Gestational Diabetes Mellitus (GDM) is glucose intolerance diagnosed in pregnancy (Metzeger, 1991), affecting up to 5% of Caucasian pregnancies (Lancet, 2008). The risk of developing GDM is higher in those of South Asian decent (Savitz, Janevic, Engel, Kaufman and Herring, 2008).
Several factors appear to contribute to developing GDM not least dietary behaviour. That is, a diet that is high in fat increases the risk of Type two Diabetes mellitus and GDM. Pregnant women consume higher amounts of fibre and fat, with the consumption of fat being higher than recommended for non-pregnant women (Verbeke & Bourdeaudhuij, 2007). Also, Kieffer, Sinco and Kim (2006) found women with GDM do not meet dietary guidelines, suggesting the importance of diet on GDM.
The occurrence of GDM is also mediated by the amount of exercise undertaken. Exercise is known to prevent the development of Diabetes Mellitus (DM) (Sigal, Kenny, Wasserman, Castaneda-sceppa & White, 2006). Smith, Cheung, Bauman, Zehle and McLean (2005) measured physical activity in women with recent GDM and found that only a small amount undertook a sufficient level of physical activity. Kim, McEwen, Kieffer, Herman and Piette (2008) also found suboptimal levels of activity in GDM sufferers. Less than one third of participants undertook vigorous physical activity for twenty minutes on three of more occasions per week.
Social support received by individuals is also central to the prevalence of GDM, not least because social support with household tasks and childcare can enable more time spent exercising (Wen et al., 2002). They found the most frequently repeated form of social support was verbal encouragement, followed by help with childcare and then assistance with household tasks. Also, over half received no social support which interfered with physical activity, suggesting the amount of social support affects levels of GDM.
Self efficacy is the perception that one can perform a particular action (Bandura, 1989) and impacts on DM because of the ability to control ones behaviour can lead to a better control of glycemic levels (Williams, McGregor, Zeldman, Freedman & Deci, 2004). They found that autonomy and competence are important in effective DM management and better glycemic control highlighting the importance of self efficacy in levels of DM because the perception that an individual can perform a particular behaviour such as a controlled exercise regime effects whether this behaviour is carried out. Self efficacy is also related to physical activity and quality of diet, in that those with high self efficacy for exercise, undertake more exercise (Smith et al., 2005; Kim et al., 2008).
Those who cope and understand their illness can potentially show an improvement in their illness manifestation (Lange & Piette, 2006). GDM is also likely to be affected by appraisal of their condition. Illness cognitions are common sense beliefs about illness (Leventhal & Nerenz, 1985) and provide schema for coping and understanding illness which can possibly effect physical and mental functioning in those with DM (Paschalides et al., 2004; Edgar & Skinner, 2003) and can be applied to GDM. Individuals who understand their illnesses are able to cope with their illness showed increased care seeking behaviour (Lawson, Lyne, Bundy & Harvey, 2006).
Risk for developing GDM is increased fro those of South Asian decent, with the highest rates being evident in Bangladeshi (21.2 percent) and Pakistani (16.2 percent) women (Savitz, Janevic, Engel, Kaufman & Herring, 2008), which may be due to cultural differences. Lawton, Ahmad, Peel and Hallowell (2007) found a difference in attitudes of Caucasian and Asian people with DM; Asian people attributed their condition to external factors such as life events and migration to Britain whereas Caucasian individuals with DM emphasised their role in DM. Stone, Pound, Pancholi, Farooqi and Khunti (2005) consider this difference as being due to diet. The South Asian diet is typically high in fat and sugar (Stone et al., 2005) and appears ingrained in culture and difficult to change.
Attitudes towards formal exercise highlight cultural differences. King, Law and Donaldson (as cited in Khunti, Kumar & Brodie, 2009) notes that many South Asian groups, including Bangladeshi and Pakistani people, view formal exercise negatively. They believe going to the gym is a source of ridicule, which receives support from Greenhalgh, Helman and Chowdhury (1998) who found that, among Bangladeshi men with DM, exercise was viewed as pointless and worsening their illness. Also, Bean, Cundy and Petrie (2006) noted differences in illness perceptions among Europeans, South Asians and Pacific Islanders which may account for differences in levels of GDM, highlighting cultural differences in illness perceptions.
RESEARCH AIMS
It can be inferred from the previous literature that despite the extensive research on GDM levels, there has been little research in looking at the differences in levels of GDM in South Asian and Caucasian women. This study aims to establish an association between illness perceptions, dietary behaviour, physical exercise, social support and self efficacy and the prevalence of GDM, and identify which factor is the greatest predictor of levels of GDM in Asian women. These factors have been found to effect levels of DM and GDM in Caucasian people (Kim et al., 2008; Smith et al., 2005) but these factors have not been yet been explored to explain the increased levels of GDM in South Asian women.
It is hypothesised that those who reported greater self efficacy and social support for exercising would report greater activity. It is also hypothesised that those who report greater social support for healthy eating would report a better quality diet. Also, it is hypothesised that those who report greater physical activity and a better quality diet will also have a better illness representation than those who exercise less and have a poorer diet.
RESEARCH METHODOLOGY
Participants
Power analysis was undertaken based on previous research (Kim et al., 2008) and establish sample size necessary to achieve power (Cohen 1988). We anticipate a modest effect size for multiple regression t up to 8 predictor variables, power of 0.8, effect size (η2= 0.13), x = 0.08 and 100 participants needed. Participants will be either of Asian or Caucasian decent and have gestational diabetes. Participants will be self selected responding to an advertisement placed on various diabetes internet forums.
Design
A questionnaire design will be employed to examine social support from friends and family, self efficacy for exercising and illness representations. The questionnaires also measure the amount an individual exercises and the quality of diet. In addition to these, descriptive statistics will be calculated which will include the grouping variable ethnic background in order to determine which of the psychosocial factors contribute to GDM development.
Materials
The questionnaire will be compiled from several previously used questionnaires, using 'Survey Monkey'. To examine self efficacy for exercise, the Marcus, Rakowski and Rossi (1992) (See Appendix A) questionnaire will be used which has adequate internal consistency and test-retest reliability with a Cronbach α of .71. Social support will be measured using the questionnaire developed by Sallis, Grossman, Pinski, Patterson and Nader (1987) (See Appendix B). The negatively worded, sabotage, rewards and punishment items will be removed because they did not correlate with dietary measures or exercise measures in Sallis et al's. (1987) study. The scales have robust findings, with a Cronbach α of .89 for social support and physical activity from friends and .91 for social support and physical activity from the family. For social support from family and healthy eating, the Cronbach α was .87 and .89 for questions examining social support from friends for a healthy diet. To measure diet, the Michigan Healthy Diet Indicator (Rafferty, Anderson, McGee & Miller, 2002) (See appendix C) will be adapted. Exercise will be measured by using questions based on Kim et al. (2008) questions. These questions were developed from the National Health Interview Survey (Gregg, Gerzoff, Caspersen, Williamson & Narayan, 2003) (See Appendix D). To measure illness representations, the Diabetes Illness Representations Questionnaire (Skinner et al., 2002) (See Appendix E) will be used, which has good internal consistency and construct validity. One potential limitation with using an online questionnaire is that only individuals who have access to the internet can take part and individuals who can read English, thus preventing participants who can not read English from taking part.
Procedure
Participants will be made aware of the aims of the study at the beginning (via the consent form, see Appendix F). After informed consent is obtained, participants complete the questionnaire, with instructions provided. When the questionnaire is completed, participants will receive a written debrief explaining the experiment and the expected results (See Appendix G).
Data Analysis
Preliminary correlations will be conducted to explore the potential associations between the psychosocial factors and characteristics of participants. Regressions will be conducted to explore potential associations between the psychosocial factors and the main outcome measures (self efficacy and social support). Multivariate regression models will then be constructed to establish which of the psychosocial factors are associated with the dependant measures identified that not due to participant characteristics. Significant correlations will be entered into several multiple regressions which will identify the greatest predictor of GDM.
