Methods: A purposive design in combination with a convenience sampling approach was used to recruit 24 women ages 26 to 55 of Malay, Indian, and Chinese descent across 15 university departments in Kuala Lumpur, Malaysia. Audio recordings of the individual semi-structured interviews were transcribed and analyzed using grounded theory.

Results: Women identified two important life transitions, getting married and having children, as particularly influential in changing eating behavior. Women reported a desire to eat healthier that was in tension with pressures to cater to the taste preferences of their household members and to accommodate their work schedules. Persistent social norms of eating in group settings and difficulty in accessing foods perceived as nutritious were reported as barriers to changing individual eating behavior. Regardless of education level and marital status, women agreed that the act of eating was closely connected with cultural values and embodied important meanings that took precedence over eating as a health-promoting behavior. While all participants expressed a desire to eat healthier, many reported limited confidence in their ability to consistently give up familiar and tasty foods for healthier alternatives. Shifts in eating behavior such as trying new healthy recipes and adopting a more restrictive diet were reported as most feasible and personally applicable after learning about a family member's declining health.

Conclusions: In this urban, multiethnic population, increasing the accessibility of nutritious foods and changing cultural perceptions of the relationship between food and nutrition may be important for enabling healthy eating behaviors.

METHODS: Consecutive biopsy-proven NAFLD patients (n = 191) and healthy controls (n = 188) were enrolled and genotyped for HLA-DQA1 and HLA-DQB1 alleles using the sequence-specific oligonucleotide-polymerase chain reaction method.

RESULTS: No association was found between the HLA alleles and NAFLD or NASH in a case-control setting. Nevertheless, among NAFLD patients, the frequency of HLA-DQB1*06 allele was significantly the lowest in NASH with significant fibrosis (10.4%) and approximately similar for NASH without significant fibrosis (22.9%) and NAFL (22.5%) (P = 0.004). It is noteworthy that the association remains significant after correction for multiple comparisons (Pc  = 0.04). Multivariate analysis revealed that HLA-DQB1*06 allele is also associated with fibrosis score (P = 0.001); the result remains significant after correction for multiple comparisons.