Mental Health Status among Married Working Women Residing in Bhubaneswar City, India: A Psychosocial Survey

Table 5

Results of binary logistic regression predicting the odds of mental health outcomes among women respondents (𝑁=216).

Independent variable	Dependent variable		“𝑃” value
	Mental health status (normal = 1, poor = 0)
	(𝛽)	Exp (𝛽)
Age group (in years)			0.203
20–30	−0.816	0.442	0.104
31–40	−0.765	0.466	0.101
41–50^R
Addiction of husband to alcohol/drugs
Yes	−0.690	0.502	0.061
No^R
Chronic illness of family members
Yes	−0.201	0.818	0.691
No^R
Workplace atmosphere			0.152
Conducive	0.735	2.086	0.115
Not conducive	−0.054	0.948	0.933
Neutral^R
Attitude of colleagues
Favourable	1.520	4.571	0.004
Unfavourable/indifferent^R
Attitude of husband/in-laws
Favourable	0.809	2.245	0.086
Unfavourable/indifferent^R
Sharing their own problems with husband
Yes	1.351	3.860	0.005
No^R
Experiencing any conflict
Yes	−0.509	0.601	0.169
No^R
Time for social obligation
Yes	0.661	1.937	0.063
No^R
Spending time for yoga/meditation/exercise			0.019
Nil	−1.932	0.145	0.018
1-2 times/week	0.867	0.420	0.365
>2 times/week^R
Participation in decision making process of family
Yes	0.553	1.738	0.207
No^R

Note: R: reference category; 𝛽: regression coefficient (log odds ratio). Exp (𝛽): odds ratio. The model fit statistics were as follows. Omnibus tests of model coefficients: 𝜒² value of 62.342 with a “𝑃” value of 0.000 tells that the final model as a whole fits significantly better than an empty model (model with no predictors). Hosmer and Lemeshow test: 𝜒² value of 3.062 with a “𝑃” value of 0.930 indicates that there is no significant difference between observed and predicted probabilities and thus the model fits.