Residential Greenness & Birth Outcomes

Zhelnov P. A critical appraisal of ‘Hu CY, Yang XJ, Gui SY, Ding K, Huang K, Fang Y, Jiang ZX, Zhang XJ. Residential greenness and birth outcomes: A systematic review and meta-analysis of observational studies. Environ Res. 2020 Dec 8:110599. doi: 10.1016/j.envres.2020.110599. Epub ahead of print. PMID: 33307084’. Zheln. 2020 Dec 16;51(1):r26d11. URI: https://zheln.com/record/2020/12/11/26/.

Zheln Review Appraisal in 10 Steps:

1. ✅ Downloaded from the PubMed Systematic Subset Daily Updates
2. ✅ Meets Shojania & Bero 2001 True Positive Criteria for Systematic Reviews by Title
3. ✅ Full Text & Other Reports Collected by Zheln
4. ✅ Generates Pragmatic Evidence Directly Relevant to Evidence-Based Practice
5. ✅ Not Found Duplicate by Zheln
6. ❌ Failed Replication
7. 🔄 Has Critical Conduct Flaws?
8. 🔄 Liked or Disliked by Zheln?
9. 🔄 Practical Implications Summarized by Zheln
10. 🔄 Appraisal Published & Call for Crowdfunding

Executive Summary

A pregnant woman in a concrete jungle. Would it not be better for the baby if her mother had more trees, and grass, and other greenery around while carrying her? A recent review by Hu and coauthors looked at adverse pregnancy outcomes, such as the rate of preterm birth or low birth weight, and checked if the amount of greenness around the mother’s home could predict them.

Plants absorb visible light because they need it to photosynthesize oxygen. At the same time, plants reflect infrared light because it would heat and hurt them in the same way a hot radiator could. In contrast, inorganic land surfaces mostly reflect visible light but absorb infrared light. This is why we can easily differentiate between those by just taking pictures of Earth from satellites.

This is what most previous reviews on the topic looked at—this index of greenness, using sophisticated statistics to measure its association with birth outcomes. The review by Hu and coauthors continues on this path and takes into account all bleeding-edge research on the topic. However, I could not access all the information needed to check the review findings because of the publisher’s paywall.

Therefore, if you want me to go on with this appraisal and report on the effect of residential vegetation on pregnancy outcomes, please help me collect these data or crowdfund their collection at Zheln.com.

Generates Pragmatic Evidence Directly Relevant to Evidence-Based Practice?

Preterm birth rate, low birth weight rate, and infant mortality are usually listed as adverse birth outcomes. These are clearly patient-important too. The review features some of these, so it does in theory generate practice-important evidence.

Is Duplicate?

I ran a low-effort Google search: green urban preterm birth meta analysis OR metaanalysis. Some very similar reviews (and recent too) instantly showed (Akaraci 2020; Lee 2020). I successfully collected all the full texts. The authors do reference Akaraci 2020 at multiple places throughout the report despite it being a very recent study (April 2020); this is commendable. However, they do not reference Lee 2020, but this is an even more recent study (August 2020). Also, no registration info is available from Lee 2020, nor does it show up in the PROSPERO advanced search results. Three records were found there for birth AND (green*):TI:

• Green spaces and birth outcomes: A systematic review and meta-analysis of observational studies [CRD42020177721] – it’s the registration by Hu et al. themselves.
• Influence of Residential Greenness on Adverse Pregnancy Outcomes: A Systematic Review and Dose-response Meta-analysis [CRD42020162387] — I located and collected the report full text (Zhan 2020); this study is also notably referenced by Hu et al. despite it being very recent (May 2020).
• The health benefits of the great outdoors: a systematic review and meta-analysis of greenspace exposure and health outcomes [CRD42015025193] — I located and collected the report full text (Twohig-Bennett 2018); this study is not referenced by Hu et al. However, reading the Twohig-Bennett 2018 report showed that this review explicitly excluded birth outcomes, so was in fact irrelevant.

From the aforementioned Google search, I also identified closely related reviews by Porpora et al. (2019) and Bekkar et al. (2020) (with a correction; all full texts collected by Zheln). These are also not referenced by Hu et al. Notably, Bekkar 2020 is another ongoing appraisal on Zheln, so have a look at it too. Anyway, in contrast to other reviews mentioned previously, these two reviews are actually not about greenery but about environmental hazards, such as air pollution. Therefore, they are not directly relevant to the topic in question.

At last, Dzhambov 2014 was cited in some of the reports. It was probably the first systematic review on this topic. I have collected its full text too.

The general design layout of these relevant reviews is taking satellite measurements of environmental greenness (usually in the form of NDVI) around the mother’s home location, on the one hand, and either birth weight or some bad-pregnancy-outcome rate, on the other hand, and calculating associations between the two variables using meta-analysis. Also, various confounders are usually accounted for in the calculation model.

So what’s so new about the review by Hu et al.? It appears that the methods were somewhat different from before and some new studies were taken into account. The authors provided a clear explanation about that. Okay, so this is not duplicate then.

Passed or Failed Replication?

No supplementary data, that reportedly contain full search strategies, is available to me because of the paywall, so I can’t replicate the searches.

Clearly, this is not a problem of the study, but this is a paywall problem. Still, I can’t go on with the appraisal without these data. I’ll be able to go on once I get these supplementaries.

I have already started appraisal of this article and have not been able to replicate it. Replicability (reproducibility) is a must-have for a great systematic review, but sometimes it is compromised mostly by bad reporting. Therefore, assessing this review comprehensively will require me to make every effort to collect all available study records, including contacting the authors, and will take considerable time. I will complete the appraisal as soon as I’m able to. Please return sometime later or follow the progress of this page here or on GitHub as I do the appraisal. Also, you can look for similar systematic review appraisals using search or by tapping on AMA specialty tags at the bottom of this page or in the side menu.

If you really need me to appraise this review ASAP, please consider buying a $1 to $600 Zheln subscription on Patreon, GitHub Sponsors, or Open Collective. Afterwards, you’ll be able to immediately PM me with the review request. Otherwise, you might as well send me a donation via PayPal with this reference in the message, and I’ll happily appraise it for you.

By all means, you are always welcome reaching out to me for whatever reason by email or on Twitter, Facebook, Instagram, or Telegram. I’m happy to get your feedback!