 (mean=12.6 years). It was then stable for several birth cohorts, but resumed its downward trend in recent cohorts (mean=12.3 years in 1990-93 cohort). Trends differed between socio-economic groups, but the recent decline was present in each group. In conclusion, menarcheal age appears to have decreased again in recent cohorts after a period of stabilisation.

Common genetic variants are significant risk factors for early menopause: results from the Breakthrough Generations Study

Women become infertile approximately 10 years before menopause, and as more women delay childbirth into their 30s, the number of women who experience infertility is likely to increase. Tests that predict the timing of menopause would allow women to make informed reproductive decisions. Current predictors are only effective just prior to menopause, and there are no long-range indicators. Age at menopause and early menopause (EM) are highly heritable, suggesting a genetic aetiology. Recent genome-wide scans have identified four loci associated with variation in the age of normal menopause (40–60 years). We aimed to determine whether theses loci are also risk factors for EM. We tested the four menopause-associated genetic variants in a cohort of approximately 2000 women with menopause ≤45 years from the Breakthrough Generations Study (BGS). All four variants significantly increased the odds of having EM. Comparing the 4.5% of individuals with the lowest number of risk alleles (two or three) with the 3.0% with the highest number (eight risk alleles), the odds ratio was 4.1 (95% CI 2.4–7.1, P = 4.0 × 10−7). In combination, the four variants discriminated EM cases with a receiver operator characteristic area under the curve of 0.6. Four common genetic variants identified by genome-wide association studies, had a significant impact on the odds of having EM in an independent cohort from the BGS. The discriminative power is still limited, but as more variants are discovered they may be useful for predicting reproductive lifespan.

Familial concordance for age at natural menopause: results from the Breakthrough Generations Study

Objective:Existing estimates of the heritability of menopause age have a wide range. Furthermore, few studies have analyzed to what extent familial similarities might reflect shared environment, rather than shared genes. We therefore analyzed familial concordance for age at natural menopause and the effects of shared genetic and environmental factors on this concordance. Methods:Participants were 2,060 individuals comprising first-degree relatives, aged 31 to 90 years, and participating in the UK Breakthrough Generations Study. Menopause data were collected using a self-administered questionnaire and analyzed using logistic regression and variance-components models. Results:Women were at an increased risk of early menopause (≤45 y) if their mother (odds ratio, 6.2; P < 0.001) or nontwin sister (odds ratio, 5.5; P < 0.001) had had an early menopause. Likewise, women had an increased risk of late menopause (≥54 y) if their relative had had a late menopause (mother: odds ratio, 6.1; P < 0.01; nontwin sister: odds ratio, 2.3; P < 0.001). Estimated heritability was 41.6% (P < 0.01), with an additional 13.6% (P = 0.02) of the variation in menopause age attributed to environmental factors shared by sisters. Conclusions:We confirm that early menopause aggregates within families and show, for the first time, that there is also strong familial concordance for late menopause. Both genes and shared environment were the source of variation in menopause age. Past heritability estimates have not accounted for shared environment, and thus, the effect of genetic variants on menopause age may previously have been overestimated.

Familial concordance for age at menarche: analyses from the Breakthrough Generations Study.

Age at menarche is correlated within families, but estimates of the heritability of menarcheal age have a wide range (0.45-0.95). We examined the familial resemblance for age at menarche and the extent to which this is due to genetic and shared environmental factors. Between 2003 and 2010 data were retrospectively collected by questionnaire from participants within the UK-based Breakthrough Generations Study. These analyses included 25,970 female participants aged 16-98 with at least one female relative who was also a study participant. A womans menarche was significantly delayed for each yearly increase in the menarcheal age of her monozygotic twin (average increase = 7.2 months, P < 0.001), dizygotic twin (average increase = 3.0 months, P = 0.03), older sister (average increase = 3.3 months, P < 0.001), mother (average increase = 3.4 months, P < 0.001), maternal grandmother (average increase = 1.5 months, P = 0.04), maternal aunt (average increase = 1.4 months, P < 0.001) and paternal aunt (average increase = 3.0 months, P < 0.001). There was not a significant association between the menarcheal ages of half-sister pairs or of paternal grandmother-granddaughter pairs, based on small numbers. Heritability was estimated as 0.57 [95% confidence interval 0.53, 0.61]. Shared environmental factors did not have an effect in the model. In conclusion, approximately half of the variation in age at menarche was attributable to additive genetic effects with the remainder attributable to non-shared environmental effects.

Population-based estimates of the prevalence of FMR1 expansion mutations in women with early menopause and primary ovarian insufficiency

Purpose:Primary ovarian insufficiency before the age of 40 years affects 1% of the female population and is characterized by permanent cessation of menstruation. Genetic causes include FMR1 expansion mutations. Previous studies have estimated mutation prevalence in clinical referrals for primary ovarian insufficiency, but these are likely to be biased as compared with cases in the general population. The prevalence of FMR1 expansion mutations in early menopause (between the ages of 40 and 45 years) has not been published.Methods:We studied FMR1 CGG repeat number in more than 2,000 women from the Breakthrough Generations Study who underwent menopause before the age of 46 years. We determined the prevalence of premutation (55–200 CGG repeats) and intermediate (45–54 CGG repeats) alleles in women with primary ovarian insufficiency (n = 254) and early menopause (n = 1,881).Results:The prevalence of the premutation was 2.0% in primary ovarian insufficiency, 0.7% in early menopause, and 0.4% in controls, corresponding to odds ratios of 5.4 (95% confidence interval = 1.7–17.4; P = 0.004) for primary ovarian insufficiency and 2.0 (95% confidence interval = 0.8–5.1; P = 0.12) for early menopause. Combining primary ovarian insufficiency and early menopause gave an odds ratio of 2.4 (95% confidence interval = 1.02–5.8; P = 0.04). Intermediate alleles were not significant risk factors for either early menopause or primary ovarian insufficiency.Conclusion:FMR1 premutations are not as prevalent in women with ovarian insufficiency as previous estimates have suggested, but they still represent a substantial cause of primary ovarian insufficiency and early menopause.Genet Med 16 1, 19–24.

Implementation of the automated Leicester Practice Risk Score in two diabetes prevention trials provides a high yield of people with abnormal glucose tolerance

Aims/hypothesisThe Leicester Practice Risk Score (LPRS) is a tool for identifying those at high risk of either impaired glucose regulation (IGR), defined as impaired glucose tolerance and/or impaired fasting glucose, or type 2 diabetes from routine primary care data. The aim of this study was to determine the yield from the LPRS when applied in two diabetes prevention trials.MethodsLet’s Prevent Diabetes (LPD) and Walking Away from Diabetes (WAD) studies used the LPRS to identify people at risk of IGR or type 2 diabetes from 54 general practices. The top 10% at risk within each practice were invited for screening using a 75 g OGTT. The response rate to the invitation and the prevalence of IGR and/or type 2 diabetes in each study were calculated.ResultsOf those invited 19.2% (n = 3,449) in LPD and 22.1% (n = 833) in WAD attended. Of those screened for LPD 25.5% (95% CI 24.1, 27.0) had IGR and 4.5% (95% CI 3.8, 5.2) had type 2 diabetes, giving a prevalence of any abnormal glucose tolerance of 30.1% (95% CI 28.5, 31.6). Comparable rates were seen for the WAD study: IGR 26.5% (95% CI 23.5, 29.5), type 2 diabetes 3.0% (95% CI 1.8, 4.2) and IGR/type 2 diabetes 29.5% (95% CI 26.4, 32.6).Conclusions/interpretationUsing the LPRS identifies a high yield of people with abnormal glucose tolerance, significantly higher than those seen in a population screening programme in the same locality. The LPRS is an inexpensive and simple way of targeting screening programmes at those with the highest risk.