Current study aimed to reveal independent risk factors of CC among children aged 0–14 years old in Armenia, thus filling an important research gap in this area. The study identified three important predictors of CC: maternal folic acid usage during pregnancy as a protective factor and two strong risk factors of CC—history of induced abortions before the pregnancy and trauma-induced stress during the pregnancy.
Similar case-control studies in other countries had comparable study populations in terms of patients’ sex distribution with male predominance, and cancer type distribution with leukemias, lymphomas and brain tumors being the most common types of CC [12, 13]. Similarly, CC cases were more commonly living in urban areas in other studies as well, and the maternal age range in this study was comparable to those in other case-control studies [13, 14].
The protective effect of maternal folic acid usage before and during pregnancy on the risk of development of childhood cancer identified in this study was consistent with the literature, as many studies found that taking folic acid before and during pregnancy was significantly inversely associated with the cancer risk (leukemia, brain tumor) in a child [15]. Indeed, several studies from the United States and Canada recorded a decrease in childhood cancer incidence (neuroblastoma, Wilms tumor and primitive neuroectodermal tumors (PNET)) after folic acid fortification of food [16]. Folate is considered as an essential nutrient for the cell multiplication (as a coenzyme for DNA synthesis) and cell homeostasis (metabolism and regeneration) [17]. According to WHO standards for maternal and neonatal care, mothers should start taking 400 µg folic acid daily two months before getting pregnant and continue taking it until the 12th week of pregnancy [18]. This study adds to the existing evidence on the importance of following this recommendation.
According to this study findings, the history of induced abortions before pregnancy increased the risk of childhood cancer almost threefold. This finding is consistent with the findings of a number of studies indicating that the past history of fetal losses including induced abortions, miscarriages and stillbirths is associated with a higher incidence of cancer (leukemia, neuroblastoma, soft tissue sarcoma) in children [19, 20]. However, some studies report lack of such association or insignificant associations [21, 22]. A matched case-control study conducted by Children's Cancer Group and the Pediatric Oncology Group identified two times higher risk of neuroblastoma in children whose mothers had a history of two or more previous induced abortions [20].
The history of maternal trauma-induced stress during pregnancy was a risk factor for developing childhood cancer in the offspring in this study. A population-based cohort study conducted in Denmark and Sweden that included 39,002 children born to women who had psychological stress (parental death) during pregnancy, identified a similar link between this experience and the risk of cancer in the offspring. In particular, the risk was increased for leukemia (standardized incidence ratio (SIR), 1.49; P = 0.004), testicular cancer (SIR, 1.80; P = 0.02) and colon cancer (SIR, 3.95; P = 0.003) before the age of 15 years [23]. There are studies indicating the role of the stress during pregnancy for increasing the risk of childhood cancer [24]. There is a proposed hypothesis explaining the pathogenesis of a stress-induced cancer by the mediated inhibition of the enzyme, which initiates activation of cancerous cells to get under immune surveillance, leading to the newly generated malignant cells to continue growing without being checked by the immune system [23]. This mechanisms could explain the findings of a cohort study conducted in Israel, which observed higher cancer incidence (for leukemia, lymphoma and melanoma) among bereaved Jewish people [25]. However, the mechanism underlying the association between maternal stress during pregnancy and carcinogenesis in the offspring is not yet clearly understood.
One of the strengths of this study was that both cases and controls were identified from the same data source, which helped to reduce the selection bias, as the participants were coming from the same base population with potentially similar exposures. Both cases and controls were selected from the same hospital, which is the only place in the county where children with hematologic and oncologic diseases can get specialized diagnosis and treatment. This means that the patients represented the entire country with the results generalizable to total under-14 years old population in Armenia.
This study did not identify significant associations between child’s second-hand smoking and childhood cancer, or other well-known environmental exposures confirmed as risk factors for childhood cancer, such as pesticides, chemicals, radiation or others. The underlying reason for this could be the small size of the studied sample, which was the major limitation of the study overall and the wide diversity of diseases with potentially different sets of risk factors combined in a heterogeneous group of CC. Recall bias could be a potential limitation of this study due to long interval between the onset of the child’s disease and the interview. To minimize this bias, we selected participants who were diagnosed or received CC treatment relatively recently. The telephone interviews were conducted by six trained interviewers; however, they were aware of the participant’s case and control status which could introduce an interviewer bias. There was also a likelihood that mothers of cases might remember some events, which mothers of controls might disregard or had forgotten. Non-coverage bias could be a potential threat to the generalizability of the study findings, as patients with missing contact information or being not registered in PCBDCA (e.g., patients with CNS tumours) could somehow differ from the participants with the contact information available, although it is unlikely that the difference, if existing, could seriously alter the study findings. Furthermore, twenty-three patients diagnosed with cancer in the PCBDCA during the same period (2017–2020) who did not survive by the time of the study were excluded from the survey because of ethical considerations of conducting interviews with their mothers. This could have influenced the study findings if the non-survivors were different than the study cases. However, the analysis of the available characteristics of the non-survivors demonstrated that they did not differ from the survivors in terms of age, sex, residency, and the cancer type. Inclusion of participants with different types of cancer as cases is another limitation of this study, as there are cancer-specific risk factors, which cannot be identified with this approach.
A number of important recommendations could be derived from the findings of this study. First, they add another important argument for folic acid supplementation before and during pregnancy, which is preventing CC. Although folic acid supplementation of pregnant women is currently recommended by the Ministry of Health of Armenia to prevent neural tube defects, there are issues with its timing and coverage [26]. Hence, efforts should be undertaken to achieve timely and universal coverage of all pregnant women with this supplementation. Also, this study highlights the need to increase the awareness among women about the importance of using safe and effective methods of birth control and family planning to avoid induced abortions. Education of reproductive age women and their family members on the importance of stress reduction during pregnancy is another important area for improvement. To investigate childhood cancer-specific risk factors, large-scale studies with a larger sample size of children with homogeneous group of CC are recommended.