Effect of Postpartum Anaemia on Maternal Health-Related Quality of Life: A Systematic Review

21 Background: Postpartum anaemia remains a persistent and severe public health issue in many parts of 22 the world. Studies have reported mixed findings on the effects of anaemia during the postpartum period 23 on maternal health-related quality of life (HRQoL). We conducted this systematic review to summarise 24 available evidence to inform public health practitioners on whether 1) anaemia negatively impact 25 maternal health-related quality of life and 2) whether iron supplementation in anaemic women can 26 improve maternal HRQoL during the postpartum period. 27 Methods: This review’s protocol was registered online with PROSPERO (CRD42020206618). We 28 extensively searched Embase, PubMed, Cochrane and Scopus through the HINARI website to identify 29 studies that reported either association or effect of postpartum anaemia on fatigue, depression and 30 mother-child interaction. We restricted our search to studies of human females published in English from 31 databases inception until August 2020. We followed a guideline for reporting systematic reviews without 32 meta-analysis to synthesise data. 33 Results: Ten out of 15 studies where the direction of effect could be determined reported a significant 34 association between lower Hb levels and physical fatigue symptoms. Fourteen out of 19 studies also 35 reported a significant association between Hb levels and postpartum depression. There was evidence in 36 six of the seven and seven of the eight randomised controlled trials that iron replenishment significantly 37 decreased physical fatigue and postpartum depression respectively. Two of the four included studies 38 showed that anaemic mothers were less responsive and had negative feelings towards their children than 39 non-anaemic mother. 40 Conclusion: Evidence from this review suggests that postpartum anaemia negatively affects health41 related quality of life and that iron replenishment improved both symptoms of fatigue and depression. 42 Nevertheless, it remains unclear whether postpartum anaemia affects mother-child interaction. 43

Postpartum anaemia remains a persistent and severe public health issue in many parts of the world. [1] 69 The World Health Organisation defines postpartum anaemia as a haemoglobin concentration of <11g/dl 70

Data extraction and management 114
A standardised, pre-piloted data extraction form was used to extract data from included studies. For 115 studies with dichotomous outcomes, we extracted the number of events and the number of participants 116 in each group. We extracted the effect measure, which included both crude and adjusted ratios with their 117 respective 95% confidence intervals and p-values. We extracted means and standard deviation for 118 continuous outcome with normally distributed data while medians, range and p-value of the non-119 parametric test were extracted for continuous skewed data. We also extracted correlation coefficients for 120 correlation studies and median or mean change from baseline for longitudinal studies. 121

Risk of bias assessment 122
Two reviewers (EM and NP) independently assessed the risk of bias in the included studies, and 123 disagreements were resolved through discussion. For randomised trials, we used a revised Cochrane risk-124 of-bias tool for randomised trials (RoB-2). Thereafter the risk of bias in the individual study was judged as 125 either "low risk" or "moderate risk" or "high-risk bias". [12] The Newcastle-Ottawa Scale was used to assess 126 the risk of bias for cohort studies and case-control studies. [14] We considered studies rated with ≥7 stars 127 as good (moderate risk).
[15] We adopted and modified a tool for evaluating the risk of bias in non-128 randomised studies of interventions (ROBINS-I) for longitudinal observational studies. [16] We assessed 129 bias due to confounding, selection of participants, missing data, measurement of outcome, and selecting 130 the reported outcomes. We dropped two domains that assess bias due to intervention classification and 131 deviations from the intended interventions as these were deemed not applicable.

Data synthesis 133
We analysed the available data and described the results from the studies with missing data. Data has 134 been grouped and analysed separately depending on whether it has reported the effect of maternal 135 anaemia on either fatigue, depression or mother-child interaction. Due to heterogeneity in the included study, a meta-analysis was not possible. Therefore, we followed reporting guidelines for Synthesis without 137 meta-analysis (SWiM) in systematic reviews. [17] 138

139
The searches in Embase, PubMed, Cochrane Central Trial, and Scopus databases through the Hinari 140 website identified 7,547 citations, of which 82 (1.1%) full articles were extracted and assessed for their 141 eligibility. Of the 82 articles, 29 (35.4%) met the eligibility criteria. We further included one article 142 identified by searching the references of the included articles ( Figure 1). We categorised and reported our 143 findings based on the three domains of HRQoL, which includes fatigue (physical health), depression 144 (mental health) and mother-child interaction (social wellbeing). 145

Postpartum anaemia and fatigue 146
Fifteen studies reported either effect or association between postpartum anaemia and fatigue (Table 1). 147 Tam et al.[18] recruited 150 anaemic women (Hb≥8≤9.9g/dl) on day two postpartum and randomly 148 allocated them to either the iron supplementation or placebo groups. At six weeks postpartum, the iron 149 supplementation group reported improved general wellbeing compared to those in the placebo group (p-150 value <0.05). The risk of bias in this study was judged to be low.
[18] Similarly, , in all their 151 three randomised studies, found that rapid improvement of body iron levels reduced maternal fatigue. after adjusting for the mode of delivery. Miller et al. [30] also reported no significant association between 180 percentage change in Hb or ferritin levels with any SF-36 and MFI sub-scales. 181  Holms et al. [23,24] found that favourable changes of haematological parameters in women who received 197 intravenous iron were associated with favourable changes in EPDS scores at all study time points (p<0.05). 198

Postpartum anaemia and depression
These findings were not different from those of Paoletti et al.[35], who also reported a significant However, the authors did not indicate whether symptoms of depression were not present at the start of 207 the study. A case-controlled study by Alharbi et al.[38] recruited 352 women (117 cases and 235 controls) 208 in Saudi Arabia and found that low haemoglobin levels were a significant risk factor for maternal 209 depression (AOR: 1.70, 95%CI; 1.05-2.74, p=0.03) at 8-12 weeks postpartum. 210 Goshtasebi et al.[39] recruited 254 women at delivery in Iran and showed that Hb of <11g/dl at delivery 211 increased the risk for maternal depression (AOR; 4.64, 95% CI: 1.33-16.08) at 4-6 weeks postpartum. Sivan et al. [42] found no association between maternal iron stores (measured by Hb, MCV, ZPP, ferritin 226 and sTfR) and postpartum depression in a Chinese population at various study time points (r=0.10). 227 Additionally, Miller et al.[30] recruited 63 women who were admitted to labour and delivery room in the 228 USA and reported no significant association between anaemia (Hb<12g/dl) and depression scores. These 229 findings were not different from those reported by Paterson et al.[25] in Britain. They enrolled 1,010 230 postnatal women and found no relationship between EPDS scores and Hb levels at 4 or 6 weeks 231 postpartum. However, women with low Hb levels had high EPDS scores, but again the difference was not 232 statistically significant (p>0.05). 233

Postpartum anaemia and mother-child interaction 234
Mixed findings were reported on the effect of postpartum anaemia on mother-child interaction (Table 3). 235 , enrolled 95 women at Khayelitsha, South Africa and randomised them into either 236 IDA-placebo (n=30) or IDA-ferrous (n=34) and non-anaemic control mothers (n=31). Mother-child 237 interaction was assessed using a 20 minutes' video recording and independently transcribed by two 238 blinded research assistants. At nine months postpartum, the control and IDA-ferrous group scored 239 significantly better on maternal sensitivity, non-hostility, and structuring scales and child responsiveness 240 scale than did the IDA-placebo group (p-value = 0.007), whose iron stores remained low. Similarly, earlier 241 in the same setting, Perez et al. [44] enrolled 81 mothers (anaemic= 51 and non-anaemic= 30). The anaemic 242 mothers were randomised to either IDA-placebo (n=21) or IDA-ferrous (n=30). The mothers were matched 243 according to age, parity and educational background. The study reported that anaemic mothers in the 244 IDA-placebo group had negative statements towards their infants, less goal setting and responsiveness 245 than mothers in non-anaemic and IDA-ferrous groups (p <0.05) at nine months postpartum. 246 Unlike the above findings, Hamm et al. [28] reported that significant improvement in Hb levels in women 247 who received multiple units of RBCs (8.7g/dl versus 7.8g/dl) did not lead to significant improvement in 248 maternal attachment scores. Dearman et al.[45] in their pilot case-control study enrolled 115 women 249 (Hb<10.5g/dl=57 and non-anaemic=58) and reported no statistical difference in maternal perception of 250 mother-infant bonding between the anaemic and non-anaemic group. 251

Discussion of Results 252
Two important findings on the effect of postpartum anaemia on maternal fatigue were noted. Firstly, 253 evidence suggests that anaemia as indicated by low Hb or depleted iron stores, is associated with maternal 254 postpartum fatigue, which can be treated by iron replenishment. Our findings are similar to a previous 255 review that investigated the impact of intravenous iron treatment on HRQoL in patients with IDA.
[46] A 256 small sample size not powered to evaluate the effect of iron replenishment on fatigue might explain 257 inconsistent findings reported by Hamm et al.[28] Secondly, little attention has been paid to the same 258 condition in developing countries, and it remains unclear whether the above findings are applicable in the 259 African context. A meta-analysis by Badr et al.[47] aimed at identifying predicting factors for maternal 260 postpartum fatigue recommended further studies to determine whether race/geographical region can 261 predict postpartum fatigue. 262 We found mixed findings on the effect or association between anaemia or iron deficiency on postpartum 263 depression. Our review couldn't determine the source of heterogeneity in the study findings. However, 264 Azami et al. (2020) found that study quality, design, time of measuring both depression and anaemia did 265 not influence their meta-analysis findings reporting postpartum anaemia as a significant risk factor for 266 postpartum depression. [48] Similarly, our findings suggest that either postpartum anaemia or iron 267 deficiency is an independent risk factor of postpartum depression and iron replenishment in anaemic 268 women lowered symptoms of postpartum depression. We recommend future studies to investigate why 269 studies conducted in different setting produce varying results. Importantly, more studies should be done 270 in Africa, considering that the prevalence of anaemia is very high in this region. 271 The association between anaemia and mother-child interaction/bonding remains unclear. Globally, 272 researchers have paid little attention to this area. We only identified four studies that showed mixed 273 findings. Two well-conducted studies reported that maternal anaemia negatively impacted mother-child 274 interaction. Other studies that reported no association also had some shortfalls. For example, the sample size in Hamm et al.[28] was not powered to evaluate the effects of anaemia on secondary outcomes such 276 as mother-child attachment. Dearman et al.[45] also acknowledged that lack of association in their study 277 might be due to the small sample size lacking power. 278

Study Limitation 279
Our study has two major limitations. Firstly, we limited our search to human studies published in English 280 only because none of the authors is conversant with other languages. Secondly, these studies' results 281 were not pooled together due to heterogeneity in study designs and reported measure of effects. 282 However, we followed acceptable methods of data synthesis as recommended by Campbell et al.[17]. 283

Conclusion 284
To our knowledge, this is the first systematic review that has determined the effect of postpartum 285 anaemia on maternal health-related quality of life in a holistic approach by incorporating all domains of 286 HRQoL. While it is clear from our findings that postpartum anaemia negatively affects health-related 287 quality of life domains of physical and mental health and that iron replenishment tremendously improved 288 symptoms of fatigue and depression, it remains unclear on its impact on mother-child interaction. There 289 is a paucity of data from developing countries on the effect of postpartum anaemia on HRQoL. We, 290 therefore, call for well-designed studies in Africa to provide contextual evidence. Nonetheless, we agree 291 and call upon clinicians in developing countries to adhere to the World Health Organisation 292 recommendation of routine iron supplementation to women until six weeks postpartum to improve 293 maternal HRQoL during the postpartum period. This systematic review was not directly funded. However, author EM is a PhD student supported by the 300 Training and Research Unit of Excellence and Consortium for Advanced Research Training in Africa 301 (CARTA), which is funded by the British Council, Sida, The African Academy for Science, Ford Foundation 302 and Wellcome Trust. 303

Availability of data and materials 304
We used data from previously published articles that have been cited. 305

Author's contributions 306
Author EM envisioned the systematic review and drafted the manuscript. NP was involved in data search 307 and assessing articles for eligibility and the risk of bias. MNM and KP provided technical review and 308 resolved disagreements between EM and NP. All authors read and approved the final version of the 309 manuscript. 310

Ethics approval and consent to participants 311
Not applicable 312

Consent for publication 313
Not applicable 314