|Home | About | Journals | Submit | Contact Us | Français|
ObjectiveTo examine negative affectivity and problem-solving abilities for lone mothers and those who are married/partnered subsequent to a child's diagnosis with cancer.MethodsNegative affectivity and problem-solving strategies were assessed for 464 mothers (87 lone and 377 married/partnered) within 2–16 weeks of their child's diagnosis with cancer.ResultsThe two groups of mothers did not differ significantly on measures of perceived posttraumatic stress or problem-solving; lone mothers reported significantly more symptoms of depression. This difference was no longer significant when maternal education was taken into account.ConclusionsNegative affectivity and problem-solving abilities were similar for lone mothers and those that are married/partnered shortly after their child has been diagnosed with cancer. Findings are discussed within the context of contemporary strategies to assess marital status as proxy variable for various underlying constructs.
The incidence of childhood cancer in the United States was estimated to be 16.2 per 100,000 children in 2003; the 5-year survival rate has consistently improved over the past 30 years (U.S. Cancer Statistics Working Group, 2007). This increase in survival rates has resulted in work examining the quality of life for parents following a child's diagnosis with cancer (Fummeler, Mullins, Van Pelt, Carpentier, & Parkhurst, 2005). This research examines maternal distress shortly after a child's diagnosis with cancer, a point in time that is particularly important given the critical, complex, and immediate decisions families make regarding treatment options for their child (Sahler et al., 2005).
Recent research has estimated that nearly 3 in 10 children live in single-parent homes (Shudy et al., 2006). While the majority of parents report increased distress following a traumatic event such as a child's diagnosis with cancer, it seems feasible that this distress might be greater for mothers who are lone parents. A National Institutes of Health (NIH) “think tank” recently delineated gaps in current research relevant to single-parent families of children with chronic illness (Brown et al., 2008).
There is a small literature investigating how single mothers fare as compared to their married counterparts in other areas. Cairney, Boyle, Offord, and Racine (2003) found that single mothers in general, compared to married mothers, had higher levels of chronic distress, experienced more negative life events, and reexperienced more of their own childhood adversities (also see Copeland & Harbaugh, 2005). Problem-solving abilities have been linked to distress (Sahler et al., 2005). To our knowledge, no study has compared the problem-solving skills of lone mothers to those that are married/partnered.
The aim of this work was to assess distress and problem-solving skills in mothers of children recently diagnosed with cancer. We hypothesized that mothers who were lone parents would report more distress than their married/partnered counterparts. In addition, we explored the problem-solving skills of mothers who were lone parents compared to the problem-solving skills of married/partnered mothers. The relationship between problem-solving skills and negative affectivity was also explored.
Participants were 464 mothers (87 lone and 377 married/partnered) whose children were recently diagnosed with any type of cancer. Participants were recruited 2–16 weeks following their child's diagnosis and were part of a larger intervention study (Sahler et al., 2005). This work is based solely on baseline data that were obtained prior to the initiation of any behavioral intervention. A cross-sectional design was chosen to compare mothers at an important time point, just after diagnosis and before behavioral treatment was initiated for mothers.
A demographic questionnaire elicited demographic information. Mothers were also asked to select their marital status from the following choices: married, single, divorced, separated, widowed, remarried, or unmarried, living with a partner. Furthermore, 100% agreement was reached for the assignment to the “lone” (single, divorced, separated, widowed) or married/partnered (married, remarried, unmarried living with partner) group based on mothers’ responses to the marital status question in a group of five researchers.
Three measures were given, the Profile of Mood States (POMS; McNair, Lorr, & Droppleman, 1992), Beck Depression Inventory—II (BDI-II; Beck, Steer, & Brown, 1996), and the Impact of Event Scale—Revised (IES-R; Weiss & Marmar, 1997). The POMS and BDI-II were used to assess mood; the IES-R was given to assess symptoms of perceived posttraumatic stress. Cronbach's α for the current sample on all measures were acceptable (.90–.97).
The SPSI-R is a 52-item instrument that measures five dimensions of problem solving (D’Zurilla & Nezu, 1990). A higher total score is reflective of better problem-solving skills. Cronbach's α for the current sample was.89.
Potential participants received a written description about the study and signed informed consent documents. Approximately 75% of mothers who were approached participated (Sahler et al., 2005). Reasons for not participating included lack of time (47%), feeling overwhelmed (14%), or not being interested (19%).
Three (3) one-tailed t-tests (α =.05) were used to evaluate differences in POMS, BDI-II, and IES-R total scores between lone mothers and married/partnered mothers. If total scores of the POMS, BDI-II, or IES-R were significantly different between the groups, exploratory analyses were planned to probe for differences on subscale scores.
Potential differences between lone versus married/partnered mothers on problem solving were examined using an exploratory two-tailed t-test (α =.05).
A power analysis (based on the POMS) using G*Power software (Erdfelder, Faul, & Buchner, 1996) determined that a sample size of 172 would be necessary to detect a medium-effect size (d =.5) at a reasonable level of power (.9) for a two-tailed t-test. Because initial comparisons (t-tests) were few in number (4) and were followed by analyses to determine what factors influenced differences, if they occurred, no correction for multiple comparisons was performed.
Demographic characteristics are shown in Table I.
Independent samples t-tests revealed no significant differences between lone and married/partnered mothers in one measure of negative mood or perceived post traumatic stress (POMS, IES-R; Table II). However, lone mothers did report significantly higher scores (negative mood/depression) on the BDI-II than married/partnered mothers. Interestingly, but not statistically significant, lone mothers showed lower scores on the IES-R than married/partnered mothers; lone mothers reported fewer total posttraumatic stress symptoms. Because the total scores for the POMS and IES-R were not significantly different between the two groups of mothers, no further analyses were conducted.
Overall, all mothers’ scores, regardless of marital status, were elevated compared to normative data. Using the BDI-II cut score guidelines, a majority of mothers (50.6%) in the current study showed minimal depression, 21.9% were mildly depressed, 17.2% were moderately depressed, and 9.1% were severely depressed.
Problem-solving skills of lone mothers were not significantly different compared to married/partnered mothers, t(461) = –1.070, p =.285, although they did have a lower mean score on the SPSI-R, suggesting that they may use more negative and fewer positive problem-solving strategies.
BDI-II (r = –.42, p <.001), POMS (r = –.35, p <.001), and IES-R (r = –.37, p <.001) scores were significantly and negatively correlated with SPSI-R scores. Mothers with more positive problem-solving skills reported better mood and fewer symptoms of perceived post traumatic stress.
Independent samples t-tests revealed significant differences in maternal education, t(457) = –2.11, p <.05 and maternal age, t(459) = –2.08, p <.05 between lone and married/partnered mothers, with lone mothers having a lower mean grade level completed (Mlone = 12.14; Mmarried/partnered = 12.98) and being younger (Mlone = 33.86 years; Mmarried/partnered = 35.71 years). Analyses of covariance (ANCOVA) were conducted to control for maternal education and age when examining differences in depression scores on the BDI-II.
When controlling for highest maternal grade completed, the overall F test of the univariate ANCOVA was significant, F(2, 458) = 9.61, p <.001. The covariate maternal grade was also significant, F(1, 458) = 14.59, p <.001. However, when maternal grade level was controlled, BDI-II scores were no longer significantly different between lone and married/partnered mothers F(1, 458) = 3.13, ns. Variance accounted for by this solution was 4.1%.
When controlling for maternal age, the overall F test was significant, F(2, 460) = 4.64, p <.01. The covariate maternal age was significant, F(1, 460) = 4.34, p <.05. Furthermore, BDI-II scores were still significantly different between lone and married/partnered mothers, F(1, 460) = 4.06, p <.05, indicating that maternal age and marital status accounted for different variance in BDI-II scores. Variance accounted for by this solution was 2.0%.
Lone mothers did not self-report more negative mood (POMS), symptoms related to trauma (IES-R), or differences in problem-solving strategies (SPSI-R). Lone mothers did report significantly more symptoms of depression (BDI-II). For both lone and married/partnered mothers, all scores on negative affectivity were elevated.
Our measure of lone parenting was associated with a range of demographic variables, especially maternal education. Although lone mothers reported significantly more symptoms of depression (BDI-II), when education was statistically controlled, the difference between the groups in depressive symptoms was no longer significant. Thus, it appears that education seems to be an important factor when assessing for depressive symptoms. However, although differences in education (and age, for that matter) were statistically significantly different between groups, their qualitative meaning is unclear given that the differences were quite small. In addition, there was considerable cultural diversity in our sample, as there were Israeli participants; these could be areas for future research.
Mothers in the current study did report more distress than would be expected given normative data; however, there were few differences in distress levels based on our measure of marital status. It is noteworthy that standard deviations obtained for measures of distress were larger than those of normative samples. This may reflect the variability in how mothers react to their child's diagnosis of cancer.
Methodological issues with the measurement of marital status must be considered when interpreting these findings. As noted in Brown et al. (2008), there is a major difference between a single mother who receives extensive family support and a single mother who lives alone with her children with little support. Furthermore, there is also a difference between a single mother who receives extended family support and a married/partnered mother who receives no extended family support and may have a conflicted relationship. Although the commonly utilized method for assessing marital status is a forced-choice scale measured only once at the outset of a study, this strategy is not an optimal approach for measurement of the underlying issues of interest. The current forced-choice format might be followed with questions related to length of relationship, support within the primary relationship, or social support more broadly. Problem-solving skills should also be addressed.
It must be argued that our finding of few differences in negative affectivity between lone and married/partnered mothers is consequential, given the attention that the NIH has focused on this topic. The present research is an initial examination of this issue. Future work can utilize a more comprehensive approach to measure marital status and the marital relationship to make more conclusive statements. It may be the case that measures of relationship status, family functioning, problem-solving skills, or general social support may be more central to key issues than our current forced-choice measures (see Noll et al., 1995). Until we can ascertain what is being measured by the forced-choice method of choosing a category of marital status and what variables we believe are most cogent, we will not completely understand the effect of lone versus married/partnered parenting on children with chronic illness.
National Cancer Institute; National Institutes of Health (R25 CA65520).
Conflict of interest: None declared.