TABLE 4.
Likelihood of Stopping Dental Care Use Among Older Adults: Health and Retirement Study, United States, 2004 and 2006
| Population Characteristic | Unadjusted OR Point Estimatea (95% CI) | Adjusted OR Point Estimateb (95% CI) |
| Age, y | ||
| 51–64 | 0.685** (0.594, 0.791) | 1.014 (0.818, 1.257) |
| 65–69 | 0.628** (0.514, 0.768) | 0.757* (0.607, 0.945) |
| 70–74 | 0.818* (0.670, 0.999) | 0.969 (0.783, 1.199) |
| ≥ 75 (Ref) | 1.000 | 1.000 |
| Gender | ||
| Women | 0.875* (0.786, 0.974) | 0.654** (0.571, 0.748) |
| Men (Ref) | 1.000 | 1.000 |
| Ethnicity/race | ||
| Black non-Hispanic | 3.025** (2.455, 3.728) | 2.052** (1.624, 2.593) |
| Hispanic | 2.897** (2.152, 3.900) | 1.540* (1.059, 2.238) |
| Other non-Hispanic | 1.943** (1.303, 2.898) | 1.747** (1.157, 2.639) |
| White non-Hispanic (Ref) | 1.000 | 1.000 |
| Family income by poverty statusb | ||
| Poor | 5.894** (4.687, 7.412) | 2.390** (1.865, 3.064) |
| Low income | 3.925** (3.348, 4.601) | 2.086** (1.758, 2.476) |
| Middle income | 2.384** (2.016, 2.818) | 1.630** (1.360, 1.953) |
| High income (Ref) | 1.000 | 1.000 |
| Education | ||
| < high-school degree | 7.532** (6.007, 9.445) | 2.815** (2.142, 3.701) |
| High-school graduate | 2.669** (2.185, 3.261) | 1.817** (1.467, 2.252) |
| College graduate (Ref) | 1.000 | 1.000 |
| Marital status | ||
| Widowed or divorced | 1.770** (1.513, 2.069) | 1.342* (1.027, 1.752) |
| Never married | 1.129 (0.787, 1.619) | 0.936 (0.588, 1.490) |
| Married (Ref) | 1.000 | 1.000 |
| Household size | ||
| 2 | 0.669** (0.572, 0.782) | 1.138 (0.880, 1.471) |
| ≥ 3 | 0.974 (0.805, 1.179) | 1.370* (1.023, 1.836) |
| 1 (Ref) | 1.000 | 1.000 |
| Loss of permanent teeth | ||
| All missing | 8.982** (7.351, 10.974) | 5.931** (4.830, 7.284) |
| None missing (Ref) | 1.000 | 1.000 |
| Health status | ||
| Good | 1.775** (1.511, 2.086) | 1.363** (1.138, 1.633) |
| Fair or poor | 3.340** (2.805, 3.978) | 1.723** (1.430, 2.077) |
| Excellent or very good (Ref) | 1.000 | 1.000 |
| Retirement or labor force status 2004 and 2006 | ||
| Entered full retirement 2006 | 1.391* (1.060, 1.825) | 0.832 (0.633, 1.094) |
| Entered partial retirement 2006 | 1.115 (0.724, 1.716) | 0.950 (0.586, 1.541) |
| Entered labor force 2006 | 1.562 (0.956, 2.555) | 1.185 (0.766, 1.831) |
| Entered not in the labor force, not retired 2006 | 2.221** (1.511, 3.263) | 0.909 (0.594, 1.390) |
| Always fully retired, 2004 and 2006 | 1.457** (1.223, 1.735) | 0.792* (0.634, 0.989) |
| Always partially retired, 2004 and 2006 | 0.832 (0.648, 1.069) | 0.686* (0.500, 0.941) |
| Always not in the labor force, not retired, 2004 and 2006 | 1.865** (1.375, 2.530) | 0.992 (0.713, 1.38) |
| Always in the labor force, 2004 and 2006 (Ref) | 1.000 | 1.000 |
| Dental coverage status 2004–2006 | ||
| Always covered, 2004–2006 | 0.456** (0.385, 0.540) | 0.533** (0.440, 0.646) |
| Lost coverage 2006 | 2.225** (1.840, 2.691) | 2.230** (1.834, 2.710) |
| Gained coverage 2006 | 1.445** (1.093, 1.912) | 1.348 (0.964, 1.886) |
| Never covered, 2004–2006 (Ref) | 1.000 | 1.000 |
Source. RAND Health and Retirement Study Data, Version H.10
Note. CI = confidence interval; OR = odds ratio. Pseudo R2 = 0.176 in adjusted regression. The sample consists of 15 787 observations, 10 116 in the stopping-dental-use equation and 5671 in the starting-dental-use equation. An original sample of 18 469 persons was reduced by 1514 persons with 0 weights, 547 persons with positive weights but not present in both 2004 and 2006, and 621 persons with missing values for any of the variables. Logistic estimates incorporated adjustments for the sample weights and the sample design. Unadjusted estimates did not control for other characteristics of the individual. Adjusted estimates included controls for other explanatory variables in the logistic equation.
Odds ratio point estimate = estimate of [Probability of dropping use/Probability of always using] for persons with row characteristic divided by [Probability of dropping use/Probability of always using] for reference group (ref). Unadjusted point estimates computed directly from Tables 1 and 2 differ slightly from the unadjusted logistic estimates in this table because of the smaller sample size for the logistic estimates.
Low income refers to persons in families with incomes 101% to 200% of the poverty line; middle income, 201% to 400% of the poverty line; and high income, > 400% of the poverty line. Poor persons are at or < 100% of the poverty line including persons in families with negative income.
*P = .05; **P = .01.