We use cookies to improve our service and to tailor our content and advertising to you.More infoYou can manage your cookie settings via your browser at any time. To learn more about how we use cookies, please see our cookies policy.
Dear editor,
We have read with great interest the meta-analysis submitted by Li and colleagues1, which investigated the association between fish consumption and depression risk. We warmly and greatly congratulate and applaud for their important work. However, an issue existed in this study should be noted.
These authors stated that observational study including cross-sectional, case-control, and cohort study was eligible for their...
Dear editor,
We have read with great interest the meta-analysis submitted by Li and colleagues1, which investigated the association between fish consumption and depression risk. We warmly and greatly congratulate and applaud for their important work. However, an issue existed in this study should be noted.
These authors stated that observational study including cross-sectional, case-control, and cohort study was eligible for their inclusion criteria in inclusion criteria subsection. However, the Preferred Reporting Items for Systematic Review and Meta-analyses (PRISMA)2 was used to guide reporting their meta-analysis. Important is that the PRISMA is designed to use mainly in systematic review and meta-analysis with randomized controlled trials (RCTs) rather than meta-analysis with observational studies in epidemiology. Controversially, the Meta-analysis of Observational Studies in Epidemiology (MOOSE)3 is developed for reporting this given meta-analysis. And thus, the authors should adopted the MOOSE to guide reporting their meta-analysis on this given topic preferably we suggested in order to further improve reporting quality.
1. Li F, Liu X, Zhang D. Fish consumption and risk of depression: a meta-analysis. J Epidemiol Community Health 2015; doi: 10.1136/jech-2015-206278.
2. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meya-analyses: the PRISMA statement. J Clin Epidemiol 2009; 62:1006-1012.
3. Stroup DF, Berlin JA, Moton SC, et al. Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis of Observational Studies in Epidemiology (MOOSE) group. JAMA 2000; 283:2008-2012.
Hintsa et al. examined the effect of effort, reward and job control
on the exit from the labour market by a 6-year follow-up study in workers
at the age of 61 years or younger [1]. The author adopted binary logistic
regression analysis by adjusting several variables, and concluded that
effort-reward imbalance (ERI), effort and job control were significant
predictors for exit from the labour market. In contrast, reward wa...
Hintsa et al. examined the effect of effort, reward and job control
on the exit from the labour market by a 6-year follow-up study in workers
at the age of 61 years or younger [1]. The author adopted binary logistic
regression analysis by adjusting several variables, and concluded that
effort-reward imbalance (ERI), effort and job control were significant
predictors for exit from the labour market. In contrast, reward was not
selected as a significant predictor. I have some concerns on their study.
First, the authors did not use effort, reward and job control
simultaneously as independent variables for predicting exit from the
labour market. Schmidt et al. recently reported that the sum score of
effort significantly increased and the sum score of reward significantly
decreased as ERI increased in 4141 samples in Germany [2]. If
multicollinearity among independent variables cannot be solved, the
simultaneous use of ERI, effort and reward should be handled with caution,
because ERI was simply calculated as the logarithmic value of [(sum score
of effort)*7]/[(sum score of reward)*3]. But inset of effort and reward
simultaneously as independent variables into logistic model seems
appropriate, because the authors measured stressful psychological work
environment by ERI model, and each factor has a different dimension for
psychometry. In addition, simultaneous use of job control from another
theoretical stress model should also be considered in combination with ERI
model after evaluating multicollinearity.
Second, statistical results differed by different combination of
adjusting variables in their study. The adjusting variables were selected
to know the net association between exit from the labour market and
effort, reward or job control, and I recommend selecting full-adjudging
model in their study. If discrepancies of statistical results by selecting
different combination of adjusting variable exist, stability of
significance cannot be guaranteed in the statistical model.
Finally, there are other reasons of exit from the labour market than
the causes of working environment. On this point, explanation rate of exit
from the labour market by factors form ERI model and job control should be
presented. The authors selected workers at the age of 61 years or younger,
and cause of exit by family support is suspected especially in women. This
would partly be related to the sex differences in the average age of
withdrawal from the labour market. Anyway, further study is commended to
confirm the causal association.
References
1 Hintsa T, Kouvonen A, McCann M, et al. Higher effort-reward
imbalance and lower job control predict exit from the labour market at the
age of 61 years or younger: evidence from the English Longitudinal Study
of Ageing. J Epidemiol Community Health 2015;69:543-9.
2 Schmidt B, Bosch JA, Jarczok MN, et al. Effort-reward imbalance is
associated with the metabolic syndrome - findings from the Mannheim
Industrial Cohort Study (MICS). Int J Cardiol 2015;178:24-8.
Xu et al report an association between deranged liver enzymes e.g. alanine transaminase (ALT) in Chinese people residing in Southern China and the incidence of diabetes [1].
The authors examined multiple potential confounding factors which could influence liver enzymes e.g. alcohol consumption & adiposity.
However, although the authors briefly mentioned chronic viral hepatitis infections i.e. chronic hepatitis B...
Xu et al report an association between deranged liver enzymes e.g. alanine transaminase (ALT) in Chinese people residing in Southern China and the incidence of diabetes [1].
The authors examined multiple potential confounding factors which could influence liver enzymes e.g. alcohol consumption & adiposity.
However, although the authors briefly mentioned chronic viral hepatitis infections i.e. chronic hepatitis B virus (HBV) & hepatitis C virus (HCV), I think they did not give this confounder sufficient weight and did not report HBV/HCV prevalence data in the cohort studied. This is a significant shortcoming of this study.
As the authors acknowledged, chronic HBV & HCV are relatively common in China, with the prevalence of chronic HBV of 11.3% for males, 8.2% for females & for chronic HCV 3.1% for males, 3.3% for females [2].
Therefore, as these are common infections in China, the underlying population prevalence of chronic HBV & HCV infection in the study cohort could account for at least some of the deranged liver enzymes in the study population. Unfortunately, the lack of data on these infections prevents quantitative analysis of the effect these hepatitides may have on the liver enzyme profiles in this cohort.
The authors acknowledged a possible association between HCV and diabetes, but there is also emerging evidence that chronic HBV infection is associated with insulin resistance [3]. Furthermore, chronic HBV complicated by cirrhosis may be associated with diabetes mellitus [4]. The apparent association between deranged liver enzymes and the development of diabetes may not be due to hepatitis/transaminitis per se, but may reflect underlying specific pathologies such as chronic HBV and HCV.
The high prevalence of chronic HBV and HCV in the Chinese population and the associations of these infections with insulin resistance & diabetes mellitus means that the authors' conclusions underplay the possible contribution of chronic viral hepatitis to the development of diabetes. Thus, the conclusions should perhaps have included recommendations that patients with deranged liver enzymes should be screened for these viral hepatitis infections, so that they can be considered for antiviral treatment which would be beneficial not just in terms of future diabetes risk, but also the well-known risks of cirrhosis & hepatocellular carcinoma in these patients.
References:
1. Xu L, Jian CQ, Schooling CM et al. Liver enzymes and incident diabetes in China: a prospective analysis of 10764 participants in the Guangzhou Biobank Cohort Study. J Epidemiol Community Health 2015;69:1040-1044
2. Huang H, Hu XF, Zhao FH et al. Estimation of cancer burden attributable to infection in Asia. J Epidemiol 2015 doi:10.2188/jea.JE20140215
3. Lee JG, Lee S, Kim YJ et al. Association of chronic viral hepatitis B with insulin resistance. World J Gastro 2012;18(42):6120-6126
4. Zhang J, Shen Y, Cai H et al. Hepatitis B virus infection status and risk of type 2 diabetes mellitus: A meta-analysis. Hepatol Res 2015 doi:10.1111.hepr.12481
Oksuzyan et al. report an association between race/ethnicity and two
subtypes of childhood leukemia: acute lymphoblastic leukaemia (ALL) and
acute myeloid leukaemia (AML).1 Accordingly, the researchers suggest that
there are genetic, cultural, and environmental factors involved in the
etiology of childhood leukaemia [1].
Importantly, Oksuzyan et al. made a significant effort to examine and
control for the poten...
Oksuzyan et al. report an association between race/ethnicity and two
subtypes of childhood leukemia: acute lymphoblastic leukaemia (ALL) and
acute myeloid leukaemia (AML).1 Accordingly, the researchers suggest that
there are genetic, cultural, and environmental factors involved in the
etiology of childhood leukaemia [1].
Importantly, Oksuzyan et al. made a significant effort to examine and
control for the potential role of socioeconomic status (SES) on this
association. In particular, the authors used paternal education levels
used as a proxy for SES given that a significant percentage of information
on maternal education was missing from the birth registry used [1]. This
strategy faces several challenges given the complex association of race,
education, and SES.
The use of paternal education might not appropriately account for the
adverse health effects which might result (at least in part) from residing
in single-parent homes. It has been widely documented that children
frequently exhibit poor health outcomes in single-parent homes due to
causal chain of effects related to parent's education, SES, and income
[2]. Moreover, the information regarding paternal education as a proxy for
SES was obtained from the California Birth Registry [1]. It is possible
that this data does not adequately represent the SES of the cases used in
the study given that the father identified might not have provided
financial support for the child.
In addition, the use of parental education as a proxy for SES may not
be an appropriate method as average income frequently varies in jobs
requiring similar education levels [3]. It is possible that these
variations occur more frequently among different ethnic/racial groups;
potentially as a result of systemic prejudice and/or unequal employment
opportunities [3].
Oksuzyan et al. report an association between race/ethnicity and
childhood leukaemia [1]. Due to potential challenges in using paternal
education as a proxy for SES, the inclusion of maternal education and a
discussion regarding SES in single-parent homes would have been valuable.
References
1 Oksuzyan S, Crespi CM, Cockburn M, Mezei G, Vergara X, Kheifets L.
Race/ethnicity and the risk of childhood leukaemia: A case-control study
in California. J Epidemiol Community Health Published Online First: 19
March 2015. doi:10.1136/jech-2014-204975
2 Gucciardi E, Celasun N, Stewart DE. Single-mother families in
Canada. Can J Public Health 2004;95(1):70-73.
3 Williams DR. Race, socioeconomic status, and health. The added
effects of racism and discrimination. Ann N Y Acad Sci 1999;896:173-188.
The author commented on the paucity of research on the point where
physical exercise in polluted air becomes more harmful than beneficial. I
would like to share our research findings conducted more than 10 years ago
in Hong Kong (Yu et al, 2004). We compared the physical fitness of school
children who regularly performed physical exercise with those who did not.
In a less polluted district (annual mean P...
The author commented on the paucity of research on the point where
physical exercise in polluted air becomes more harmful than beneficial. I
would like to share our research findings conducted more than 10 years ago
in Hong Kong (Yu et al, 2004). We compared the physical fitness of school
children who regularly performed physical exercise with those who did not.
In a less polluted district (annual mean PM10=44.9 ug/m3), children who
regularly did physical exercise had significantly better cardiopulmonary
fitness (with a higher predicted maximum oxygen update of 1.8 mL/Kg/min
among children who did regular exercise). By contrast, among children in a
'high pollution district' (PM10=57.6 ug/m3), there was no significant
difference in their cardiopulmonary fitness whether they exercised
regularly or not (the difference in VO2 max between children with regular
physical exercise and those without was insignificant, at 0.6 mL/Kg/min).
The concentration of PM10 in mainland Chinese cities are much higher than
our 'high pollution district', and outdoor physical exercise may be more
harmful than beneficial.
Tze Wai Wong
Reference: Yu ITS, Wong TW, Liu HJ. Impact of air pollution on
cardiopulmonary fitness of schoolchildren. Journal of Occupational and
Environmental Medicine 2004; 46:946-954.
Batty et al. conducted a follow-up study to know the effect of
passive smoking on subsequent mortality [1]. In men in their study, self-
reported passive smoking, not salivary cotinine, could predict mortality.
From their Tables 1 and 2, salivary cotinine level was categorized into
three groups, and self-reported passive smoking was categorized binary. I
have a query on the association between self-reported passive smokin...
Batty et al. conducted a follow-up study to know the effect of
passive smoking on subsequent mortality [1]. In men in their study, self-
reported passive smoking, not salivary cotinine, could predict mortality.
From their Tables 1 and 2, salivary cotinine level was categorized into
three groups, and self-reported passive smoking was categorized binary. I
have a query on the association between self-reported passive smoking and
salivary cotinine.
Martinez-Sanchez et al. reported that geometric mean (95% confidence
interval (CI)) of salivary cotinine in general population was 1.62 (1.41-
1.87) in 210 men and 1.34 (1.19-1.51) in 299 women, whose age ranged from
16 to 64 years. They also reported that the number of second-hand smoke
(SHS) at home was 133 among 509 non-smokers, and that at work was 153
among 378 non-smoking workers. The geometric mean (95% CI) of salivary
cotinine in subjects with SHS exposure at home was 1.57 (1.32-1.86) and
that without SHS exposure was 1.42 (1.27-1.58). In contrast, the geometric
mean (95% CI) of salivary cotinine in subjects with SHS exposure at work
was 1.44 (1.23-1.69) and that without SHS exposure was 1.42 (1.24-1.63).
From this study, information on self-reported passive smoking does not
reflect the level of salivary cotinine concentration.
Taking together, I recommend Batty et al. presenting data on the
association between self-reported passive smoking and salivary cotinine
concentration by presenting 2*3 cross tables, stratified by sex. By this
information, superiority of self-reported passive smoking on the
predictive ability against salivary cotinine concentration for mortality
would be partly clarified.
References
1 Batty GD, Gale CR, Jefferis B, et al. Passive smoking assessed by
salivary cotinine and self-report in relation to cause-specific mortality:
17-year follow-up of study participants in the UK Health and Lifestyle
Survey. J Epidemiol Community Health 2014;68:1200-3.
2 Martinez-Sanchez JM, Fu M, Perez-Rios M, et al. Comparing salivary
cotinine concentration in non-smokers from the general population and
hospitality workers in Spain. Eur J Public Health. 2009;19:662-4.
Dear Sir,
Thank you very much for your interest in our research. In our JECH paper,
we analyzed data on a broad range of pediatric diseases and found,
generally, children aged 10-14 years are more vulnerable to both hot and
cold effects, compared with children of other age groups. In the OEM
paper, we analyzed pediatric asthma data and found, specifically, children
aged 10-14 years are more sensitive to the adverse impact...
Dear Sir,
Thank you very much for your interest in our research. In our JECH paper,
we analyzed data on a broad range of pediatric diseases and found,
generally, children aged 10-14 years are more vulnerable to both hot and
cold effects, compared with children of other age groups. In the OEM
paper, we analyzed pediatric asthma data and found, specifically, children
aged 10-14 years are more sensitive to the adverse impact of cold on
asthma. We used consistent statistical approach in the two papers and have
controlled for relative humidity and air pollutants. Due to limited space,
we were not be able to present the difference between different age groups
in terms of vulnerability to temperature effects on every disease.
Best,
Zhiwei
Cohort 1: born 1970s and followed between ~15-35 years of age
Cohort 2: born 1950s and followed between ~35-55 years of age
Cohort 3: born 1930s and followed between ~55-75 years of age
This study design has serious limitations for the investigation of
cohort differences in BMI trajectories (and therefore also for the
investigation of...
Cohort 1: born 1970s and followed between ~15-35 years of age
Cohort 2: born 1950s and followed between ~35-55 years of age
Cohort 3: born 1930s and followed between ~55-75 years of age
This study design has serious limitations for the investigation of
cohort differences in BMI trajectories (and therefore also for the
investigation of cohort differences in the associations of socio-economic
position with BMI trajectory). Looking at Figure 1 - how, for example, can
you calculate cohort differences in BMI at age 45 years when only cohort 2
had data at that age? How, in fact, can you make any cohort comparison of
trajectories at ages when only one cohort had data? The mixed effects
model used in the present paper will provide estimates, but these are
based on trajectories fitted beyond the age range of the data (for at
least one cohort). The only ages were cohort differences could have
reliably been estimated are those were there is overlap (e.g., difference
in BMI at age 35 years between cohort 1 and cohort 2), and this could have
been done with cross-sectional analyses.
Failure to consider the limitations of the data has led to
potentially misleading interpretation of the results, such as "adiposity
increased most quickly with age in the youngest cohort". This is expected
given that cohort 1 was of an age where BMI is known to increase more
rapidly than later in life. The design of the study makes it impossible to
disentangle age and cohort effects (except at overlapping ages). For
example, is BMI at baseline greater in cohort 3 than in cohort 1 because
they are different cohorts (exposed to different environments) or because
they are different ages? The difference will, of course, be due to a
combination of age and cohort effects, plus any possible period effect.
Cross-cohort comparisons of trajectories are a powerful strategy, but
in nearly all instances they require the trajectories for each cohort to
span the same or similar age range.
These impressive data have been widely reported. Eating plenty of
fruit and vegetables seems to be a good idea, but I am concerned at how
the need to eat 7+ portions a day to obtain maximum benefit has been
reported. As far as I can tell, the estimates for "daily" consumption were
based on a single day. Few people eat exactly the same every day and
regression to the mean suggests that most of those who ate 7+ portions on...
These impressive data have been widely reported. Eating plenty of
fruit and vegetables seems to be a good idea, but I am concerned at how
the need to eat 7+ portions a day to obtain maximum benefit has been
reported. As far as I can tell, the estimates for "daily" consumption were
based on a single day. Few people eat exactly the same every day and
regression to the mean suggests that most of those who ate 7+ portions on
the day of the survey will not have sustained such high levels of
consumption over the 7.7 years of follow-up. While these people probably
continued to eat a lot of fruit and vegetables and gain health benefits
from doing so, isn't it misleading to extrapolate the quantity of fruit
and vegetable eaten in one day into a recommendation for regular daily
consumption?
Please see our supplementary material where you can see the questions
that were asked during the interview. You will note that survey
participants were explicitly advised not to include potatoes when
considering their answers.
Please see our supplementary material where you can see the questions
that were asked during the interview. You will note that survey
participants were explicitly advised not to include potatoes when
considering their answers.
Hintsa et al. examined the effect of effort, reward and job control on the exit from the labour market by a 6-year follow-up study in workers at the age of 61 years or younger [1]. The author adopted binary logistic regression analysis by adjusting several variables, and concluded that effort-reward imbalance (ERI), effort and job control were significant predictors for exit from the labour market. In contrast, reward wa...
Xu et al report an association between deranged liver enzymes e.g. alanine transaminase (ALT) in Chinese people residing in Southern China and the incidence of diabetes [1].
The authors examined multiple potential confounding factors which could influence liver enzymes e.g. alcohol consumption & adiposity. However, although the authors briefly mentioned chronic viral hepatitis infections i.e. chronic hepatitis B...
Oksuzyan et al. report an association between race/ethnicity and two subtypes of childhood leukemia: acute lymphoblastic leukaemia (ALL) and acute myeloid leukaemia (AML).1 Accordingly, the researchers suggest that there are genetic, cultural, and environmental factors involved in the etiology of childhood leukaemia [1].
Importantly, Oksuzyan et al. made a significant effort to examine and control for the poten...
Dear sir
The author commented on the paucity of research on the point where physical exercise in polluted air becomes more harmful than beneficial. I would like to share our research findings conducted more than 10 years ago in Hong Kong (Yu et al, 2004). We compared the physical fitness of school children who regularly performed physical exercise with those who did not. In a less polluted district (annual mean P...
Batty et al. conducted a follow-up study to know the effect of passive smoking on subsequent mortality [1]. In men in their study, self- reported passive smoking, not salivary cotinine, could predict mortality. From their Tables 1 and 2, salivary cotinine level was categorized into three groups, and self-reported passive smoking was categorized binary. I have a query on the association between self-reported passive smokin...
Dear Sir, Thank you very much for your interest in our research. In our JECH paper, we analyzed data on a broad range of pediatric diseases and found, generally, children aged 10-14 years are more vulnerable to both hot and cold effects, compared with children of other age groups. In the OEM paper, we analyzed pediatric asthma data and found, specifically, children aged 10-14 years are more sensitive to the adverse impact...
The three cohorts in this study were as follows:
Cohort 1: born 1970s and followed between ~15-35 years of age
Cohort 2: born 1950s and followed between ~35-55 years of age
Cohort 3: born 1930s and followed between ~55-75 years of age
This study design has serious limitations for the investigation of cohort differences in BMI trajectories (and therefore also for the investigation of...
These impressive data have been widely reported. Eating plenty of fruit and vegetables seems to be a good idea, but I am concerned at how the need to eat 7+ portions a day to obtain maximum benefit has been reported. As far as I can tell, the estimates for "daily" consumption were based on a single day. Few people eat exactly the same every day and regression to the mean suggests that most of those who ate 7+ portions on...
Dear Prof. Bland,
Please see our supplementary material where you can see the questions that were asked during the interview. You will note that survey participants were explicitly advised not to include potatoes when considering their answers.
http://jech.bmj.com/content/suppl/2014/03/04/jech-2013- 203500.DC1/jech-2013-203500supp1.pdf
Conflict of Interest:
N...
Pages