Page 68 - GHES-1-2
P. 68

Global Health Econ Sustain                                         Stochastic modeling of age at menopausal



            considered as the proportion of survival. The following   Table 1. Observed and expected value of the number of
            procedures were used to construct the menopausal life table.  women by different distributions (2005)
              Thus, from the fitted results of the logistic and RGLLog   Age group  Observed  Expected number of women by
            models, it was assumed that F  is the empirical result of the         Normal  Weibull  LLog  Logistic
                                   x
            proportion of women having menopause at age x year or   Below 35  1    0.256  2.358  0.476  0.411
            less and S  = 1−F  is the survived proportion or not reached   35–37  2  0.943  2.614  1.002  1.995
                   x
                         x
            menopause at age x. From the concept of the actuarial life
            table technique, let l  be the cohort of women, followed   37–39  5    3.147  4.988  2.776  3.294
                             0
            by the number of women not having menopause at age x   39–41    7      7.940  8.898  6.894  7.279
            (l ) using the equation l = l  S . The probability of reaching   41–43  11  15.157  14.504  14.384  14.240
             x
                              x
                                   x
                                 x
            menopause from age x to x+1 year (q ) was computed as:  43–45  24     21.887  20.743  22.760  22.217
                                         x
                   F   − F                                     45–47       25     23.911  24.321  24.951  24.832
               q =  x+1 − F  x                         (15)    47–49       18     19.763  20.983  18.974  19.240
                x
                    1
                        x
                                                               49–51       10     12.358  11.308  11.065  11.101
              Other symbols such as  L  is the total number of
                                     x
            women-years having menopause by the cohort at age x, T    51–53  6     5.846  2.998  5.595  5.347
                                                          x
            is the total number of women-years having menopause by   53–55  5      2.793  0.284  5.123  4.043
            the cohort above age x, and e  is the mean expected age   α             -     13.638  20.562  -
                                    x
            of menopause at age x were calculated using the following   β          3.742  46.478  45.679  2.241
            formula as in the actuarial life table:            γ                  45.634    -    0.000  45.598
                   k(l + l  )     n                            χ 2                 2.601  5.811  2.379  1.661
                              x ∑
                          +
               L =   x   x k  , T =  L + L X1 +…+ L X n   (16)  NLL               −43.027  −41.016  −40.989 −39.086
                                                   +
                x
                                          +
                                     X
                       2
                                  x
                                                               AIC                90.054  86.033  85.977  82.172
              Where k is the number of years between years n to n+k.  BIC         90.849  86.829  86.773  82.968
                                   T
               T x+1  = T − L , and  e =  l  x         (17)    Table 2. Descriptive statistics of age at menopause data of
                                x
                     x
                         x
                                    x
                                                               2018 according to their occupation
            4. Results and discussion
                                                                Occupation of   Mean      No. of      Standard
            The distributional pattern of age at menopause of Nepalese   respondents      women      deviation
            women was fitted using the four probability models and the   Service  49.667    6          0.516
            fitted results along with different test statistics results are   Business  50.667  12     1.969
            presented in Table 1 for the first data set and Table 2 for the   Agriculture  49.968  31  2.273
            second data set. The logistic distribution was significantly   Housemaker  49.653  98      1.663
            better fitted with the menopause data as compared to other
            distributions. The observed value of AIC, BIC, SSQ, and χ    Others  49.000     7          0.817
                                                          2
            were found as the lowest for logistic distribution, and the   Total  49.766    154         1.785
            result of NLL was found as the highest for logistic distribution
            as compared to other distributions used in this research.  considered when girls have menarche. If the average age of
              The graphical representation between the observed and   menarche was considered as 15 years then the expected age
            expected frequencies of women at the age of menopause   of menopause was after 30.507 years of menarche.
            is  shown  in  Figure  1.  From  the  analysis  and  graphical   If the woman was 30  years old, the expected age of
            presentation, the logistic distribution was found to be a   menopause was estimated as after 15.565  years. The
            good fit so it was recommended to use this distribution for   probability of having menopause when the age of women was
            describing the age pattern of women at menopause.
                                                               35 years is 0.018 (1.8%) whereas when the age of women was
              After fitting the logistic distribution, the fitted value   45 years the chances of having menopause is 0.484 (48.4%).
            of cumulative distribution was used to compute the
            other components of the menopausal life table which   Several parameter probability distributions were also
            are shown in  Table  3. From the analysis, the expected   computed to better predict the waiting time for menopause
            age of menopause at the birth of girls was estimated as   and the probability of having menopause at a particular
            45.511 years. However, the expected age of menopause was   age (Table 4).
            Volume 1 Issue 2 (2023)                         5                        https://doi.org/10.36922/ghes.1239
   63   64   65   66   67   68   69   70   71   72   73