Adolescent survivors of childhood cancer are at an increased risk for bone-related morbidity,1
but evidence supporting interventions to promote bone health behaviors for this population, such as milk consumption, remains limited.9
The TTM has been widely used to inform behavioral interventions targeting dietary behaviors25
and the impact of interventions to promote milk consumption among young survivors could be enhanced by tailoring content to individual survivors’ SOC.10
Our findings provide preliminary support for the construct validity of a milk consumption SOC algorithm based on behavioral criteria, and highlight important topics for research.
Adolescent survivors who expressed readiness to change for consuming 2 and 4 daily servings of milk reported more frequent milk consumption, were more likely to drink 4 servings of milk a day, and consumed more dietary calcium. These findings are consistent with prior studies that sought to validate the algorithm for use in other populations, such as college-age young adults.14,15
We used the 24-hour recall method to measure dietary calcium consumption which is a significant methodological improvement over prior studies using less robust measures of calcium intake.14,15,22
The algorithm also incorporates several validity-enhancing strategies, including using simple behavioral goals, applying a clear SOC classification scheme, and specifying objective behaviors to avoid misclassification.26
We did not observe associations between SOC and knowledge or self-efficacy. This finding may be because our measures for these constructs were not based directly on the TTM, but were developed to capture broader bone health constructs.27
We also evaluated validity based on a combined SOC outcome for 2 and 4 daily servings of milk, rather than examining these behaviors separately. This may have affected our ability to detect differences based on theoretical constructs. It is critical that future studies demonstrate the validity of the algorithm relative to constructs derived more specifically from TTM, such as processes of change and decisional balance.26
Future studies can build from this work by examining milk consumption behaviors independently.
Our findings highlight other important considerations for future studies seeking to develop bone health behavior interventions targeting young cancer survivors. While milk is a primary source of dietary calcium, survivors may consume a variety of calcium-rich foods to improve bone health (e.g., leafy green vegetables). Young survivors in later SOC were more likely to meet recommendations for daily calcium consumption, but just over one quarter met recommendations. A comprehensive approach to bone health interventions that motivates young survivors to consume a variety of bone health-promoting foods and engage in healthful physical activity appears needed to reduce survivors’ risk for bone health morbidity.9
Future studies in this area may also warrant comparing self-report behavioral assessments to objective measures of bone health (e.g., bone density scans). That our assessment did not differentiate between types of milk (e.g., non-fat milk, whole milk) is an important consideration to address in future studies, as some commonly-consumed forms of milk may have negative health effects (e.g., milk flavored with sugar).7
Our findings should be interpreted in light of study limitations. The small sample restricted our ability to analyze construct validity separately for both milk consumption behaviors and limits generalizability of findings. The broad age range of survivors may have spanned distinct developmental periods, which could influence our findings. We relied on self-report assessments, some of which were not developed based on the TTM or were brief and adapted from prior work (e.g., milk consumption frequency).
Implications for Research and Practice
Despite study limitations, our findings provide preliminary support for the validity of the milk consumption SOC algorithm among adolescent survivors of childhood cancer relative to behavioral criteria. Additional research is needed to evaluate the construct validity of the algorithm relative to TTM constructs (e.g., processes of change) among larger, more diverse samples. Interventions that take a broad approach by motivating young survivors to consume milk and a variety of other healthful foods and engage in other bone health-promoting behaviors may be ideal to reduce the risk for bone health morbidity in this vulnerable population and could be informed using this SOC algorithm.