Lung function impairment in pediatric patients with sickle cell anemia from Nigeria is associated with low steady state hemoglobin

Severe anemia at steady state is a risk factor for pulmonary hypertension in patients with sickle cell anemia (SCA).1 We hypothesized that lower steady state blood hemoglobin (Hb) levels could be also associated with more severe lung function impairment in SCA pediatric patients from sub-Saharan Africa. We evaluated the relationship between different predictors, including low steady state Hb level, and spirometry results in children and adolescents with SCA from Nigeria.

In a cross-sectional study, patients with Hb phenotype SS aged 6-18 years, followed at the Barau Dikko Teaching Hospital, Kaduna, Nigeria, underwent spirometry and anthropometry on the occasion of a follow-up appointment. The most recent steady state full blood cell count was recorded. Steady state was defined according to Ballas.2 A control group for spirometry outcomes of age-matched non-SCA individuals was enrolled in some schools in Kaduna state. The study was approved by the local Ethics Committee (HREC 16-0017) and informed consent was obtained from parents. Information on clinical history was collected from medical records and by interviewing patients' caregivers. Exclusion criteria were: no blood tests performed at steady state in the last 6 months, current hydroxyurea treatment, respiratory symptoms or feeling unwell on the test day, SCA-related acute events (eg, pain crises) in the last month. An Easy-on-PC spirometer (ndd, Zurich, Switzerland) was used. The principal investigator performed all tests and spirometry data were included if they satisfied the American Thoracic Society quality standards adapted for children.3 Spirometry z-scores were derived according to the Global Lung Initiative 2012 reference values.4 The lower limit of normal (LLN) for spirometry was established at −1.64 z-scores (5th percentile). Spirometry patterns were classified as normal (FEV₁ and FVC ≥ LLN), obstructive (FEV₁/FVC < LLN), restrictive (FVC < LLN) or mixed (FEV₁/FVC and FVC < LLN). Group differences were tested using unpaired t test, χ² or Fisher's exact test as appropriate. A P-value <0.05 was considered as statistically significant. A multiple logistic regression model for “restrictive spirometry” (vs normal) was built in two steps: first, in addition to “Hb level in the lowest quartile”, the following potential confounders were included in a fully adjusted model: sex, age, white blood cell count in the highest quartile, body mass index z-score (zBMI) < −2, history of acute chest syndrome, history of asthma (parent report of physician-diagnosed asthma ever), at least 3 pain crises in the last year. There were no significant interactions between the predictors included in the model. Covariates with P < 0.20 were then retained in the final model (Table 1).

A total of 186 patients with SCA and 434 controls were initially enrolled. Only one SCA child was on hydroxyurea. After exclusions, data from 126 patients (mean ± SD age 11.5 ± 3.1 year, 53% boys) and 364 controls (mean ± SD age 10.4 ± 2.4 years, 52% boys) were analyzed. Median (1st-3rd quartile) Hb value was 7.8 g/dL (7.3-8.4). A restrictive spirometry was detected in 29.3% of the patients (7.9% in controls), an obstructive spirometry in 4.3% (3.6% in controls) and a mixed pattern in 2.9% (0.5% in controls). Frequency of a restrictive spirometry pattern was higher in patients with Hb level in the lowest quartile (53.6%) than in the other patients (25.0%) (P = 0.004), as well as frequency of a FEV₁ z-score < 5° percentile (46.4% vs 23.9%, P = 0.02). In SCA patients, the lowest quartile of Hb level was associated with a 3.9-fold increased risk of restrictive spirometry pattern (95% confidence intervals 1.4 to 10.6; Table 1) compared to the highest Hb quartile. Frequency of Hb level in the lowest quartile was similar between SCA patients with wasting (zBMI < −2), who represented 30.8% of the study group, and those normotrophic (P = 0.2).

We found that a low steady state Hb level was associated with increased risk of restrictive spirometry in pediatric African patients with SCA. A pathological FEV₁ was also common in patients with low Hb level, and it was most frequently due to the presence of a restrictive pattern.

In adults with SCA from the United States, an association between a restrictive lung disease and a more severe HbSS phenotype, including lower Hb levels and hematocrit, was reported.5 Lunt et al, instead, showed that among children and young adult SCA patients from the United Kingdom, those with the most severe anemia had increased pulmonary capillary blood volume and respiratory system resistance, associated with a mixed obstructive-restrictive physiology.6 The author hypothesized that pulmonary vascular engorgement in these patients could cause compression of extrinsic small airway, increasing airway resistance.6 A similar mechanism could be present also in our patients and the predominance of a restrictive over a mixed obstructive-restrictive spirometry pattern could depend on a more advanced stage of chronic lung injury in Nigerian pediatric SCA patients compared to their counterparts in the UK.7

In a large study on SCA patients from West Africa, the lowest quartile of Hb level was associated with elevated tricuspid regurgitant jet velocity (suggesting pulmonary hypertension) and microalbuminuria.8 We report a further association of the lowest quartile of Hb level with restrictive spirometry in pediatric patients with SCA from Nigeria. A lower steady state Hb level could be related to a higher rate of chronic hemolysis. In the pulmonary microcirculation hemolysis triggers a cascade of events, including endothelial dysfunction, oxidative stress, inflammation, vaso-occlusion with ischemia and reperfusion injury, which may result in cumulative lung injury and restrictive lung function abnormalities. However, a low steady state Hb in African patients with SCA could also depend on other causes of anemia, such as malnutrition-related iron and micronutrients deficiency or chronic parasitosis, including chronic malaria. In this study severity of anemia was not related to the presence of undernutrition (zBMI < −2) in SCA patients.

This study has several limitations: lung function assessment was based only on spirometry; hemolysis indices and other causes of anemia were not investigated; lung function and Hb level were based on a single record instead of a more accurate longitudinal evaluation. A strength is that this is the first study investigating the relationship between Hb level and lung function in patients with SCA from sub-Saharan Africa. Furthermore, we took into account several potential predictors of impaired lung function beyond a low Hb, in order to limit biases.

In conclusion, this study shows that a steady state low Hb level in African pediatric patients with SCA is associated with more severe lung function impairment and should prompt respiratory assessment when is found. Longitudinal studies should clarify whether the relationship between a low Hb and restrictive spirometry is tracked over time and its prognostic meaning in African patients with SCA.