Lin et al. conducted across-sectional study to evaluate the relationship between the prevalence of hyperuricaemia (HUA) and non-valvular atrial fibrillation (NVAF) (1). Age, living in urban areas, alcohol consumption, central obesity, elevated fasting plasma glucose level, elevated blood pressure, lower high-density lipoprotein cholesterol level and elevated triglycerides level were significantly associated with increased risk of HUA, and HUA were at higher risk for NVAF. Regarding sex difference, serum uric acid level had a modest predictive value for NVAF in women. I have some concerns about their study.
First, Tu et al. conducted a prospective study to evaluate the risk of gout in patients with alcohol-related diseases and alcohol dependence syndrome (2). Alcohol-related diseases were significantly associated with gout risk, and severe alcohol-dependent patients were significantly associated with an increased risk of gout. The authors recommended that alcohol use assessment and measures to prevent alcohol dependence in patients with gout. I agree with their recommendation, which would partly lead to prevent NVAF.
Second, Yu et al. presented strategies to improve gout/hyperuricemia management for preventing acute arthritis attacks, cardiovascular disease, kidney disease, and other urate-related disorders (3). Alcohol consumption is closely related to gout/hyperuricemia, and psycho-social factors should also be considered for the management.
Lin et al. conducted across-sectional study to evaluate the relationship between the prevalence of hyperuricaemia (HUA) and non-valvular atrial fibrillation (NVAF) (1). Age, living in urban areas, alcohol consumption, central obesity, elevated fasting plasma glucose level, elevated blood pressure, lower high-density lipoprotein cholesterol level and elevated triglycerides level were significantly associated with increased risk of HUA, and HUA were at higher risk for NVAF. Regarding sex difference, serum uric acid level had a modest predictive value for NVAF in women. I have some concerns about their study.
First, Tu et al. conducted a prospective study to evaluate the risk of gout in patients with alcohol-related diseases and alcohol dependence syndrome (2). Alcohol-related diseases were significantly associated with gout risk, and severe alcohol-dependent patients were significantly associated with an increased risk of gout. The authors recommended that alcohol use assessment and measures to prevent alcohol dependence in patients with gout. I agree with their recommendation, which would partly lead to prevent NVAF.
Second, Yu et al. presented strategies to improve gout/hyperuricemia management for preventing acute arthritis attacks, cardiovascular disease, kidney disease, and other urate-related disorders (3). Alcohol consumption is closely related to gout/hyperuricemia, and psycho-social factors should also be considered for the management.
Third, Kuwabara et al. clarified the role of HUA as an independent risk factor for atrial fibrillation (AF) (4). Adjusted odds ratio (95% confidence interval) of HUA for AF was 3.19 (1.81-5.62). As they used AF as a dependent variable instead of NVAF, sub-analysis is needed to verify the association.
Regarding the mechanism of the association between HUA and NVAF, Liu et al. investigated the relationship between HUA and left atrial (LA) thrombus/spontaneous echo contrast and determined the predictive value of HUA in patients with NVAF (5). They concluded that HUA was closely associated with LA stasis, and it could predict and refine LA stasis risk in patients with NVAF. Maharani et al. investigated the link between HUA and AF with special reference to hyperuricemia-induced atrial remodeling (6). They recommended that uric acid transporters might become a potential strategy to reduce the risk of hyperuricemia-induced AF. Anyway, further studies are needed to accept the mechanism of the association.
References
1. Lin WD, Deng H, Guo P, et al. High prevalence of hyperuricaemia and its impact on non-valvular atrial fibrillation: the cross-sectional Guangzhou (China) Heart Study. BMJ Open. 2019;9(5):e028007. doi: 10.1136/bmjopen-2018-028007
2. Tu HP, Tung YC, Tsai WC, et al. Alcohol-related diseases and alcohol dependence syndrome is associated with increased gout risk: A nationwide population-based cohort study. Joint Bone Spine. 2017;84(2):189-196. doi: 10.1016/j.jbspin.2016.02.024
3. Yu KH, Chen DY, Chen JH, et al. Management of gout and hyperuricemia: Multidisciplinary consensus in Taiwan. Int J Rheum Dis. 2018;21(4):772-787. doi: 10.1111/1756-185X.13266
4. Kuwabara M, Niwa K, Nishihara S, et al. Hyperuricemia is an independent competing risk factor for atrial fibrillation. Int J Cardiol. 2017;231:137-142. doi: 10.1016/j.ijcard.2016.11.268
5. Liu FZ, Liao HT, Lin WD, et al. Predictive effect of hyperuricemia on left atrial stasis in non-valvular atrial fibrillation patients. Int J Cardiol. 2018;258:103-108. doi: 10.1016/j.ijcard.2018.01.080
6. Maharani N, Kuwabara M, Hisatome I. Hyperuricemia and atrial fibrillation. Int Heart J. 2016;57(4):395-9. doi: 10.1536/ihj.16-192
It has been five years since Hedegaard and colleagues published their paper on stillbirth and labour induction [1]. In Denmark, its findings are still often brought forward as weighty documentation in public and professional discussions on the appropriate timing of recommended induction of labour in low-risk pregnancies after due date. Thus, as late as in January 2020, one of the main Danish broadcasting companies (TV2) covered the subject, and a curve almost identical to the top graph of the study’s Figure 2 was presented, alongside statements from one the authors. The graph was entitled "Number of dead fetuses per 1000 pregnant women for the period 2000-2008", and the accompanying text read “[…] the numbers show that the number of dead fetuses increases markedly by the beginning of 41 completed gestational weeks and even more after 42 weeks. […] the number of fetuses that died in the mother’s womb after 37th week decreased from 120 to 60 when the regulations were changed in 2011.”
Following its publication in 2014, the study was subject to much attention from peers. The responses focused on data quality, methodological issues, and whether data did or did not support the authors’ conclusions. However, the problem regarding competing risks was not touched upon, and since ignoring this problem will make the risk appear greater than it actually is, we find it crucial to point out that competing risks was not taken into account in this study’s assessment of...
It has been five years since Hedegaard and colleagues published their paper on stillbirth and labour induction [1]. In Denmark, its findings are still often brought forward as weighty documentation in public and professional discussions on the appropriate timing of recommended induction of labour in low-risk pregnancies after due date. Thus, as late as in January 2020, one of the main Danish broadcasting companies (TV2) covered the subject, and a curve almost identical to the top graph of the study’s Figure 2 was presented, alongside statements from one the authors. The graph was entitled "Number of dead fetuses per 1000 pregnant women for the period 2000-2008", and the accompanying text read “[…] the numbers show that the number of dead fetuses increases markedly by the beginning of 41 completed gestational weeks and even more after 42 weeks. […] the number of fetuses that died in the mother’s womb after 37th week decreased from 120 to 60 when the regulations were changed in 2011.”
Following its publication in 2014, the study was subject to much attention from peers. The responses focused on data quality, methodological issues, and whether data did or did not support the authors’ conclusions. However, the problem regarding competing risks was not touched upon, and since ignoring this problem will make the risk appear greater than it actually is, we find it crucial to point out that competing risks was not taken into account in this study’s assessment of the risk of stillbirth.
Figure 2 shows the Nelson-Aalen estimates for the outcome ‘fetal death’ cumulated from 37 weeks of gestation for the two periods 2000-08 and 2009-12, where the curve for 2000-2008 illustrates the “background for the new induction paradigm in Denmark”. It appears that the curve for the latter period is well below that for 2000-08, however, what is estimated using these curves are not the cumulative proportions of fetal deaths but rather cumulative rates (or ‘hazards’) of fetal death. Cumulative rates are sometimes good approximations to cumulative proportions, namely when (1) the outcome is rare, and (2) there are no ‘competing risks’, and while (1) is well fulfilled here, (2) is certainly not. This is because the vast majority of ongoing pregnancies will end in a live birth and only very few in a fetal death. This means, that in this competing risks-situation the cumulative risk should be estimated as what is often denoted the ‘cumulative incidence’ using the so-called Aalen-Johansen estimator [2].
Thus, interpreting what is estimated and shown in Figure 2 as cumulative proportions is highly misleading and, in fact, using the cumulative rate instead of the correct cumulative incidence will overestimate the risk. The correctly estimated cumulative proportions are given in Table 1, namely 2.23 per 1000 and 1.52 per 1000 for the two periods (because when there are no ‘censored observations’, i.e. all pregnancies end in either live birth or fetal death, the Aalen-Johansen estimator is just a simple relative frequency).
1. Hedegaard M, Lidegaard Ø, Skovlund CW, Mørch LS & Hedegaard M. Reduction in stillbirths at term after new birth induction paradigm: Results of a national intervention. BMJ Open 4, 1–8 (2014).
2. Andersen PK, Geskus RB, De witte T & Putter H. Competing risks in epidemiology: Possibilities and pitfalls. Int. J. Epidemiol. 41, 861–870 (2012).
Seo et al. conducted a prospective study to investigate the risk of 3-year all-cause mortality in elderly patients with heart failure with reduced ejection fraction (HFrEF) with special reference to the clinical characteristics and treatment strategies (1). Adjusted hazard ratio (HR) (95% CI) of guideline-directed medical therapy (GDMT) for all-cause mortality was 0.47 (0.39 to 0.57). In addition, adjusted HR (95% CI) of beta-blockers only and renin-angiotensin system inhibitors only for all-cause mortality was 0.57 (0.45 to 0.73) and 0.58 (0.48 to 0.71), respectively. I have a query about their study with special emphasis on sex difference.
Motiejunaite et al. conducted a prospective study to evaluate sex difference in 1-year all-cause mortality in patients with acute heart failure (AHF) (2). In the GREAT registry mostly from Europe and Asia, adjusted hazard ratio (HR) (95% CI) of women against men for 1-year mortality was 0.86 (0.79-0.94), which was obvious in northeast Asia, presenting 0.76 (0.67-0.87). In the OPTIMIZE-HF registry from the USA, adjusted HR (95% CI) of women against men for 1-year mortality was 0.93 (0.89-0.97). They concluded that male patients with AHF had an increased mortality risk than female patients. Although they handled patients with AHF, the follow-up period and GDMT should be paid attention for the mortality risk assessment. In contrast, Seo et al. should conduct stratified analysis by sex.
Seo et al. conducted a prospective study to investigate the risk of 3-year all-cause mortality in elderly patients with heart failure with reduced ejection fraction (HFrEF) with special reference to the clinical characteristics and treatment strategies (1). Adjusted hazard ratio (HR) (95% CI) of guideline-directed medical therapy (GDMT) for all-cause mortality was 0.47 (0.39 to 0.57). In addition, adjusted HR (95% CI) of beta-blockers only and renin-angiotensin system inhibitors only for all-cause mortality was 0.57 (0.45 to 0.73) and 0.58 (0.48 to 0.71), respectively. I have a query about their study with special emphasis on sex difference.
Motiejunaite et al. conducted a prospective study to evaluate sex difference in 1-year all-cause mortality in patients with acute heart failure (AHF) (2). In the GREAT registry mostly from Europe and Asia, adjusted hazard ratio (HR) (95% CI) of women against men for 1-year mortality was 0.86 (0.79-0.94), which was obvious in northeast Asia, presenting 0.76 (0.67-0.87). In the OPTIMIZE-HF registry from the USA, adjusted HR (95% CI) of women against men for 1-year mortality was 0.93 (0.89-0.97). They concluded that male patients with AHF had an increased mortality risk than female patients. Although they handled patients with AHF, the follow-up period and GDMT should be paid attention for the mortality risk assessment. In contrast, Seo et al. should conduct stratified analysis by sex.
Regarding the first query, Lainscak et al. assessed age- and sex-related differences in 1-year all-cause mortality and hospitalization in chronic heart failure patients (3). Rates of use of GDMT such as angiotensin-converting enzyme inhibitors/angiotensin receptor blockers, beta-blockers and mineralocorticoid receptor antagonists were 85.7%, 88.7% and 58.8%, respectively. All-cause mortality and all-cause hospitalization increased with greater age in both sexes, and sex was not an independent predictor of GDMT or 1-year all-cause mortality in patients with LVEF =<45%. This paper presented no sex difference in 1-year mortality in patients with HFrEF. Longer follow-up is needed to verify the sex difference in mortality, especially in elderly patients. In addition, risk of different medications on the prognosis should also be verified.
References
1. Seo WW, Park JJ, Park HA, et al. Guideline-directed medical therapy in elderly patients with heart failure with reduced ejection fraction: a cohort study. BMJ Open. 2020;10(2):e030514. doi: 10.1136/bmjopen-2019-030514
2. Motiejunaite J, Akiyama E, Cohen-Solal A, et al. The association of long-term outcome and biological sex in patients with acute heart failure from different geographic regions. Eur Heart J. 2020;41(13):1357-1364. doi: 10.1093/eurheartj/ehaa071
3. Lainscak M, Milinkovi? I, Polovina M, et al. Sex- and age-related differences in the management and outcomes of chronic heart failure: an analysis of patients from the ESC HFA EORP Heart Failure Long-Term Registry. Eur J Heart Fail. 2020;22(1):92-102. doi: 10.1002/ejhf.1645
I read with interest the article by Karlsen, et al. concluding that here was an association between clinically significant uterine fibroids and pre-term birth. (1) The study has several significant shortcomings. The likely inaccuracies of the use of a large retrospective, administrative, coded database and the low prevalence of fibroids only identified clinically and not confirmed by ultrasound or other imaging modality were mentioned by the authors in the limitations section of the article. However, the degree of these limitations makes the data uninterpretable and the conclusions unfounded.
Administrative databases are inadequate to answer clinical questions as they are subject to clinical misdiagnosis and coding errors which are blind to the goals of the study. The only way to assure accuracy of such a data set is to adjudicate the data by chart review, which was not done by the authors.
Reliance on clinical examination to determine the presence of uterine fibroids, especially submucous fibroids that might more significantly impact pregnancy, is inaccurate. (3) Virtually all women in the developed world currently have an ultrasound as part of their early pregnancy evaluation. Most studies using ultrasound for the diagnosis of fibroids in early pregnancy report a prevalence of 3-4%. One such study found that among 64,047 pregnant women, 2058 had fibroids diagnosed with 1st trimester US, a prevalence of 3.2%. (Stout)
I read with interest the article by Karlsen, et al. concluding that here was an association between clinically significant uterine fibroids and pre-term birth. (1) The study has several significant shortcomings. The likely inaccuracies of the use of a large retrospective, administrative, coded database and the low prevalence of fibroids only identified clinically and not confirmed by ultrasound or other imaging modality were mentioned by the authors in the limitations section of the article. However, the degree of these limitations makes the data uninterpretable and the conclusions unfounded.
Administrative databases are inadequate to answer clinical questions as they are subject to clinical misdiagnosis and coding errors which are blind to the goals of the study. The only way to assure accuracy of such a data set is to adjudicate the data by chart review, which was not done by the authors.
Reliance on clinical examination to determine the presence of uterine fibroids, especially submucous fibroids that might more significantly impact pregnancy, is inaccurate. (3) Virtually all women in the developed world currently have an ultrasound as part of their early pregnancy evaluation. Most studies using ultrasound for the diagnosis of fibroids in early pregnancy report a prevalence of 3-4%. One such study found that among 64,047 pregnant women, 2058 had fibroids diagnosed with 1st trimester US, a prevalence of 3.2%. (Stout)
The current study captured 124 women with fibroids coded among 92,696 pregnancies, for a prevalence of 0.13%. If we assume that the prevalence of fibroids is approximately that found in other studies, then we would expect to find 2,966 pregnancies with concurrent fibroids in the Karlsen study. Therefore, 2,842 pregnancies were likely undetected clinically or improperly coded and thus misclassified as not having fibroids. The authors suggest that mistakenly including women with fibroids in the control group would attenuate the association between fibroids and pregnancy outcomes. I disagree. If these women had adverse outcomes, the findings would be diluted by the outcomes of the other 89,730 pregnancies. However, if the women had favorable outcomes, then that data would significantly decrease the adverse outcome conclusions reported in this study. To emphasize this point, 95% of the data is missing and basing any conclusions on this degree of incomplete data is misleading and irresponsible.
Another important question considered in this study, which remains controversial in the literature, is whether myomectomy prior to pregnancy improves outcomes. However, based on the inaccurate data, I am disappointed that no conclusions can be made for this or any other fibroid-related pregnancy outcome the authors attempted to study.
William H. Parker, MD
Department of Ob Gyn and Reproductive Sciences
UC San Diego
La Jolla, California
(1) Karlsen K, Schiøler Kesmodel U, Mogensen O, Humaidan P, Ravn P. Relationship between a uterine fibroid diagnosis and the risk of adverse obstetrical outcomes: a cohort study. BMJ Open. 2020;10:e032104
(2) L. Flickinger, G. D’Ablaing III, D.R. Mishell Jr. Size and weight determinations of nongravid enlarged uteri Obstet Gynecol, 68 (Dec (6)) (1986), pp. 855-858
(3) Stout MJ, Odibo AO, Graseck AS, Macones GA, Crane JP, Cahill AG. Leiomyomas at routine second-trimester ultrasound examination and adverse obstetric outcomes. Obstet Gynecol. 2010;116:1056-63.
Peng et al. conducted a cross-sectional study to determine the relationship between depressive symptoms and arterial stiffness in general population (1). The Patient Health Questionnaire-9 (PHQ-9) was used to assess the degree of depressive symptoms: 0-4 no depressive symptoms, 5-9 mild depressive symptoms and 10-27 moderate to severe depressive symptoms. Brachial-ankle pulse wave velocity (baPWV) was measured to determine arterial stiffness. By multivariate linear regression analysis, standardized regression coefficients (95% CIs) of mild and moderate to severe depressive symptoms to predict baPWV were 40.3 (6.6-74.1) and 87.7 (24.0-151.5), respectively. In addition, there were significant interactions in subjects whose blood pressures are beyond the optimal range and combined with diabetes mellitus. Two major metabolic components, increased blood pressure and glucose intolerance, were closely associated with the relationship between depressive symptoms and arterial stiffness. I have two concerns about their study.
First, Liu et al. examined the association between arterial stiffness and type 2 diabetes mellitus (T2DM) with special reference to interaction by white blood cell (WBC) counts (2). They observed a dose-response relationship between increased baPWV and elevated risk of T2DM, presenting the adjusted odds ratio (OR) (95% CI) of highest tertile of baPWV for T2DM risk being 2.29 (1.32-3.98). In addition, mediation analyses indicated that total WBC count medi...
Peng et al. conducted a cross-sectional study to determine the relationship between depressive symptoms and arterial stiffness in general population (1). The Patient Health Questionnaire-9 (PHQ-9) was used to assess the degree of depressive symptoms: 0-4 no depressive symptoms, 5-9 mild depressive symptoms and 10-27 moderate to severe depressive symptoms. Brachial-ankle pulse wave velocity (baPWV) was measured to determine arterial stiffness. By multivariate linear regression analysis, standardized regression coefficients (95% CIs) of mild and moderate to severe depressive symptoms to predict baPWV were 40.3 (6.6-74.1) and 87.7 (24.0-151.5), respectively. In addition, there were significant interactions in subjects whose blood pressures are beyond the optimal range and combined with diabetes mellitus. Two major metabolic components, increased blood pressure and glucose intolerance, were closely associated with the relationship between depressive symptoms and arterial stiffness. I have two concerns about their study.
First, Liu et al. examined the association between arterial stiffness and type 2 diabetes mellitus (T2DM) with special reference to interaction by white blood cell (WBC) counts (2). They observed a dose-response relationship between increased baPWV and elevated risk of T2DM, presenting the adjusted odds ratio (OR) (95% CI) of highest tertile of baPWV for T2DM risk being 2.29 (1.32-3.98). In addition, mediation analyses indicated that total WBC count mediated 4.5% of the association between increased baPWV and elevated T2DM risk. Although small mediation effect was identified, inflammation might interact on the association between baPWV and elevated risk of T2DM. Other inflammation markers should also be evaluated by further studies.
Second, Xu et al. examined the association between the skeletal muscle mass-to-visceral fat area ratio (SVR) and arterial stiffness in patients with T2DM (3). Adjusted ORs of the lowest SVR tertile against highest for high baPWV were 4.33 in men and 4.66 in women. They also evaluated by receiver operating characteristic curve analysis that the optimal cutoffs values of SVR for detecting high baPWV were 191.7 g/cm2 for men and 157.3 g/cm2 for women. Obesity is a fundamental metabolic component, and SVR was closely associated with arterial stiffness in patients with T2DM. Although cross-sectional study cannot present causal direction, inter-relationship among metabolic components are suspected to be related to arterial stiffness. In addition, psycho-physiological relationship should be paid with caution in subjects with T2DM.
References
1. Peng L, Bi S, Liu X, et al. Association between depressive symptoms and arterial stiffness: a cross-sectional study in the general Chinese population. BMJ Open. 2020;10(2):e033408. doi: 10.1136/bmjopen-2019-033408
2. Liu Y, Lai X, Guo W, et al. Total white blood cell count mediated the association between increased arterial stiffness and risk of type 2 diabetes mellitus in Chinese adults. Arterioscler Thromb Vasc Biol. 2020;40(4):1009-1015. doi: 10.1161/ATVBAHA.119.313880
3. Xu J, Pan X, Liang H, et al. Association between skeletal muscle mass to visceral fat area ratio and arterial stiffness in Chinese patients with type 2 diabetes mellitus. BMC Cardiovasc Disord. 2018;18(1):89. doi: 10.1186/s12872-018-0827-z
The suggestion that cloth masks can lead to increased infection compared to no mask is not substantiated The control arm had less than 1% of no mask use and therefore the statement is an assumption not a proof . It could be possible for bacteria or fungi that could multiply on the mask but for virus it would have to shown there is more bioavailability than no mask .
The authors do not state what kind / standard of cloth masks were used. Is there any chance they could do this. Given the huge shortage of FFP2 and FFP3 masks, it would be helpful to know the authors’ views of advice from the CDC in the US. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/diy-cloth...
Katzke et al. conducted a prospective study to assess the relationships of alkaline phosphatase (ALP), gamma-glutamyltransferase (GGT), alanine aminotransferase (ALT), aspartate transaminase (AST) and the De Ritis ratio (AST/ALT) with mortalities (1). There were significant positive associations of all-cause mortality with ALP, GGT and AST. In addition, a combined liver risk score had positive associations with all-cause and cause-specific mortality. I have two concerns about their study.
First, Oh et al. conducted a cross-sectional study to investigate the association between renal hyperfiltration and serum ALP level (2). Renal hyperfiltration was defined as exceeding the age- and sex-specific 97.5th percentile, and odds ratio (95% confidence interval [CI]) of the highest ALP quartile for renal hyperfiltration was 1.624 (1.204-2.192). Higher ALP levels are significantly associated with renal hyperfiltration, and renal hyperfiltration was closely associated with several metabolic disorders. As renal hyperfiltration can be considered as early-stage of renal damage, serum ALP level might be related to renal damage. Mechanism of the association should be specified by further study.
Second, Kunutsor et al. conducted a meta-analysis to evaluate the associations of baseline levels of liver enzymes with all-cause mortality in general populations (3). The pooled relative risks (95% CI) per 5 U/l increment in GGT and ALP levels were 1.07 (1.04-1.10) and 1.03 (1.01-1.0...
Katzke et al. conducted a prospective study to assess the relationships of alkaline phosphatase (ALP), gamma-glutamyltransferase (GGT), alanine aminotransferase (ALT), aspartate transaminase (AST) and the De Ritis ratio (AST/ALT) with mortalities (1). There were significant positive associations of all-cause mortality with ALP, GGT and AST. In addition, a combined liver risk score had positive associations with all-cause and cause-specific mortality. I have two concerns about their study.
First, Oh et al. conducted a cross-sectional study to investigate the association between renal hyperfiltration and serum ALP level (2). Renal hyperfiltration was defined as exceeding the age- and sex-specific 97.5th percentile, and odds ratio (95% confidence interval [CI]) of the highest ALP quartile for renal hyperfiltration was 1.624 (1.204-2.192). Higher ALP levels are significantly associated with renal hyperfiltration, and renal hyperfiltration was closely associated with several metabolic disorders. As renal hyperfiltration can be considered as early-stage of renal damage, serum ALP level might be related to renal damage. Mechanism of the association should be specified by further study.
Second, Kunutsor et al. conducted a meta-analysis to evaluate the associations of baseline levels of liver enzymes with all-cause mortality in general populations (3). The pooled relative risks (95% CI) per 5 U/l increment in GGT and ALP levels were 1.07 (1.04-1.10) and 1.03 (1.01-1.06), presenting linear dose-response relationships. As the magnitude of effect sizes are not so large, further studies are required to verify the association.
References
1. Katzke V, Johnson T, Sookthai D, et al. Circulating liver enzymes and risks of chronic diseases and mortality in the prospective EPIC-Heidelberg case-cohort study. BMJ Open. 2020;10(3):e033532. doi: 10.1136/bmjopen-2019-033532
2. Oh SW, Han KH, Han SY. Associations between renal hyperfiltration and serum alkaline phosphatase. PLoS One. 2015;10(4):e0122921. doi: 10.1371/journal.pone.0122921
3. Kunutsor SK, Apekey TA, Seddoh D, et al. Liver enzymes and risk of all-cause mortality in general populations: a systematic review and meta-analysis. Int J Epidemiol. 2014;43(1):187-201. doi: 10.1093/ije/dyt192
Galland-Decker et al. conducted a cross-sectional study to assess factors associated with fatigue in the middle-aged general population (1). Adjusted odds ratios (ORs) (95% confidence intervals [CIs]) of obesity, depression, and anaemia for fatigue were 1.40 (1.03-1.91), 3.26 (2.38-4.46), and 1.70 (1.00-2.89), respectively. I have two concerns about their study with special reference to sex difference and sleep parameters.
First, Broström et al. conducted a cross-sectional study to examine the associations between self-reported sleep duration, depressive symptoms, anxiety, fatigue and daytime sleepiness in older inhabitants (2). Subjects were divided into short sleepers (≤6 hours), normal sleepers (7-8 hours), and long sleepers (≥9 hours). Depressive symptoms were associated with short sleep in men, and fatigue was associated with both short and long sleep duration in men. As there is a significant association between sleep duration, depressive symptoms and fatigue only in men, mechanism of sex difference on the association should be specified by prospective studies, including sex-related variables.
Second, Lang et al. examined the effect of co-morbid insomnia on depression in community-dwelling middle-aged men (3). Insomnia was defined as difficulty initiating/maintaining sleep (DIMS). Co-morbid insomnia with daytime fatigue/OSA was defined as DIMS-F/COMISA. Standardized regression coefficients (95% CIs) of DIMS-F and COMISA for depression, using Patient Hea...
Galland-Decker et al. conducted a cross-sectional study to assess factors associated with fatigue in the middle-aged general population (1). Adjusted odds ratios (ORs) (95% confidence intervals [CIs]) of obesity, depression, and anaemia for fatigue were 1.40 (1.03-1.91), 3.26 (2.38-4.46), and 1.70 (1.00-2.89), respectively. I have two concerns about their study with special reference to sex difference and sleep parameters.
First, Broström et al. conducted a cross-sectional study to examine the associations between self-reported sleep duration, depressive symptoms, anxiety, fatigue and daytime sleepiness in older inhabitants (2). Subjects were divided into short sleepers (≤6 hours), normal sleepers (7-8 hours), and long sleepers (≥9 hours). Depressive symptoms were associated with short sleep in men, and fatigue was associated with both short and long sleep duration in men. As there is a significant association between sleep duration, depressive symptoms and fatigue only in men, mechanism of sex difference on the association should be specified by prospective studies, including sex-related variables.
Second, Lang et al. examined the effect of co-morbid insomnia on depression in community-dwelling middle-aged men (3). Insomnia was defined as difficulty initiating/maintaining sleep (DIMS). Co-morbid insomnia with daytime fatigue/OSA was defined as DIMS-F/COMISA. Standardized regression coefficients (95% CIs) of DIMS-F and COMISA for depression, using Patient Health Questionnaire 9-item version, were 2.3 (1.2-3.5) and 4.1 (3.0-5.1), respectively. Significant Increase of depression by co-morbid insomnia with fatigue/OSA was clarified in this study, and sleep parameters should be included for the risk assessment of fatigue. In addition, combination effect on depression should also be checked in middle aged women.
References
1. Galland-Decker C, Marques-Vidal P, Vollenweider P. Prevalence and factors associated with fatigue in the Lausanne middle-aged population: a population-based, cross-sectional survey. BMJ Open. 2019;9(8):e027070. doi: 10.1136/bmjopen-2018-027070
2. Broström A, Wahlin Å, Alehagen U, et al. Sex-specific associations between self-reported sleep duration, depression, anxiety, fatigue and daytime sleepiness in an older community-dwelling population. Scand J Caring Sci. 2018;32(1):290-298. doi: 10.1111/scs.12461
3. Lang CJ, Appleton SL, Vakulin A, et al. Co-morbid OSA and insomnia increases depression prevalence and severity in men. Respirology. 2017;22(7):1407-1415. doi: 10.1111/resp.13064
We read Dr. Karlsen’s paper entitled “Relationship between a uterine fibroid diagnosis and the risk of adverse obstetrical outcomes: a cohort study” with great interest. We understand the authors evaluated a large and complicated set of data and appreciate their effort. However, the authors used a database fraught with errors without validation of said database, made incorrect assumptions and conclusions about the data, and incompletely discussed some of the major queries in the discussion.
The first issue surrounds the use of the Danish National Birth Cohort, Patient Registry and Birth Registry. Although the authors write about the limitations of a registry that uses codes “more closely connected to hospital budgets than hospital diagnosis codes”, they rely on a single study to prove the quality of their database.
Multiple authors have shown that administrative databases are inadequate to answer clinical questions. They are are set up for billing purposes only, not epidemiological research (1,2), and elaborate statistical machinations do not overcome the impact of anomalous data (3). Several studies have shown that the very registry used in this report; the Danish National Patient Registry, is highly inaccurate from a clinical standpoint. The authors from a study in 2009 have a self-proclaimed 10% inaccuracy rate in their diagnoses addressing oral contraceptive research (4). Other authors have found higher erroneous rates for diagnoses of hypertension, rhe...
We read Dr. Karlsen’s paper entitled “Relationship between a uterine fibroid diagnosis and the risk of adverse obstetrical outcomes: a cohort study” with great interest. We understand the authors evaluated a large and complicated set of data and appreciate their effort. However, the authors used a database fraught with errors without validation of said database, made incorrect assumptions and conclusions about the data, and incompletely discussed some of the major queries in the discussion.
The first issue surrounds the use of the Danish National Birth Cohort, Patient Registry and Birth Registry. Although the authors write about the limitations of a registry that uses codes “more closely connected to hospital budgets than hospital diagnosis codes”, they rely on a single study to prove the quality of their database.
Multiple authors have shown that administrative databases are inadequate to answer clinical questions. They are are set up for billing purposes only, not epidemiological research (1,2), and elaborate statistical machinations do not overcome the impact of anomalous data (3). Several studies have shown that the very registry used in this report; the Danish National Patient Registry, is highly inaccurate from a clinical standpoint. The authors from a study in 2009 have a self-proclaimed 10% inaccuracy rate in their diagnoses addressing oral contraceptive research (4). Other authors have found higher erroneous rates for diagnoses of hypertension, rheumatoid arthritis and uterine rupture (5). Another independent review of the database revealed a 41% misclassification rate in women with alleged hormone-related thromboembolic events (6) In fact, in 2010, Shapiro et al (7) recommended ignoring data from unconfirmed data base reports, citing unreliable data from the Danish National Patient Registry.
The attempt by the authors to compare women with symptomatic uterine fibroids is laudable, as this has not been covered in the existing literature, but unfortunately it is not clear that they were able to accomplish this. Simply having a diagnostic code for “fibroid” does not ensure that the fibroid was indeed symptomatic; in fact, an unknown number may have been given the diagnosis due to ultrasound detection in the absence of symptoms. Unfortunately, the authors did not attempt to validate this assumption.
There are other issues with this manuscript. Hysteroscopic myomectomy was presumed if the operation codes included “resection and excision of pathological tissue”, but we do not know this to be the case, and again no attempt at validation was conducted by the authors. When a fibroid diagnosis was given, there was no concern regarding number or size of fibroids, both potential confounders that were not corrected for in the regression analysis. Finally, the authors included women with multiple separate pregnancies, stating that this was valid due to “robust standard errors”. Which standard errors were robust, how robust were they, and why did this correct for an obvious methodologic flaw of including some women multiple times?
The conclusions reached by the authors regarding preterm delivery and caesarean section are incompletely reviewed in the discussion. To date, ten studies have addressed the effect of fibroids on the rate of preterm delivery, with six showing an increased risk with fibroids and four showing no difference. However, five of the six that adjusted the risk ratio for known confounders found an increase, similarly found this study (8). Including this information would have strengthened their findings. Regarding caesarean section, 13 of 15 published studies show an increase in the presence of fibroids, including all 11 studies that adjusted for confounders via regression analysis (as does this manuscript) (8). The authors, however, do not focus on cesarean sections that were acute in nature; these were not different between the two groups. This is an interesting and novel finding, implicating surgeon or patient choice as the driving force behind increased cesarean sections in women with fibroids.
Whether myomectomy prior to pregnancy improves outcomes for women and their children is the crucial unanswered question in the current literature. The authors state that risk of preterm birth is decreased, and cesarean section rates increased among women with untreated fibroids compared with women with surgical excision. However, this appears to be based upon raw numbers and not statistical analysis, as the two groups (fibroids in situ versus fibroids removed) were never compared. At first glance, it appears that there is no significant difference between these groups for any outcome measure, although without appropriate statistical analysis we cannot be sure. This shortcoming diminishes the value in addressing this important issue.
In conclusion, this paper has important methodologic flaws, and deficient conclusions. We would be interested in further publications from these authors.
(1) Grimes DA. Epidemiologic research using administrative databases. Obstet Gynecol 2010;116(5):1018-19.
(2) Grimes DA. Epidemiologic research with administrative databases: red herrings, false alarms and pseudo-epidemics. Hum Reprod 2015;30(8):1749-52.
(3) Ault MR. Combating the garbage-in, gospel-out syndrome. Rad Protect Manage 2004;20:26-30.
(4) Lidegaard O, Lokkegaard, E, Svendsen AL, Agger C. Hormonal contraception and risk of venous thromboembolism: national follow-up study. BMJ 2009;339:b2890.
(5) Pedersen M, Klarlund M, Jacobsen S, Svendsen AJ, Frisch M. Validity of rheumatoid arthritis diagnoses in the Danish National Patient Registry. Eur J Epidemiol 2004;19:1097-1103.
(6) Severinsen MT, Kristensen SR, Overvad K, Dethlefsen C, Tjonneland A, Johnsen SP . Venous thromboembolism discharge diagnoses in the Danish National Patient Registry should be use with caution. J Clin Epidemiol 2010;63:223-8.
(7) Shapiro S, Dinger J. Risk of venous thromboembolism among users of oral contraceptives: a review of two recently published studies. J Fam Plann Reprod Health Care 2010;36:33-8.
(8) Pritts EA, Olive DL. Fibroids and reproduction. In Modern Management of Uterine Fibroids. Cambridge University Press, 2020.
Lin et al. conducted across-sectional study to evaluate the relationship between the prevalence of hyperuricaemia (HUA) and non-valvular atrial fibrillation (NVAF) (1). Age, living in urban areas, alcohol consumption, central obesity, elevated fasting plasma glucose level, elevated blood pressure, lower high-density lipoprotein cholesterol level and elevated triglycerides level were significantly associated with increased risk of HUA, and HUA were at higher risk for NVAF. Regarding sex difference, serum uric acid level had a modest predictive value for NVAF in women. I have some concerns about their study.
First, Tu et al. conducted a prospective study to evaluate the risk of gout in patients with alcohol-related diseases and alcohol dependence syndrome (2). Alcohol-related diseases were significantly associated with gout risk, and severe alcohol-dependent patients were significantly associated with an increased risk of gout. The authors recommended that alcohol use assessment and measures to prevent alcohol dependence in patients with gout. I agree with their recommendation, which would partly lead to prevent NVAF.
Second, Yu et al. presented strategies to improve gout/hyperuricemia management for preventing acute arthritis attacks, cardiovascular disease, kidney disease, and other urate-related disorders (3). Alcohol consumption is closely related to gout/hyperuricemia, and psycho-social factors should also be considered for the management.
Third, Ku...
Show MoreIt has been five years since Hedegaard and colleagues published their paper on stillbirth and labour induction [1]. In Denmark, its findings are still often brought forward as weighty documentation in public and professional discussions on the appropriate timing of recommended induction of labour in low-risk pregnancies after due date. Thus, as late as in January 2020, one of the main Danish broadcasting companies (TV2) covered the subject, and a curve almost identical to the top graph of the study’s Figure 2 was presented, alongside statements from one the authors. The graph was entitled "Number of dead fetuses per 1000 pregnant women for the period 2000-2008", and the accompanying text read “[…] the numbers show that the number of dead fetuses increases markedly by the beginning of 41 completed gestational weeks and even more after 42 weeks. […] the number of fetuses that died in the mother’s womb after 37th week decreased from 120 to 60 when the regulations were changed in 2011.”
Following its publication in 2014, the study was subject to much attention from peers. The responses focused on data quality, methodological issues, and whether data did or did not support the authors’ conclusions. However, the problem regarding competing risks was not touched upon, and since ignoring this problem will make the risk appear greater than it actually is, we find it crucial to point out that competing risks was not taken into account in this study’s assessment of...
Show MoreSeo et al. conducted a prospective study to investigate the risk of 3-year all-cause mortality in elderly patients with heart failure with reduced ejection fraction (HFrEF) with special reference to the clinical characteristics and treatment strategies (1). Adjusted hazard ratio (HR) (95% CI) of guideline-directed medical therapy (GDMT) for all-cause mortality was 0.47 (0.39 to 0.57). In addition, adjusted HR (95% CI) of beta-blockers only and renin-angiotensin system inhibitors only for all-cause mortality was 0.57 (0.45 to 0.73) and 0.58 (0.48 to 0.71), respectively. I have a query about their study with special emphasis on sex difference.
Motiejunaite et al. conducted a prospective study to evaluate sex difference in 1-year all-cause mortality in patients with acute heart failure (AHF) (2). In the GREAT registry mostly from Europe and Asia, adjusted hazard ratio (HR) (95% CI) of women against men for 1-year mortality was 0.86 (0.79-0.94), which was obvious in northeast Asia, presenting 0.76 (0.67-0.87). In the OPTIMIZE-HF registry from the USA, adjusted HR (95% CI) of women against men for 1-year mortality was 0.93 (0.89-0.97). They concluded that male patients with AHF had an increased mortality risk than female patients. Although they handled patients with AHF, the follow-up period and GDMT should be paid attention for the mortality risk assessment. In contrast, Seo et al. should conduct stratified analysis by sex.
Regarding the first query, Lainscak et...
Show MoreTo the Editor,
I read with interest the article by Karlsen, et al. concluding that here was an association between clinically significant uterine fibroids and pre-term birth. (1) The study has several significant shortcomings. The likely inaccuracies of the use of a large retrospective, administrative, coded database and the low prevalence of fibroids only identified clinically and not confirmed by ultrasound or other imaging modality were mentioned by the authors in the limitations section of the article. However, the degree of these limitations makes the data uninterpretable and the conclusions unfounded.
Administrative databases are inadequate to answer clinical questions as they are subject to clinical misdiagnosis and coding errors which are blind to the goals of the study. The only way to assure accuracy of such a data set is to adjudicate the data by chart review, which was not done by the authors.
Reliance on clinical examination to determine the presence of uterine fibroids, especially submucous fibroids that might more significantly impact pregnancy, is inaccurate. (3) Virtually all women in the developed world currently have an ultrasound as part of their early pregnancy evaluation. Most studies using ultrasound for the diagnosis of fibroids in early pregnancy report a prevalence of 3-4%. One such study found that among 64,047 pregnant women, 2058 had fibroids diagnosed with 1st trimester US, a prevalence of 3.2%. (Stout)
The c...
Show MorePeng et al. conducted a cross-sectional study to determine the relationship between depressive symptoms and arterial stiffness in general population (1). The Patient Health Questionnaire-9 (PHQ-9) was used to assess the degree of depressive symptoms: 0-4 no depressive symptoms, 5-9 mild depressive symptoms and 10-27 moderate to severe depressive symptoms. Brachial-ankle pulse wave velocity (baPWV) was measured to determine arterial stiffness. By multivariate linear regression analysis, standardized regression coefficients (95% CIs) of mild and moderate to severe depressive symptoms to predict baPWV were 40.3 (6.6-74.1) and 87.7 (24.0-151.5), respectively. In addition, there were significant interactions in subjects whose blood pressures are beyond the optimal range and combined with diabetes mellitus. Two major metabolic components, increased blood pressure and glucose intolerance, were closely associated with the relationship between depressive symptoms and arterial stiffness. I have two concerns about their study.
First, Liu et al. examined the association between arterial stiffness and type 2 diabetes mellitus (T2DM) with special reference to interaction by white blood cell (WBC) counts (2). They observed a dose-response relationship between increased baPWV and elevated risk of T2DM, presenting the adjusted odds ratio (OR) (95% CI) of highest tertile of baPWV for T2DM risk being 2.29 (1.32-3.98). In addition, mediation analyses indicated that total WBC count medi...
Show MoreThe suggestion that cloth masks can lead to increased infection compared to no mask is not substantiated The control arm had less than 1% of no mask use and therefore the statement is an assumption not a proof . It could be possible for bacteria or fungi that could multiply on the mask but for virus it would have to shown there is more bioavailability than no mask .
The authors do not state what kind / standard of cloth masks were used. Is there any chance they could do this. Given the huge shortage of FFP2 and FFP3 masks, it would be helpful to know the authors’ views of advice from the CDC in the US. https://www.cdc.gov/coronavirus/2019-ncov/prevent-getting-sick/diy-cloth...
Katzke et al. conducted a prospective study to assess the relationships of alkaline phosphatase (ALP), gamma-glutamyltransferase (GGT), alanine aminotransferase (ALT), aspartate transaminase (AST) and the De Ritis ratio (AST/ALT) with mortalities (1). There were significant positive associations of all-cause mortality with ALP, GGT and AST. In addition, a combined liver risk score had positive associations with all-cause and cause-specific mortality. I have two concerns about their study.
First, Oh et al. conducted a cross-sectional study to investigate the association between renal hyperfiltration and serum ALP level (2). Renal hyperfiltration was defined as exceeding the age- and sex-specific 97.5th percentile, and odds ratio (95% confidence interval [CI]) of the highest ALP quartile for renal hyperfiltration was 1.624 (1.204-2.192). Higher ALP levels are significantly associated with renal hyperfiltration, and renal hyperfiltration was closely associated with several metabolic disorders. As renal hyperfiltration can be considered as early-stage of renal damage, serum ALP level might be related to renal damage. Mechanism of the association should be specified by further study.
Second, Kunutsor et al. conducted a meta-analysis to evaluate the associations of baseline levels of liver enzymes with all-cause mortality in general populations (3). The pooled relative risks (95% CI) per 5 U/l increment in GGT and ALP levels were 1.07 (1.04-1.10) and 1.03 (1.01-1.0...
Show MoreGalland-Decker et al. conducted a cross-sectional study to assess factors associated with fatigue in the middle-aged general population (1). Adjusted odds ratios (ORs) (95% confidence intervals [CIs]) of obesity, depression, and anaemia for fatigue were 1.40 (1.03-1.91), 3.26 (2.38-4.46), and 1.70 (1.00-2.89), respectively. I have two concerns about their study with special reference to sex difference and sleep parameters.
First, Broström et al. conducted a cross-sectional study to examine the associations between self-reported sleep duration, depressive symptoms, anxiety, fatigue and daytime sleepiness in older inhabitants (2). Subjects were divided into short sleepers (≤6 hours), normal sleepers (7-8 hours), and long sleepers (≥9 hours). Depressive symptoms were associated with short sleep in men, and fatigue was associated with both short and long sleep duration in men. As there is a significant association between sleep duration, depressive symptoms and fatigue only in men, mechanism of sex difference on the association should be specified by prospective studies, including sex-related variables.
Second, Lang et al. examined the effect of co-morbid insomnia on depression in community-dwelling middle-aged men (3). Insomnia was defined as difficulty initiating/maintaining sleep (DIMS). Co-morbid insomnia with daytime fatigue/OSA was defined as DIMS-F/COMISA. Standardized regression coefficients (95% CIs) of DIMS-F and COMISA for depression, using Patient Hea...
Show MoreWe read Dr. Karlsen’s paper entitled “Relationship between a uterine fibroid diagnosis and the risk of adverse obstetrical outcomes: a cohort study” with great interest. We understand the authors evaluated a large and complicated set of data and appreciate their effort. However, the authors used a database fraught with errors without validation of said database, made incorrect assumptions and conclusions about the data, and incompletely discussed some of the major queries in the discussion.
The first issue surrounds the use of the Danish National Birth Cohort, Patient Registry and Birth Registry. Although the authors write about the limitations of a registry that uses codes “more closely connected to hospital budgets than hospital diagnosis codes”, they rely on a single study to prove the quality of their database.
Multiple authors have shown that administrative databases are inadequate to answer clinical questions. They are are set up for billing purposes only, not epidemiological research (1,2), and elaborate statistical machinations do not overcome the impact of anomalous data (3). Several studies have shown that the very registry used in this report; the Danish National Patient Registry, is highly inaccurate from a clinical standpoint. The authors from a study in 2009 have a self-proclaimed 10% inaccuracy rate in their diagnoses addressing oral contraceptive research (4). Other authors have found higher erroneous rates for diagnoses of hypertension, rhe...
Show MorePages