Preconception lifestyle advice for people with infertility

In the paper

“No studies reported on other behavioural changes.”

In the paper

“The other six studies focused solely on women ( Alibeigi 2020, Belan 2019, Hughes 2000, Mutsaerts 2016, Ng 2018, Rossi 2013).”

In the paper

“One study reported on behavioural changes for smoking but not as defined in the review methods.”

In the paper

“In one study, women were included with their male partner (i.e., as a couple) (Oostingh 2020); no studies focused on men only.”

In the paper

“No studies reported on live birth, ongoing pregnancy, adverse events, or miscarriage.”

In the paper

“We included seven parallel-design RCTs with a total of 2130 participants, including Hughes 2000, which was the only included study in the 2010 version of this systematic review (Anderson 2010).”

What the sources say

• Not checked yetcites Anderson 2010 source

In the paper

“No studies reported on adverse events or other behavioural changes.”

In the paper

“Three studies assessed preconception lifestyle advice on a combination of topics (Alibeigi 2020, Ng 2018, Oostingh 2020); the remaining four studies assessed preconception lifestyle advice on one topic: weight (Belan 2019, Mutsaerts 2016), alcohol intake (Rossi 2013), or smoking (Hughes 2000).”

In the paper

“One study reported on behavioural changes for alcohol consumption but not as defined in the review methods.”

In the paper

“No studies reported on other behavioural changes.”

In the paper

“For secondary outcome parameters on lifestyle behavioural changes, three studies reported on maintaining a healthy weight (Alibeigi 2020, Belan 2019, Mutsaerts 2016).”

In the paper

“Miscarriage: No studies reported on miscarriage.”

In the paper

“Four studies reported on following a healthy diet (Belan 2019, Mutsaerts 2016, Ng 2018, Oostingh 2020).”

In the paper

“We found seven randomized controlled trials (RCTs) comparing preconception lifestyle advice with routine care or attention control in a total of 2130 women and men with infertility.”

In the paper

“For the primary safety outcomes, two studies reported on adverse events including gestational diabetes and hypertension (Belan 2019, Mutsaerts 2016), and two studies reported on miscarriage.”

In the paper

“Considerable statistical heterogeneity (P = 0.0003 and I2 = 92%) and clinical heterogeneity between intervention and comparison were evident.”

In the paper

“Two studies reported on stopping smoking and other substance abuse.”

In the paper

“With respect to primary effectiveness outcomes, three studies reported on live birth in the study publication (Belan 2019, Mutsaerts 2016, Rossi 2013), and one study reported data on live birth on request (Oostingh 2020).”

In the paper

“All studies described their interventions in their publications; only Mutsaerts 2016 provided a full reproducible protocol on how the intervention was built.”

In the paper

“Any adverse event: No studies reported on adverse events.”

In the paper

“One study provided preconception lifestyle advice on smoking cessation (Hughes 2000).”

In the paper

“All studies defined preconception lifestyle advice as the intervention, whereas the control group received routine care or attention control.”

In the paper

“Live birth or ongoing pregnancy was reported in three studies (Belan 2019, Mutsaerts 2016, Rossi 2013), and one study provided data on live birth on request (Oostingh 2020).”

In the paper

“Seven studies met the inclusion criteria for this review, only three of which reported on live birth or ongoing pregnancy (Belan 2019, Mutsaerts 2016, Rossi 2013), and one provided data on live birth on request (Oostingh 2020).”

What the sources say

• Not checked yetcites Belan 2019
• Not checked yetcites Mutsaerts 2016 source
• Not checked yetcites Rossi 2013 source

In the paper

“One study provided preconception lifestyle advice on alcohol intake (Rossi 2013).”

In the paper

“The mean delta stage of change was 0.31 for the experimental group and 0.26 for the control group.”

What the sources say

• Not checked yetcites Hughes 2000 source

In the paper

“Given the limited number of identified studies, subgroup analyses on participant or intervention characteristics were not feasible, and inclusion of such a wide range of participants and interventions resulted in clinical heterogeneity.”

In the paper

“Although lifestyle behavioural changes after the intervention were reported in all studies, a variety of measures and definitions were used to report these changes.”

In the paper

“We identified one other systematic review on the effects of preconception lifestyle interventions on people with infertility (Lan 2017).”

What the sources say

• Not checked yetcites Lan 2017 source

In the paper

“In the first edition of this Cochrane systematic review, the need for further research into this subject was highlighted, as no available RCTs had assessed the effects of preconception advice on the chance of live birth or other fertility outcomes in people with diagnosed infertility (Anderson 2010).”

What the sources say

• Not checked yetcites Anderson 2010 source

In the paper

“Background and Methods updated to meet current Cochrane standards”

In the paper

“Decrease in number of weeks drinking above the safety daily limit in the past 6 months (mean ± SD) 5.6 ± 9 in experimental group (n = 16) 1.7 ± 3.5 in control group (n = 21) MD 3.90, 95% CI -0.76 to 8.56”

What the sources say

• Not checked yetcites Rossi 2013 source

In the paper

“Decrease in number of binges in the past 6 months (mean ± SD) 14.5 ± 44.4 in experimental group (n = 16) 1.2 ± 4.5 in control group (n = 21) MD 13.30, 95% CI -8.54 to 35.14”

What the sources say

• Not checked yetcites Rossi 2013 source

In the paper

“We identified 6 studies in addition to the first version of this review (Alibeigi 2020 Belan 2019 Mutsaerts 2016 Ng 2018 Oostingh 2020 Rossi 2013)”

In the paper

“New citation required and conclusions have changed”

In the paper

“Hughes 2000 Delta stage-of-change = difference in stage of motivation to change smoking behaviour (mean) Experimental group (mean): 0.31 (no SD reported) (n = 47) Control group (mean): 0.26 (no SD reported) (n = 47)”

What the sources say

• Not checked yetcites Hughes 2000 source

In the paper

“The total number of records found: 668”

In the paper

“Searches and results updated.”

In the paper

“18 February 2019 New search has been performed”

In the paper

“Rate of maintained cessation rose from 4% to 24% (P < 0.001)”

What the sources say

• Not checked yetcites Hughes 2000 source

In the paper

“Evidence suggests that adherence to a diet rich in vegetables, fruits, whole grains, fish, and poultry is related to better fertility and higher live birth rates among women and improved semen quality in men, and unhealthy diets may have the opposite effect (Chiu 2018, Gaskins 2018a, Giahi 2016, Grieger 2018, Salas-Huetos 2017, Vujkovic 2010).”

What the sources say

• Not checked yetcites Chiu 2018 source
• Not checked yetcites Gaskins 2018a
• Not checked yetcites Giahi 2016 source

In the paper

“These food items may contain salmonella, toxoplasmosis, campylobacter, or listeria, which may cause infection and consequently adverse pregnancy outcomes.”

What the sources say

• Not checked yetcites Government of South Australia 2015
• Not checked yetcites Ross 2006

In the paper

“Therefore, lifestyle modification is first-line treatment for women and men with obesity (Best 2017, Practice Committee 2015).”

What the sources say

• Not checked yetcites Best 2017 source
• Not checked yetcites Practice Committee 2015

In the paper

“Iron supplementation can prevent maternal anaemia and improve birth outcomes including birth weight (Haider 2017, World Health Organization 2012).”

What the sources say

• Not checked yetcites Haider 2017 source
• Not checked yetcites World Health Organization 2012

In the paper

“Three review authors (TB, ACV, MV) independently reviewed the included studies for risk of bias using the Cochrane 'Risk of bias' assessment tool”

In the paper

“women in the intervention group had a greater number of previous IVF attempts than women in the control group (P = 0.029)”

In the paper

“Median duration of time trying to conceive was longer in the intervention group than in the control group in Mutsaerts 2016 (P = 0.037);”

In the paper

“Three review authors (TB, ACV, MV) independently extracted data from eligible studies using a data extraction form”

In the paper

“and women in the intervention group were more highly educated than control group women (P = 0.004).”

In the paper

“This review does not provide clear guidance for clinical practice in this area.”

In the paper

“We analysed the data on an intention-to-treat basis as far as possible”

In the paper

“we aimed to minimise their potential impact by ensuring a comprehensive search for eligible studies and by being alert for duplication of data”

In the paper

“Evidence was insufficient to allow conclusions on the effects of preconception lifestyle advice on adverse events and miscarriage and on safety”

In the paper

“At baseline, participants in the intervention and control groups were comparable, with the exception of Alibeigi 2020: women in the intervention group had a greater number of previous IVF attempts than women in the control group (P = 0.029); and women in the intervention group were more highly educated than control group women (P = 0.004).”

In the paper

“However, current data suggest that non-prescription and recreational drug use during pregnancy is associated with increased risk of foetal death, low birth weight, and preterm birth (Gouin 2011, Gunn 2016Ladhani 2011, Metz 2017).”

What the sources say

• Not checked yetcites Gouin 2011 source
• Not checked yetcites Gunn 2016Ladhani 2011
• Not checked yetcites Metz 2017 source

In the paper

“The main limitations of the evidence were poor study methods in included studies (lack of blinding) and lack of (precision in) findings for live birth, safety outcomes, and reported behavioural changes.”

In the paper

“Three review authors (TB, ACV, MV) independently performed an initial screen of titles and abstracts retrieved by the search for potentially relevant studies.”

In the paper

“in Rossi 2013 the intervention group had on average 2.1 drinks/drinking per day and the control group 1.8.”

In the paper

“Analysis of the literature indicates that active and passive (second-hand) smoking of tobacco is associated with decreased fertility and reduced chance of a healthy, live birth both in the general population and among couples with infertility (Augood 1998, Hyland 2016, Practice Committee 2018, Radin 2014, Sharma 2016, Waylen 2008).”

What the sources say

• Not checked yetcites Augood 1998 source
• Not checked yetcites Hyland 2016
• Not checked yetcites Practice Committee 2018

In the paper

“Over past years, fertility clinics have tended to see an increasing number of patients with an unhealthy lifestyle, and they have acknowledged more and more the importance of providing preconception lifestyle advice (Gormack 2015, Homan 2018).”

What the sources say

• Not checked yetcites Gormack 2015 source
• Not checked yetcites Homan 2018 source

In the paper

“The importance of providing preconception lifestyle advice is also reflected in the guidelines on routine psychosocial care in infertility and medically assisted reproduction of the European Society of Human Reproduction and Embryology (ESHRE) (ESHRE guideline 2015).”

What the sources say

• Not checked yetcites ESHRE guideline 2015

In the paper

“The study details of Hughes 2000 could not be verified, as no study protocol was available and the study was not prospectively registered.”

In the paper

“We therefore rated these studies at unclear risk of attrition bias at this point.”

In the paper

“Protocols were available for six studies (Alibeigi 2020, Belan 2019, Mutsaerts 2016, Ng 2018, Oostingh 2020, Rossi 2013); four were prospectively registered (Alibeigi 2020, Belan 2019, Mutsaerts 2016, Oostingh 2020).”

In the paper

“Another study was rated at unclear risk of other bias, as it was not clear which lifestyle questionnaire was used, and if this questionnaire had been validated (Oostingh 2020).”

In the paper

“Finally, Rossi 2013 was not prospectively registered.Although these study authors did report on live birth, they did not report on quality of life (a pre-specified outcome), and baseline data on alcohol intake were missing (Rossi 2013).”

In the paper

“We considered studies to be at low risk of reporting bias if a protocol was prospectively registered and live birth or ongoing pregnancy was included as an outcome measure.”

In the paper

“We considered that not blinding outcome assessors was unlikely to influence the primary review outcomes (live birth, ongoing pregnancy, adverse events, and miscarriage) and certain secondary outcomes, including birth weight, gestational age, clinical pregnancy, and time to pregnancy.”

In the paper

“One study was rated at unclear risk of selection bias of sequence generation (Alibeigi 2020).”

In the paper

“In Alibeigi 2020, missing data and reasons for discontinuation were documented and were balanced between groups (five in the intervention group, six in the control group), and all randomized women were included in the ITT analysis.This study was rated at low risk of attrition bias.”

In the paper

“Alibeigi 2020 had a prospectively registered protocol but did not report on live birth or ongoing pregnancy.In addition, data on dietary behaviour were measured but were not reported.”

In the paper

“Oostingh 2020 had an unreported pre-specified outcome (Big 3 complications including smallfor-gestational-age, premature birth, and congenital malformations and reported an outcome (lifestyle risk score) not specified in the protocol.”

In the paper

“These two studies were rated at unclear risk of this bias (Alibeigi 2020, Rossi 2013).”

In the paper

“Sample size calculation: no, sub-analyses of larger trial”

In the paper

“outcome assessor is participant and is not blinded; reported behavioural changes are likely to be influenced.Eating healthy and moving is socially desirable”

In the paper

“Failure to report live birth in infertility trials is a major source of bias, as it should be the default primary outcome in such studies (Harbin Consensus Conference Workshop Group 2014).”

What the sources say

• Not checked yetcites Harbin Consensus Conference Workshop Group 2014 source

In the paper

“the power calculations were based on an assumption of an increase in the rate of vaginal birth of a healthy singleton at term from 45% in the control group to 60% in the intervention group, a 20% discontinuation rate during the lifestyle intervention, and a 5% loss to follow-up.We calculated that a sample of 285 women per group would provide the trial with a power of 80% at a two sided alpha level”

In the paper

“only abstract with limited data is available as of February 2021, so risk of attrition bias is unclear at this point ITT analysis is planned.400 women are recruited, 236 randomised (131 intervention, 133 control); no further information on missing data is available (yet)”

In the paper

“Baseline differences: no”

In the paper

“prospectively registered study protocol is available.Not all pre-specified outcomes are reported at this point.Only abstracts with limited data were available as of February 2021, so risk of reporting bias is high at this point”

In the paper

“Not clear which lifestyle questionnaire was used, and if this questionnaire was validated”

In the paper

“based on the outcome of live birth and the Fertility Fitness Programme, the inclusion of 58 women per group provides 80% power to identify a doubling in live-birth rates with this programme (25% to 50%, α=5%).Assuming a dropout rate of 10% the final estimate sample size will be 128 women”

In the paper

“randomization was performed online and was stratified according to trial centre and ovulatory status”

In the paper

“This review was conducted according to Cochrane methods, with a pre-defined protocol to minimize potential bias.”

In the paper

“Not applicable; no objective outcome measures in this study”

In the paper

“outcome assessor is participant and is not blinded; reported behavioural changes in smoking are likely to be influenced.Having quit smoking is a more desirable answer”

In the paper

“Risk of bias assessment for reporting bias was independently conducted by two review authors.”

In the paper

“only abstracts with interim analyses are available”

In the paper

“ITT analysis, but study authors did not impute on ITT those lost to follow-up; we assumed in our analysis that they did not have a clinical pregnancy.Missing data documented but not balanced between groups.30.2% in intervention group vs 18.2% in control group.Reasons for discontinuation not reported”

In the paper

“Not applicable; no objective outcome measures in this study”

In the paper

“for each stratum, a permuted block design is used and programmed before-hand”

In the paper

“participating women will be randomised in programme by the computer generation of a series of validation codes which is unique for each participant”

In the paper

“randomised using a computer-generated, blocked schedule”

In the paper

“prospectively registered study protocol is available.Not all pre-specified outcomes are reported at this point.Only abstracts with limited data are available as of February 2021, so risk of reporting bias is high at this point”

In the paper

“no, study author replied by email that there were no baseline differences be-tween groups”

In the paper

“Study appears free of other sources of bias”

In the paper

“We aimed to minimize reporting bias by conducting a comprehensive search for eligible studies and by being alert for duplication of data.”

In the paper

“Sample size calculation: yes, but not (yet) achieved: protocol says 440, abstract reported on 264”

In the paper

“Due to the small number of studies included in the review, we were not able to construct a funnel plot, and consequently, we were not able to examine the presence of publication bias.”

In the paper

“central randomisation centre”

In the paper

“Baseline differences: yes, there were significant differences between groups for the median duration of time attempting to conceive (P = 0.037)”

In the paper

“the random allocation sequence was generated by an investigator who was not involved in eliciting data or conducting the study”

In the paper

“the sample size was calculated using a significance level of 0.05 and a power of 80% to detect at least a difference of 25% between groups in 1 outcome of the study for the baseline success rate of 30%.”

In the paper

“computer generation of a series of validation codes which is unique for each participant.After completion of the baseline questionnaires, women will be randomised to the intervention or the control group.The randomisa tion process will be concealed”

In the paper

“randomization was stratified according to age groups.According to the order of entrance to the study, based on permuted blocks of size four, the patients in each of the 4 age groups of < 30, 30-34, 35-37, and 38-40 years were randomly assigned to the intervention and control groups”

In the paper

“Reporting bias was assessed through careful assessment of included studies for failure to report obvious outcomes or for reporting them in insufficient detail.”

In the paper

“Infertility is a couple's condition, and couples may have a correlated lifestyle, for example, in weight and diet (Best 2017).”

What the sources say

• Not checked yetcites Best 2017 source

In the paper

“although clinicians were blinded throughout, due to the nature of the study, it was not possible to blind women who were randomised”

In the paper

“no study protocol was available to further verify selective outcome reporting.Study authors were contacted for the previous version of this systematic review on reporting of live birth.They reported that this information was not collected.this information could have been collected at 12-month follow-up.Secondary outcome (rate of maintained cessation): no data on infertile and pregnant women were”

In the paper

“there were no statistical or clinical differences between the intervention and control groups for the mPP analysis”

In the paper

“blinding was not possible.However, we specified the type of infertility treatment before randomisation in order to minimize differences in treatment assignments; this led to similar distributions of infertility treatment in both groups”

In the paper

“only abstracts with limited data are available”

In the paper

“sample size estimation was based on what was considered to be a clinically significant difference between intervention and control in delta stage-of-change of 1.0.Standard assumptions of alpha (2-tailed) = 0.05 and beta = 0.20 were used.The sample required was 107 patients per study arm.Although this target was reached for combined infertile and pregnant participants (N = 214), it was not attained”

In the paper

“This review does not provide clear guidance for clinical practice on preconception lifestyle advice.”

In the paper

“not clear if outcome assessors were blinded.Clinical pregnancy not likely to be influenced, as this is an observer-reported outcome not involving judgement”

In the paper

“Baseline differences: yes, differences in in previous IVF attempts (P = 0.029) and differences in education level (P = 0.004) between control and intervention”

In the paper

“However, it highlights the lack of evidence on this topic, despite recommendations by international guidelines to provide counselling to all couples with infertility.”

In the paper

“Because these interventions were not initiated as preconception lifestyle advice alone, do not reflect daily clinical practice, and do not allow for assessment of lifestyle behavioural changes, we did not include such studies in our review.”

In the paper

“No changes were made as a result of the findings of included studies.”

In the paper

“outcome assessors were not blinded.BMI and clinical pregnancy were not likely to be influenced, as they are observer-reported outcomes not involving judgement”

In the paper

“only abstracts with limited data were available as of February 2021, so risk of attrition bias is unclear at this point”

In the paper

“prospectively registered study protocol was available.Most out-comes were analysed and reported in accordance with the protocol.Change in dietary behaviour and adherence to intervention were measured but not re-ported.Live birth or ongoing pregnancy was not included”

In the paper

“no baseline differences in vegetable intake, fruit intake, or folic acid intake; no further details on other baseline characteristics”

In the paper

“prospectively registered study protocol was available; outcomes were analysed and were reported in accordance with the protocol.”

In the paper

“the study flowchart for IVF is shown in Figure 1.Five women in the intervention group and 6 women in the control group were excluded due to lack of willingness and cooperation.Their data were used to assess the outcome of spontaneous pregnancy, but spontaneous pregnancies did not occur to them”

In the paper

“Sample size calculation: yes, but not achieved.Aimed for 3000 (general preconception cohort + sub-fertile cohort) in trial register.Aimed for 1000 general and 1000 subfertile in protocol”

In the paper

“however, all outcomes were measured based on clinical tests.They were all objective and quantitative not qualitative, so they could not be biased”

In the paper

“Other sources of bias Low risk Comment: study appears free of other sources of bias”

What the sources say

• Not checked yetcites Chang 2011 source

In the paper

“GRADE assessment was performed on the outcomes”

In the paper

“Incomplete outcome data High risk Comment: in Chang 2011 (table 2), 161 out of 511 participants randomised are reported to have infertility.However outcome data are available only for 37 participants.Study authors were contacted twice for clarification, but no reply was received as of February 2021”

What the sources say

• Not checked yetcites Chang 2011 source

In the paper

“Blinding of outcome as sessors Objective outcomes Low risk Comment: outcome assessors were not blinded.Live birth and miscarriage were not likely to be influenced, as they are observer-reported outcomes not involving judgement”

What the sources say

• Not checked yetcites Chang 2011 source

In the paper

“Blinding of participants and personnel All outcomes High risk Quote:”

What the sources say

• Not checked yetcites Chang 2011 source
• Not checked yetcites Saitz et al., 2007

In the paper

“Blinding of outcome as sessors Patient reported outcome measures High risk Comment: outcome assessor is participant and is not blinded; reported behavioural changes regarding drinking are likely to be influenced, as less drinking is known to be socially desirable”

What the sources say

• Not checked yetcites Chang 2011 source

In the paper

“Sequence generation Low risk Quote:”

What the sources say

• Not checked yetcites Chang 2011 source

In the paper

“TB, ACV, MV, CM, ED, and SLF declare that they do not have any conflicts of interest.”

In the paper

“the review authors agreed to use a random-effects model instead of a fixed-effect model because large variation between studies is expected (in populations and/or interventions)”

In the paper

“a narrative review summary format was chosen as the method for synthesis when it was not possible to conduct meta-analyses”