Displaying 1-10 letters out of 460 published
Suggesting that sugars are 'hidden' in drinks marketed to children is misleading
Efforts to reduce sugar intake among children are required as intakes are much higher than recommended. Sweetened drinks are one of the main sources of dietary sugars, particularly for children and adolescents, and sugar intakes are around 2-3 times more than recommended by the Scientific Advisory Committee on Nutrition. That said, juices and smoothies make a nutritionally valuable contribution to children's diets as they count as one of their 5-a-day. The National Diet and Nutrition Survey revealed that only 10% of boys and 7% of girls aged 11 - 18 years met the 5-a-day recommendation. The Government's latest advice is that juices and smoothies should be limited to a combined total of one small (150ml) glass per day and to keep them to mealtimes to reduce the risk of tooth decay.
In determining the sugar content of the drinks surveyed, the authors themselves have used the clearly labelled amounts of sugars present in products. One of the suggestions in the paper to tackle unacceptably high sugar intakes is that the food industry stops adding unnecessary sugar to sweeten drinks. Pure fruit juices and fruit juice smoothies are made from whole fruit, crushed fruit and fruit purees, and are not permitted to contain added sweeteners or to dilute the naturally-occurring sugar content contributed by the fruit with water. Reducing the sugar content of these products is only possible by using lower sugar variants of fruits or vegetables. Fruit juice-based drinks, such as squashes and cordials, may include additional ingredients such as sugars, sweeteners and added water, which must be declared in the ingredients list. Reducing the sugar content of these products is possible, and some manufacturers have taken steps by replacing all or some of the added sugars with artificial sweeteners.
Scientific Advisory Committee on Nutrition. Carbohydrates and Health. The London: TSO, 2015.
National Diet and Nutrition Survey. Results from Years 1, 2, 3 and 4 (combined) of the Rolling Programme (2008/2009 - 2011/2012). A survey carried out on behalf of Public Health England and the Food Standards Agency. London: TSO, 2014.
Conflict of Interest:
Re :APOC3 testing in CHD
Thanks for the comments on our paper titled "Association between apolipoprotein C3 Sst I, T-455C, C-482T and C1100T polymorphisms and risk of coronary heart disease". In the letter, the author mentioned that we missed lots of studies on Sst I1-5. Among these studies, Baroni et al reported that S2 frequency was higher in CHD patients than controls (0.112 vs. 0.099, p>0.05) , however, odd ratio (OR) under dominant model could not be calculated; the study conducted by Masana et al was also excluded because of insufficient data ; Ding et al evaluated -455T > C polymorphism, reporting frequency of "C" allele was 26.9% in ACS group and 17.3% in control group (p<0.05); however, we found obvious mistakes in Table 2 of the paper, which reported numbers of TT, TC, CC were 46, 115, 68 in ACS group and 67, 137, 50 in controls (more C alleles than T) and OR was unavailable . We admit that the study by Paulweber et al was missed, which was published in 1988 and the full-text was not accessible in our first literature process and thus it was not carefully reviewed . Besides, the study by Tobin et al was also missed in previous pooled analysis of Sst I because they used C3175G as Sst I , while most of the studies used C3238G as Sst I polymorphism. We have compared the primers of the included studies and confirmed that both C3175G and C3238G were rs5128 polymorphism. Bhanushali et al reported a higher frequency of S2 in CHD cases (0.34) than in controls (0.27), with an OR of 2.34 (0.98-5.57); besides, they also reported C1100T polymorphism . We included this study in the previous pooled analysis of Sst I and C1100T polymorphisms but the ORs were not calculated under dominant model. The study by Wong et al assessed CHD risk in a prospective cohort, using a HR to estimate CHD risk according to kinds of polymorphisms, including Sst I, -482C>T and 1100C>T . However, HR under dominant model was available only in Sst I polymorphism. Additionally, we re-conduct the literature search process and several other studies on Sst I polymorphism was identified [8-10].
A total of 21 studies evaluated Sst I polymorphism. Among them, Baroni et al  and Bhanushali et al  reported a higher frequency of S2 allele in CHD cases, while the data were insufficient to calculate ORs under dominant model. Wong et al reported a HR of 0.72 (95% CI 0.45-1.15) for the association between Sst I polymorphism and CHD risk . Finally ORs under dominant model were available in 18 studies, and the pooled OR was 1.18 (95% CI 1.03-1.36).For T-455C polymorphism, Olivieri et al  and Martinelli et al  used data from the same project, while we failed to find the duplication because the authors and sample size were different. The study of Martinelli et al  was adopted in the current pooled analysis of T-455C for the larger sample size than Olivieri et al. The current pooled analysis represented an OR of 1.17 (95% CI 0.99-1.38). After excluding the study by Bhanushali et al , the pooled OR was 1.01 (95% CI 0.74-1.40) for C1100T polymorphism. No change has been made for the analysis of C-482T polymorphism.
Taken together, the conclusions reported in the previous manuscript were not significantly affected, and the re-pooled analysis also suggested a marginally significant association between APOC 3 Sst I and T-455C polymorphisms and CHD risk.
1. Baroni MG, Berni A, Romeo S, et al. Genetic study of common variants at the Apo E, Apo AI, Apo CIII, Apo B, lipoprotein lipase (LPL) and hepatic lipase (LIPC) genes and coronary artery disease (CAD): variation in LIPC gene associates with clinical outcomes in patients with established CAD. BMC Med Genet, 2003;4:8.
2. Masana L, Febrer G, Cavanna J, et al. Common genetic variants that relate to disorders of lipid transport in Spanish subjects with premature coronary artery disease. Clin Sci (Lond), 2001;100:183-90.
3. Ding Y, Zhu MA, Wang ZX, et al. Associations of polymorphisms in the apolipoprotein APOA1-C3-A5 gene cluster with acute coronary syndrome. J Biomed Biotechnol, 2012;2012:509420.
4. Paulweber B, Friedl W, Krempler F, et al. Genetic variation in the apolipoprotein AI-CIII-AIV gene cluster and coronary heart disease. Atherosclerosis, 1988;73:125-33.
5. Tobin MD, Braund PS, Burton PR, et al. Genotypes and haplotypes predisposing to myocardial infarction: a multilocus case-control study. Eur Heart J, 2004;25:459-67.
6. Bhanushali AA, Das BR. Influence of genetic variants in the apolipoprotein A5 and C3 gene on lipids, lipoproteins, and its association with coronary artery disease in Indians. J Community Genet, 2010;1:139-48.
7. Wong WM, Hawe E, Li LK, et al. Apolipoprotein AIV gene variant S347 is associated with increased risk of coronary heart disease and lower plasma apolipoprotein AIV levels. Circ Res, 2003;92:969-75.
8. Bai H, Saku K, Liu R, et al. Association between coronary heart disease and the apolipoprotein A-I/C-III/A-IV complex in a Japanese population. Hum Genet, 1995;95:102-4.
9. Ordovas JM, Civeira F, Genest J, Jr., et al. Restriction fragment length polymorphisms of the apolipoprotein A-I, C-III, A-IV gene locus. Relationships with lipids, apolipoproteins, and premature coronary artery disease. Atherosclerosis, 1991;87:75-86.
10. Aalto-Setala K, Kontula K, Sane T, et al. DNA polymorphisms of apolipoprotein A-I/C-III and insulin genes in familial hypertriglyceridemia and coronary heart disease. Atherosclerosis, 1987;66:145-52.
11. Olivieri O, Stranieri C, Bassi A, et al. ApoC-III gene polymorphisms and risk of coronary artery disease. J Lipid Res, 2002;43:1450-7.
12. Martinelli N, Trabetti E, Bassi A, et al. The -1131 T>C and S19W APOA5 gene polymorphisms are associated with high levels of triglycerides and apolipoprotein C-III, but not with coronary artery disease: an angiographic study. Atherosclerosis, 2007;191:409-17.
Conflict of Interest:
Marital status, self-rated health and dementia
Sundstrom et al. examined the association between marital status and dementia in adults by a 10-year cohort study . The authors handled 2,288,489 induviduals, and 31,572 dementia individuals were observed. Marital status was classified into married, single, divorced or widowed, and the incidence of dementia was monitored. Cox regression showed that significant increase in hazard ratios of single, divorced and widowed against married individuals were observed both in men and women, which was also recognized in young-old and middle-old individuals. The authors concluded that living alone, non-married, was a risk for dementia. I have some concerns in their study.
First, St John and Montgomery determined that self-rated health (SRH) in cognitively intact older adults could predict dementia, but SRH in older adults with cognitive impairment could not predict dementia by a prospective cohort study with multinomial regression models . This means that SRH monitoring is effective to prevent dementia in individuals without cognitive impairment, and SRH monitoring in combination with the screening for early cognitive impairment in individuals living alone is strongly recommended.
Relating to the first query, I previously reported that male, having disease, physical complaints, depressive state, lower levels of education and being divorced resulted a risk for poor self-rated health . Taken together, inter-relationship among marital status, self-rated heath and dementia should be considered simultaneously.
Finally, the authors concluded that social relationship should be considered as a risk factor for dementia. Kuiper et al. conducted a systematic review with a meta-analysis, and social relationship factors including a lack of social interaction were associated with incident dementia . Marital status and other information on social network should be evaluated simultaneously for the risk of dementia.
1. Sundstrom A, Westerlund O, Kotyrlo E. Marital status and risk of dementia: a nationwide population-based prospective study from Sweden. BMJ Open, 2016;6:e008565.
2. St John P, Montgomery P. Does self-rated health predict dementia? J Geriatr Psychiatry Neurol, 2013;26:41-50.
3. Kawada T, Suzuki S. Marital status and self-rated health in rural inhabitants in Japan: a cross-sectional study. J Divorce Remarriage, 2011;52:48-54.
4. Kuiper JS, Zuidersma M, Oude Voshaar RC, et al. Social relationships and risk of dementia: A systematic review and meta-analysis of longitudinal cohort studies. Ageing Res Rev, 2015;22:39-57.
Conflict of Interest:
Re:No evidence for kanuka honey being an effective treatment of rosacea
Messrs Schmidt and Schmidt, in their letter published in BMJ Open on 24 March 2016, assert that the study we reported about the effects of Kanuka honey in rosacea has a number of important flaws including problems with the experimental treatments, study design, and statistical analysis.
With respect to the study medications the aim of the study was not to estimate the effect of a preparation of Kanuka honey with another honey but to compare the properties of a honey treatment for rosacea against a control skin treatment. In this respect, the choice of treatment arms is clearly appropriate. Glycerine as an excipient, a pharmacologically inactive component, in the Kanuka honey preparation allows both ease of administration to the skin and also prolonged contact of the Kanuka honey with the skin. The use of excipients for this purpose in dermatological products is common. The control arm was Cetomacrogol cream. As outlined in the paper this is a non-ionic compound which itself comprises a number of excipients and which is also often used as a delivery agent for other dermatological agents. The study therefore compares, as much as possible, the potential active properties of Kanuka honey with a dermatologically neutral agent.
We agree that it is not possible to mask participants as to the active or control therapies in a study of this nature, due to the characteristic smell and taste of honey, and that this could potentially lead to bias in the estimates of treatment effects. We have not only highlighted this potential bias in the study report but have also been scrupulous in other aspects of the study design such as masked randomisation and masked assessment of the primary outcome variable of rosacea severity. This was achieved through the same masked investigator making the visual assessments of rosacea severity without any communication with the study participants.
With regard to the comment about statistical power, the paper presents a point estimate and confidence interval for the difference between the active and comparator arms of the trial and because the lower limit of the confidence interval is above the null, the issue of Type II error does not arise. Readers can judge for themselves the relevance of the lower or upper confidence limits of the difference. We did not specify or set out to carry out a Bayesian analysis of the experimental result in light of the anticipated rates of response. Even if we had done so, the point estimate of the relative response achieved, a relative risk of 2, was actually identical to that anticipated in the power calculation. We further note that the point estimate and confidence interval for the response rate in the active treatment group, 24/68, is 35.5% (24.1 to 47.8); and in the comparator group 12/69, it is 17.4% (9.3 to 28.4). Both of these are concordant with the response rates anticipated in the sample size calculation of 50% and 25% respectively.
Conflict of Interest:
The authors of this letter undertook the kanuka honey study, which is the subject of this correspondence. The study was funded by HoneyLab, who supplied the Kanuka honey product (Honevo).
Representation of drug adverse effects in product information
We strongly endorse the conclusion of Cornelius et al. regarding the necessity for improved communication of benefit-risk information to medication prescribers and users. Based on our extensive experience in evaluation and communication of medication-associated risks, we have previously proposed improvements in the processes for selecting and effectively communicating adverse drug reactions and interactions in HPI (1-3).
In our view, their article exemplifies the opacity underlying risk information in current regulatory Healthcare Professional Information (HPI) and points to the necessity for its readers to be familiar with past and present regional standards and approaches for its creation, especially when comparing HPI across regions. We would like to draw attention to potential additional reasons for the observed adverse reaction profile differences, and to the need their analysis illustrates for an understanding of the specialized rules, terminologies, and other aspects of HPI creation to avoid misinterpreting its content.
The decision whether to list an undesirable event as an ADR in HPI ("ADR for the purposes of HPI") involves judgement about both the strength of evidence for a causal association with the drug and the significance of that potential ADR for the prescribing decision. In general, the causal evidence threshold is not fixed, but rather adjusted according to the clinical importance of the event, being lower for those with greater potential for severe harm to the patient. However, we do not believe this to be a major cause of the substantial differences in ADR enumeration between US and EU HPI in the subset described. Rather, given the age range of the products studied, it appears that these primarily reflect differences between the EU approach to HPI of including only ADRs (i.e. those with evidence supporting a causal association) vs. the then- prevalent US approach of including both ADRs and adverse events with only a temporal relationship to drug, which resulted in the inclusion in US HPI of many reported adverse events, especially from clinical trials, without a reasonable causal association with drug. In 2006, FDA ended this decades-long practice and once again began to require that ADRs meet the "reasonable causal association" evidentiary standard to be included in HPI.
Older US HPIs in particular may also have used the term "adverse event" to encompass both actual "adverse reactions" and "adverse events without reasonable causal association". Furthermore, the commonly seen explanation for displaying adverse reactions/events as being those occurring in e.g. ? 5% of patients and more frequently than with placebo does not necessarily represent a criterion for categorising an event as an ADR, but may simply reflect the criterion for including it in that table or list. More specific criteria, such as "reported in ? 5% and at least twice as frequently than with placebo" represent surrogate causality criteria.
The degree of confidence regarding an ADR's causal association is only exceptionally described in HPI, and it is often unclear whether or to what extent listings may contain items other than "ADRs for the purposes of HPI". It should also be noted that HPI readers unfamiliar with pharmaceutical regulatory definitions may not appreciate that terms such as adverse reaction, suspected adverse reaction, etc. may have markedly different meanings in different contexts, e.g. for obligatory reporting of adverse events to regulatory authorities vs. for informing the medical community or the public. For example, an event from a clinical trial that is deemed a "suspected adverse reaction" (SAR) for the purposes of mandatory regulatory reporting may not meet criteria for inclusion in HPI.
Given the foregoing and other limitations of the safety information presented in the HPI, it is essential to consult sources other than HPI for additional information and interpretation regarding actual and potential risks. For example, several regulatory authorities have begun publishing detailed "assessment reports" on the internet. In those jurisdictions, more recently created regulatory product information thus consists of 3 layers: patient-directed information; "traditional" HPI (such as the EU SmPC and US PI); and the more detailed information in public assessment reports.
Finally, the following caveats must always be borne in mind when assessing the benefits and risks of medicines. For a new medicine, it is essential to understand that the safety profile in the HPI is purely provisional, as only the most common adverse reactions have any reasonable likelihood of detection in clinical trials. The totality of risks, especially uncommon ones, can only be ascertained after exposure of much larger patient populations with clinical characteristics that often differ from those in clinical trials, creating an additional challenge for assessing causality with the typically limited case data available. In both clinical trial and licensed clinical use, there is almost always a significant "background noise" of undesirable occurrences that occur in the target patient population and are unrelated to the drug. These events typically account for the bulk of reported safety data, creating a "needle -in-the-haystack" situation that may obscure the smaller subset of attributable harms (ADRs). Various informatics methods have been developed to address this issue, but clinical judgement remains the final determinant of causality and hence HPI inclusion.
Given the importance of presenting clear, accurate, and complete safety information to all medication users, it is disappointing that the international harmonisation of many aspects of drug development and licensing achieved during the past 25+ years remains elusive for this critical regulatory function.
Drazen J et al. Institute of Medicine (IOM) 2007. Adverse Drug Event Reporting: The Roles of Consumers and Health-Care Professionals, Workshop Summary. Washington, DC: The National Academies Press.
Fontaine AL. Current Requirements and Emerging Trends for Labelling as a Tool for Communicating Pharmacovigilance Findings. Drug Safety, 2004; 27: 579-589
Kahn SN. Signal Detection and Evaluation in Pharmacovigilance: Analytical and Regulatory Aspects. Regulatory Affairs Focus, 2006; 11(4); 14-19.
Conflict of Interest:
Critical appraisal of the article
I wish to provide some perspective from epidemiology with regard to a recent study that appeared in your journal. First, this study is the first population-based study to look at the relationship between MS and retinopathy in a North Chinese population, but the study did not mention why it uses Shenyang people as the study population. Second, the author provided numerous studies on MS, DM. But the stereo fundus photography was not mentioned in the introduction, which is used to detect retinopathy. Third, logistic regression analyses for retinopathy were only adjusted for age, gender, drinking, smoking and health conditions (BMI, HbA1c, duration of diabetes, SBP, DBP), but there are more important factors that influence people's retinopathy, such as demo-socioeconomic status, occupations and other health conditions. Such factors also should be taken into account. Fourth, the tables and figures were clearly presented - however, the P -value did not show in the Table 2. It is better to add the P-value so as to show how significant it is. Fifth, diabetes status was defined according to self-reported physician diagnosis or the use of antidiabetic agents. Nevertheless, the type of diabetes was not investigated.
In summary, there are some limitations to this study. I hope this letter can point them out, which could help future similar studies.
Conflict of Interest:
The need to further examine the role of religion and its association to STD incident
Is there an association between church membership and lower sexually transmitted disease (STD) incident among selected Danish religious cohorts? A large body of literature has shown a positive link between religion and health, particularly it's role in reducing lifetime risk for cardiovascular disease and selected cancers (1, 2). However, little research has looked at the effects of religion on sexual behavior and incidence of STDs.
The authors in this study sought to investigate this association among Danish Seventh Day Adventists (SDAs) and Baptists. They hypothesized that religious societies in the selected Danish cohorts will have a lower incidence of STDs. Choosing a retrospective cohort design, the authors followed two selected cohorts from 1977 to 2009 by linking national registers of inpatient and outpatient care. They then compared STDs incidents for syphilis, gonorrhea and chlamydia among the selected cohorts and the general Danish population. The results showed lower incidence of STDs were observed for both cohorts.
The study selected a large Danish religious cohort of 3,119 SDAs and 1,856 Baptists, setting it apart from other similar studies with small sample sizes (3, 4, 5, 6). Furthermore, the study took necessary steps to ensure that identification of the selected cohort was done correctly and that the study represented the Danish population. However, the study has a few limitations that need to be addressed. Most of the study data was derived from the national patient register and unfortunately these records were not designed for meeting the specific objectives of the study; therefore, it is very possible that some of the variables needed for the study were not made available in the records. Furthermore, the patient recorded data may have been completed by different health care professions and therefore prone to consistency and/or inaccuracy. Additionally, the article mentions that only 26 out of 46 Baptist communities took part in the study making these individual prone to selection bias.
It is also not clear why the authors chose to observe the STDs syphilis, gonorrhea and chlamydia. Were these the most frequently reported STDs during this time or were they selected based on available data from the national patient registers? What about other STDs such as HPV (discussed in article conclusion but not part of study analysis), genital herpes and HIV? The authors also chose to adjust for age and gender which are important confounders but many other external factors (race, individual sexual beliefs and attitudes, friends, family and peer influences, social norms and culture, etc) were not taken into consideration. Lastly, church membership and/or affiliation tell us very little about how religious a person is and may therefore provide very little information regarding the link between religion and STD incident. There is a need to further examine the link between religion and STDs incidents by possibly conducting a prospective cohort study that looks at the association between religious practice (church attendance, participation/activity), attitudes, and beliefs, and how these relate to STDs incident.
1. Wallace JM, Forman TA. Religion's role in promoting health and reducing risk among American youth. Health Educ Behav, 1998;25:721-41.
2. Krup et al. Association between sexually transmitted disease and church membership. A retrospective cohort study of two Danish religious minorities. BMJ Open, 2016;6:10128
3. Landor A, Simons LG, Simons RL, Brody GH, Gibbons FX. The Role of Religiosity in the Relationship Between Parents, Peers, and Adolescent Risky Sexual Behavior. Journal of Youth and Adolescence, 2011;40(3), 296-309.
4. Rostosky SS, Wilcox BL, Wright MLC, et al. The impact of religiosity on adolescent sexual behavior: a review of the evidence. J Adolesc Res, 2004;19:677-97.
5. McCree DH, Wingood GM, DiClemente R, et al. Religiosity and risky sexual behavior in African-American adolescent females. J Adolesc Health, 2003;33:2-8.
6. Lefkowitz ES, Gillen MM, Shearer CL, et al. Religiosity, sexual behaviors, and sexual attitudes during emerging adulthood. J Sex Res, 2004;41:150-9.
Conflict of Interest:
Early Goal Directed Therapy (EGDT) for sepsis is dead and buried
Three large, international, multicenter randomized trials have convincingly demonstrated that Early Goal Directed Therapy (EGDT) does not improve the outcome of patients with severe sepsis and septic shock. A meta-analysis by the ARISE, ProCESS and ProMISe investigators concluded that "EGDT is not superior to usual care for ED patients with septic shock but is associated with increased utilisation of ICU resources". The meta-analysis by Yu and colleagues appears to cast doubt regarding these conclusions. It is evident from their meta-analysis that the studies by Rivers et al  and Yan et al  are "profound" statistical outliers which raises serious concerns regarding the validity of the findings of these studies. [5-7] Furthermore, the results of these studies are on face value biologically implausible. The basic premise of EGDT is to optimize tissue oxygen delivery to reverse tissue hypoxia with the use of continuous monitoring to prespecified physiological targets. This basic tenant of EGDT is without scientific foundation. [8,9] Furthermore, none of the component interventions of EGDT have been individually demonstrated to improve patient outcome.[8,9] It would therefore appear absurd to propose that the combination of a number of non-beneficial therapies should prove to be beneficial. It is abundantly clear that the EGDT protocol for sepsis does not improve patient outcomes and that this therapeutic strategy must be abandoned. It is misguided for Yu and colleagues to propose "that further well-designed studies should eliminate all potential source of bias to determine if EGDT has a mortality benefit". EGDT for the management of sepsis is dead and buried; there is no place for an exhumation and further meta-analyses on this topic should now cease.
1. Angus DC, Barnato AE, Bell D et al. A systematic review and meta- analysis of early goal-directed therapy for septic shock: the ARISE, ProCESS and ProMISe investigators. Intensive Care Med, 2015; 41:1549-60.
2. Yu H, Chi D, Wang S et al. Effect of early goal-directed therapy on mortality in patients with severe sepsis or septic shock: a meta- analysis of randomised controlled trials. BMJ Open, 2016; 6:e008330.
3. Rivers E, Nguyen B, Havstad S et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med, 2001; 345:1368-77.
4. Early goal-directed therapy collaborative group of Zhejiang Province: The effect of early goal-directed therapy on treatment of critical patients with severe sepsis/septic shock: A multicener, prospective, randomized controlled study [in Chinese]. Zhongguo Wei Zhong Bing Ji Jiu Yi Xue, 2010; 6:331-34.
5. Ionnidis JP. Research accomplishments that are too good to be true. Intensive Care Med, 2014; 40:99-101.
6. Ioannidis JP. Why most published research findings are false. PLOS Med, 2005; 2:e124.
7. Burton TM. New therapy for sepsis infection raises hope but many questions (lead article). The Wall Street Journal, 2008; 14 Aug.
8. Marik PE. Surviving sepsis: going beyond the guidelines. Ann Intensive Care, 2011; 1:17.
9. Marik PE. The demise of early goal-directed therapy for severe sepsis and septic shock. Acta Anaesthesiol Scand, 2015; 59:561-67.
Conflict of Interest:
Sugar in children's fruit drinks
Growing public awareness of the detrimental effect sugar sweetened drinks has on children's teeth and weight has prompted many parents to choose seemingly healthier fruit juice and smoothie alternatives. However, the recent study by Boulton et al. has found that fruit based drinks, such as fruit juices and smoothies, specially marketed to UK children contain "unacceptably high" levels of sugar. 
New UK guidelines extend the advice that only a 150 ml serving of fruit juice counts as one of the recommended five portions of fruit and vegetables a day and this applies to smoothies - which previously contributed to two 5 A DAY portions. This is in acknowledgement of the high sugar content of smoothies.  The study by Boulton et al. found that only six products were actually sold in 150 ml bottles/cartons, which is of no help to parents trying to make a healthy choice for their children. 
For confused parents looking to do the best for their children, the message seems to be "don't assume fruit juices are low sugar and always check the label to find out".
Whole fruit is the healthier alternative because it gives the children an intake of fibre from the fruit, which is often missing in fruit juices.
For parents looking to avoid sugary drinks, it seems that fruit juices and smoothies cannot be considered safe options. The zero risk approach is water.
1. Boulton J, Hashem KM, Jenner KH, Lloyd-Williams F, Bromley H, Capewell S. How much sugar is hidden in drinks marketed to children? A survey of fruit juices, juice drinks and smoothies. BMJ Open, 2016;6:e010330.
2. Public Health England. The Eatwell Guide. March 2016. https://www.gov.uk/government/publications/the-eatwell-guide
Conflict of Interest:
Response to Gattrell et al., Professional Medical Writing Support and the Quality of Randomised Controlled Trial Reporting: A Cross-Sectional Study
I read with interest the study by Gattrell et al., Professional Medical Writing Support and the Quality of Randomised Controlled Trial Reporting: A Cross-Sectional Study. As a medical-technical writer/editor for more than 40 years, I was pleased to see such serious research into the value of medical writing.
The authors conclude that "the enhanced reporting seen in industry-funded trials may be attributable to professional medical writing support." However, "medical writing support" may not indicate the contribution of medical writers as much as it reflects the advantages of an in-house publication process organized as well to meet regulatory requirements and that coincidentally includes medical writers. Given its cost, medical writing is most often provided by in-house medical writers who prepare and edit draft manuscripts for in-house researchers and sometimes for external researchers. Medical communications companies also offer medical writing, but almost always to corporate clients, not to individual authors. More than 98% of supported articles were industry-sponsored; thus, with the exception noted below, many, if not most, of these articles may have been prepared in-house. The authors assumed (I think incorrectly) that the responsibility of including the CONSORT information rests primarily with medical writers. However, anyone in an internal review or the peer review process can insist on following CONSORT. Company policy may also require adherence to the guidelines. If so, the link between medical writing support and adherence to CONSORT may reflect organizational characteristics, not the initiative of medical writers.
Of the 12 guidelines studied, the median percentages of articles that did and did not have acknowledged medical writing support were within 10% of each other for 7, although adherence was always higher in supported articles. The below analysis of the remaining 5 guidelines (6a, 7a, 8b, 14a, and 23) again suggests that the difference between groups can also be explained by the organizational context in which the articles were prepared.
Because industry-sponsored research has regulatory implications, key information will likely be collected and reported in manuscripts. For example, the primary outcome variable (Guideline 6a) will probably be specified in an article, as indeed it was in 89% of supported articles (vs. 50% for unsupported articles). In contrast, in my experience, many individual authors, who do not have to meet regulatory requirements, often do not explicitly state the primary outcome because they assume the endpoint is obvious, even if it isn't. For the same reason, the details of trial registration (Guideline 23) are perhaps more likely to be reported in articles prepared in-house (81% vs. 49%), as are the dates of recruitment and followup (Guideline 14a; 38% vs. 19%).
Supported articles more often included a sample-size calculation (Guideline 7a). Median sample size was 43 (vs 159) for unsupported articles. With samples that small, a sample size calculation may not have been done and hence could not have been reported. Likewise, the type of random assignment (ie., whether stratification and blocking were used; Guideline 8b) was reported in only 13% of non-supported articles: small samples don't lend themselves to stratification. What appear to be omissions in reporting in either group may actually be omissions that no medical writer can fix. One could argue that the guidelines should be addressed in both the affirmative and the negative, but I have never read a report of an RCT that explained why a sample size calculation or the dates of recruitment were not provided. In the comparison limited to industry-supported articles, again it is unclear whether the authors did (or could) distinguish between articles produced in-house and those produced out-of-house. Partially industry-funded articles were defined as those written by external researchers who received a supply of drugs to be tested. Of these 25 articles, 23 reported no medical writing support, which is consistent with an absence of in-house medical writing support. If medical writers are responsible for meeting the guidelines, the obvious question is why supported articles were deficient in any of them. Adherence was less than 50% for 9 of the 12 guidelines. For example, a flow chart is an obvious and unambiguous requirement, but the percentages of both groups including a chart were only 40% and 32% for supported and non-supported articles, respectively, although this information may be in the text of articles with smaller samples. The authors state that "Declared professional medical writing support was associated with improved completeness of reporting in...reports of RCTs..." It may be more accurate to say that medical writing support was associated with articles that were not as poorly reported as those without declared support. If medical writers get credit for improving reports of RCTs, it follows that they should be held responsible for any shortcomings in such reports. In fact, I don't believe the situation is that simple. I have added the required CONSORT information only to have authors remove it so they could expand the discussion section, for example. The English was "acceptable" in 80% of supported articles and in 45% of unsupported articles (the percentage of corresponding authors from countries where English is the first language was the same in both groups). The relative contributions of authors, writers, and reviewers is unknown, and the number of people reviewing the manuscript is likely to be greater in manuscripts developed in-house. Curiously, Biomed Central endorses CONSORT, yet their journals obviously don't insist that the guidelines be followed. Remember that CONSORT includes a checklist for authors to indicate the manuscript page on which each guideline is reported. The checklist is designed specifically to aid reviewers in assuring that each guideline was met. Why were these articles accepted for publication if reporting was so incomplete? In conclusion, although I am firmly convinced that skilled medical writing and editing do substantially improve the quality of scientific articles - if for no other reason than the fact that so many individuals and companies are willing to pay so well for it - I'm not persuaded that this article provides evidence that medical writers are uniquely and directly associated with the quality of methodological reporting.
1. Gattrell WT, Hopewell S, Young K, et al. Professional medical writing support and the quality of randomised controlled trial reporting: a cross-sectional study. BMJ Open 2016;6:e010329. doi:10.1136/bmjopen-2015 -010329.
Conflict of Interest: