Introduction
Patients with head and neck cancer are known to be at risk of nutritional deficiencies as the result of a variety of factors, including mechanical obstruction to swallowing caused by the tumour, treatment side effects, cancer cachexia and pre-existing malnourishment. Whilst many measures of nutritional status are available, including weight, body mass index (BMI), skinfold thickness and upper arm circumference,Reference Frisancho1 hand-grip strength has increasingly been investigated as a simple and non-invasive anthropometric marker of nutritional status.Reference Norman, Stobäus, Gonzalez, Schulzke and Pirlich2
In states of malnourishment, protein is lost preferentially from muscle mass, to compensate for the reduced nutritional input.Reference Daniel, Pratt and Spargo3, Reference Heymsfield, Arteaga, McManus, Smith and Moffitt4 As muscle function is known to be associated with whole body protein,Reference Peng, Plank, McCall, Gillanders, McIlroy and Gane5, Reference Windsor and Hill6 body-cell massReference Norman, Schutz, Kemps, Josef Lubke, Lochs and Pirlich7 and BMI,Reference Norman, Schutz, Kemps, Josef Lubke, Lochs and Pirlich7–Reference Pieterse, Manandhar and Ismail9 loss of weight and muscle mass results in weakness and deteriorating function which hand-grip strength may be able to approximate.
The use of hand-grip strength may not only be limited to approximating nutritional status. It has been shown to be strongly correlated with post-operative complications,Reference Guo, Zhang, Ma, Zhang and Huang10–Reference Webb, Newman, Taylor and Keogh12 length of hospital stay,Reference Bohannon13, Reference Kerr, Syddall, Cooper, Turner, Briggs and Sayer14 functional status,Reference Kerr, Syddall, Cooper, Turner, Briggs and Sayer14, Reference Humphreys, de la Maza, Hirsch, Barrera, Gattas and Bunout15 short-termReference Vecchiarino, Bohannon, Ferullo and Maljanian16 and medium-termReference Bohannon, Maljanian and Ferullo17 survival, and dependency for activities of daily living.Reference Rantanen, Avlund, Suominen, Schroll, Frändin and Pertti18
Previous research has investigated hand-grip strength as a predictor of post-operative complications and hospital stay in patients with head and neck cancer.Reference Guo, Zhang, Ma, Zhang and Huang10 However, to our knowledge, no previous study has investigated its use in the routine monitoring of head and neck cancer patients' nutritional state, despite its inclusion in national guidelines for nutritional assessment.Reference Roland and Paleri19 As there is a correlation between weight and hand-grip strength in healthy individuals,Reference Chandrasekaran, Ghosh, Prasad, Krishnan and Chandrasharma20 we aimed to prospectively compare the weight loss and hand-grip strength of patients with head and neck cancer at baseline and 3, 6 and 12 months post-treatment to see if changes in weight correspond with changes in hand-grip strength.
Materials and methods
A prospective study of nutritional and swallowing support in patients receiving treatment for head and neck cancer was undertaken at four independent sites in the North of England Cancer Network (Cumberland Infirmary, Carlisle; James Cook Hospital, Middlesbrough; Freeman Hospital, Newcastle upon Tyne; and Sunderland Royal Hospital, Sunderland). Consecutive patients were recruited over a three-month period between June and September 2009.
Patients received nutritional assessments and had appropriate interventions at baseline and during treatment and follow up, as recommended by national guidelines.Reference Roland and Paleri19 Data on nutritional status, including weight and hand-grip strength, were collected at baseline, and 3, 6 and 12 months after treatment. Other patient details, including tumour and treatment demographics and World Health Organization performance status classification data,Reference Oken, Creech, Tormey, Horton, Davis and McFadden21 were obtained from the data submitted to the National Head and Neck Cancer Audit (‘DAHNO’). Hand-grip strength was measured in the non-dominant arm, with the patient standing and their elbow extended. The best of three measurements was recorded.Reference Todorovich and Micklewright22 Weights were recorded on approved scales in follow-up clinics.
Ethical considerations
Weight is recorded in all patients undergoing treatment and is thus an integral part of practice. Hand-grip strength is measured in two centres in the region and was rolled out to other centres for the purpose of the audit. The current study involved analysis of the existing data obtained at routine follow-up visits. Ethical approval was therefore not required.
Statistical analysis
Descriptive statistics were used for the following variables: age, gender, tumour site, tumour stage and treatment modality. Data on weight loss and hand-grip strength were analysed using repeated measures analysis of variance (ANOVA) (with Mauchly's sphericity test) and Bonferroni correction for pairwise comparisons. A p-value of 0.05 was considered significant (two-tailed test) unless post-hoc analyses were performed. Bonferroni adjusted alpha was used in post-hoc analysis. Statistical analysis was performed using the IBM® SPSS® Statistics software package, version 20.
Results
A total of 114 patients were recruited for the audit. The demographics for these patients are shown in Table I.
Table I Patient demographics
WHO = World Health Organization; RT = radiotherapy
Data on pre- and post-treatment weights and hand-grip strengths are shown in Tables II and III. Repeated measures ANOVA showed that weight loss was statistically significant (p < 0.003, with Mauchly's sphericity <0.05). However, changes in hand-grip strength were not statistically significant over the time period (p < 0.126, with Mauchly's sphericity <0.05).
Table II Pre- and post-treatment average weight
Pts = patients; SD = standard deviation
Table III Pre- and post-treatment average hand-grip strength
Pts = patients; SD = standard deviation
Further pairwise comparisons (Bonferroni correction applied) showed that weight loss was significant at both 3 and 6 month intervals, but not at 12 months. However, hand-grip strength was not significantly different at any of the time intervals (Table IV).
Table IV Pairwise comparison of pre- and post-treatment mean weight and hand grip strength
SEM = standard error of the mean; mth = months
Discussion
This prospective multicentre audit showed that the statistically significant amount of weight lost by head and neck cancer patients whilst undergoing treatment does not coincide with a drop in hand-grip strength as might be expected.
There are a few possible explanations as to why a fall in body weight may not have coincided with a fall in hand-grip strength in this study. Firstly, the weight loss seen in this study may have represented fat loss rather than lean mass. Given that changes in body fat composition have little effect on strength,Reference Goodpaster, Park, Harris, Kritchevsky, Nevitt and Schwartz23 hand-grip strength may not have been expected to fall in these patients. Furthermore, we did not routinely measure body fat composition in this study and we are unable to investigate this further. However, in healthy individuals, hand-grip strength is correlated less strongly with weight than other strength measures such as knee extension testing, probably resulting from the lower limb muscular strength being more dependent on weight for activities of daily living.Reference Bohannon, Magasi, Bubela, Wang and Gershon24 Despite this, hand-grip strength does show a correlation with lower limb muscular strength and we would have expected malnourished cancer patients experiencing weight loss such as those in this study to have a lower hand-grip strength.Reference Norman, Stobäus, Smoliner, Zocher, Scheufele and Valentini25
Another possible explanation for the lack of change in hand-grip strength could be related to it being a poor predictor of nutritional status and weight loss. Whilst hand-grip strength has been shown to improve in studies of nutritional intervention in malnourished patients with inflammatory bowel disease,Reference Christie and Hill26 tuberculosisReference Paton, Chua, Earnest and Chee27 and benign gastrointestinal disease,Reference Norman, Kirchner, Freudenreich, Ockenga, Lochs and Pirlich28 stroke patients,Reference Ha, Hauge, Spenning and Iversen29 and those undergoing surgery,Reference Beattie, Prach, Baxter and Pennington30 results are inconsistent in other populations such as the elderly.Reference Edington, Barnes, Bryan, Dupree, Frost and Hickson31–Reference Smoliner, Norman, Scheufele, Hartig, Pirlich and Lochs33 A meta-analysis of 10 studies failed to show any positive effect of active nutritional intervention on hand-grip strength in older people,Reference Milne, Avenell and Potter34 suggesting that hand-grip strength may be more of a reflection of function rather than nutritional status.Reference Milne, Avenell and Potter34 This is supported by data from the Danish Head and Neck Cancer ’DAHANCA 25’ study, where head and neck cancer patients were randomised to either progressive strength training and nutritional supplementation, or progressive strength training and placebo, post-treatment. Both groups had equal strength outcomes as measured by knee extension.Reference Lønbro, Dalgas, Primdahl, Overgaard and Overgaard35
Whilst clearly hand-grip strength may be an important outcome itself in terms of strength and function, it seems to be less important as a nutritional marker. This represents a potential need for change in the nutritional assessment of patients with head and neck cancer. Indeed, the relatively larger standard deviations seen in the values for hand-grip strength compared with body weight would suggest a larger natural variation in hand-grip strength than in weight. Therefore, it may not be a useful measurement for making inferences related to associated nutritional variables such as weight loss.
Hand-grip strength may also be affected by a number of other factors, such as psychological well-being and impairments in dexterity secondary to co-morbidities, which are independent of cancer treatment such as joint disease. It may also be affected by factors that are dependent on treatment, such as primary site, the extent of primary therapy and the nature of adjuvant treatment. However, the heterogeneity of treatments for head and neck cancer at different subsites would make further subgroup analysis of little value in this study.
Limitations
The Parenteral and Enteral Nutrition Group of the British Dietetic Association changed their recommendations regarding hand-grip strength testing after our data collection, and advised recording an average of three tests rather than a best of three tests (which is what they had previously advised).Reference Todorovich and Micklewright36 Hand-grip strength measurements were standardised in this audit and readings were consistent throughout data collection. Furthermore, there is evidence to suggest that there is no difference between measuring hand-grip strengths in these ways.Reference Innes37
• Hand-grip strength has been shown to correlate with post-operative complications, functional status and survival
• Hand-grip strength does not correlate with weight loss in head and neck cancer patients
• Hand-grip strength may provide a more useful reflection of functional state than nutritional state
Some patients did not undergo hand-grip strength or weight testing at all three time points. Although we therefore do not have a complete data set for all patients, we have a relatively large sample size and we did adjust for incomplete data in our statistical analysis.
Conclusion
Changes in weight do not correspond with changes in hand-grip strength in patients receiving treatment for head and neck cancer. Nutritional assessment should be multifactorial and hand-grip strength should not be relied upon to routinely monitor changes in nutritional status.
Acknowledgements
The authors are grateful to the North of England Cancer Network for supporting the audit data collection process. We would like to additionally acknowledge Alan Davidson (Dietetic Assistant) for his role in the collection of data.
